Similarity Analysis and the Translation Profile

Andrew Chesterman
University of Helsinki

Abstract
Similarity entails both sameness and difference, and translation research needs to keep
both these aspects in focus. There is a curious asymmetry in many current analyses of
translation equivalences and differences, in that pride of place is most often given to
differences. Two key notions for the understanding of similarity in translation are the
nature of similarity as a multi-place predicate, and the difference between divergent
and convergent similarity. Similarity judgements also underlie quality assessment. A
proposal is made for a general framework for the similarity analysis of a translation
profile, i.e. the linguistic form of a translation. Such an analysis makes reference at
least to the source text, non-translated texts in the target language, other translations,
and learner texts. Statements about similarities are descriptive statements; but as
formulations of similarities become more abstract, showing relations across broader
systems and different fields, they become increasingly explanatory in nature.

1. Sameness and difference

In Translation Studies (TS), similarity has been used primarily as a way of
conceptualizing the relation between source text (ST) and target text (TT),
usually known as equivalence. This relation is at the textual centre of TS.
Originally, the focus was on “sameness”: recall the early debates about the
necessarily literal nature of the translation of sacred texts. True, this sameness
is seldom, if ever, absolute. Theories of different kinds of equivalence have
been developed by Catford (1965), Nida (1964), Koller (1979) and others
(including Chesterman 1998, who argues for an interpretation of the concept as
“relative similarity”). Equivalence is of course also a central concept in
terminology studies and in practical terminological work.
But more recently, the focus has been on difference: much attention has
been given to the notions of shifts and/or strategies, both of which imply
changes, non-identity. The very notion of equivalence has been under attack
54 ANDREW CHESTERMAN

(see e.g. Pym 1995 for a defence of it). The status of the differences that
translators make (rather than the samenesses that they seek to preserve) is also
emphasized by the so-called Manipulation School (see e.g. Hermans 1985), the
functionalists such as skopos theory (e.g. Vermeer 1996), target-oriented
approaches (e.g. Toury 1995), and more recently by Venuti’s use of the
concept of the remainder (following Lecercle 1990). Differences are often seen
as revealing things that are (inevitably) “lost in translation”, but Venuti’s
interpretation of the remainder is expressed in terms of translation gains:
Translators [...] can never entirely avoid the loss that the translating process
enforces on the foreign text, on its meanings and structures, figures and
traditions. And translators cannot obviate the gain in their translating, the
construction of different meanings, structures, figures and traditions and there-
by the creation of textual effects that go far beyond the establishment of a
lexicographical equivalence to signify primarily in terms of the translating
language and culture. Following the important work of Jean-Jacques Lecercle
(1990), I call these effects the ‘remainder’ in a translation. (Venuti 2002: 219)
The remainder is here defined in terms of textual effects. Does this mean actual
textual features? Venuti goes on to speak of the remainder as consisting of
“linguistic forms and textual effects” (Venuti 2002: 219), but later of the ways
in which the remainder can be “released”, intentionally or otherwise (2002:
220). The remainder remains a complex concept, and not entirely clear, but we
do not need to go further into it here. My main point is Venuti’s positive focus
on difference. Looking for similarities means taking account of both
“samenesses” and differences. The two are logically related, in that we proceed
up a conceptual hierarchy, in the direction of synthesis, by defining same-
nesses; and we proceed downward, in the direction of analysis, by defining
differences.

2. The translation profile

I now introduce the term “translation profile”. I define this as the linguistic
form of a given translation (or set of translations), comprising all its relevant
textual features, as described in comparison with the linguistic profiles of other
relevant texts. A translation profile is thus described in terms of textual
relations. In other words, it is the sum total of a translation’s similarities
(including implied differences) with other relevant texts. What are the other
relevant texts?
In the first place, with respect to the ST. This aspect of the translation
profile is covered by a whole range of equivalence-types and shifts or
strategies/techniques/procedures, which have been defined and classified in
 SIMILARITY ANALYSIS AND THE TRANSLATION PROFILE 55

many ways. This is familiar ground to all translation scholars, but I will
introduce a small illustrative set of data at this point, to which we will return as
the discussion proceeds. The data are the multilingual label on a pair of Giorgio
Ricci shoes, made in Italy, bought in Finland in 2005. I assume the source
language is Italian; the text of the label is reproduced below, with the original
layout.

La speciale lavorazione Die spezielle

a sacchetto “Sacchetto-Beutenaht”
e la talloneria anatomica, und das anatomisch geformte Fussbett sind
frutto di una lunga esperienza artigianale das Ergebnis einer langjährigen
e di una raffinata tecnica produttiva, Handwerkstradition sowie einer weitentwickelten
conferiscono alle nostre calzature Produktionstechnik. Hierdurch erhalten die
una morbidezza senza confronti. Modelle einen unvergleichbaren Tragecomfort.
Il piede che le indossa Somit kann sich der Fuss maximaler
può muoversi con la massima libertà, Bewegungsfreiheit bei gleichzeitig höchster
valorizzando la sobria eleganza Eleganz erfreuen.
dei modelli.
The special Le montage spécial
“Sacchetto” stitching “façon sac (sacchetto)”
and the anatomic heel support et la semelle anatomique,
are the result of a long experienced issus d’une longue expérience artisanale,
workmanship and refined production skills donnent à ces chaussures
and make our shoes une souplesse sans égal.
specially soft to ware so that feet Entièrement cousues et fabriquées à la main
can move comfortably adding value avec des peaux de qualité,
and style to the shoes. elles suivent la forme et le mouvement
de vos pieds assurant un confort maximum.

A literal gloss-translation of the Italian: ‘The special working / in “little bag” style /
and the anatomical heel, / fruit of a long handicraft experience / and of a refined
productive technique, / confer to our shoes / a softness without rivals. / The feet which
put them on / can move in the maximum liberty, / giving value to the sober elegance /
of the models.’

Looking just at the German target text here, we can note several obvious
samenesses (equivalences) of content and style, such as the transfer of the key
term sacchetto. There are also obvious shifts, such as a new sentence break
(Hierdurch...), many examples of transposition (può muoversi > Bewegungs-
freiheit) and other structural and semantic shifts (frutto > Ergebnis), including
what we could call explicitation (> gleichzeitig) and implicitation (che le
indossa >). This facet of the translation profile can thus be described as a list of
the equivalences and shifts observed with respect to the source text.
56 ANDREW CHESTERMAN

3. Asymmetry: does it matter?

It is a curious fact that ST has developed much more detailed taxonomies and
terminologies of difference than of sameness. On the sameness side, we have
rather simple classifications of kinds of equivalence: for instance, two types in
Nida, five in Koller. But on the difference side, we have much more rich and
complex proposals for taxonomies of shifts etc. (see e.g. van Leuven-Zwart
1989/1990, Molina and Hurtado Albir 2002, and the critical survey in Chester-
man 2005).
Why this focus on difference? Because of the trope of loss in our Western
tradition, after Babel? Is difference intrinsically more interesting? Because
difference runs counter to our basic expectation of non-difference? Do people
tend “to assume the essential likeness of items that are ostensibly related”
(Wendland 2004: 340), and thus focus on features that do not conform to this
initial assumption? Translations are supposed to be “the same as” the original,
and so when we realize they are not, and cannot be, we then look at the
differences. But in terms of gaining a balanced picture of the overall translation
profile, our perspective will be rather one-sided if we focus more closely on
differences than on samenesses.
The terminological jungle itself is daunting (see Chesterman 2005): we
find differences classified in terms of shifts, strategies, procedures, techniques,
solution-types, methods... Here I will by-pass this terminological debate and
focus just on shifts, defined as textual differences between ST and TT (i.e. as
the textual results of strategies or techniques). >From the point of view of the
translation profile, it would be worth developing parallel taxonomies of
sameness and difference, as illustrated in Table 1 on the next page:
 SIMILARITY ANALYSIS AND THE TRANSLATION PROFILE 57

Types of sameness (equivalences) Types of difference (shifts)

formal syntactic
transposition
unit shift (e.g. clause > phrase)
...
semantic semantic
modulation
generalization
...
stylistic stylistic
variation (e.g. of dialect)
foreignization
...
pragmatic pragmatic
addition
omission
explicitation, implicitation
reduction
...
Table 1. Types of sameness (equivalences) – Types of difference (shifts)

Note the lack of delicacy on the left; we have a rather more detailed conceptual
repertoire available on the right.

4. Non-translated texts
The source text is not the only reference point for a similarity analysis. We also
relate translations to non-translated texts, of the same type and function and
perhaps also subject matter, in the target language. I will call these NTs, for
non-translated texts. These are also known as comparable texts or parallel
texts. Less happily, they are also sometimes referred to as native texts (a
misleading term, for not all such texts are by native speakers or by
monolinguals); or as authentic texts (are translations not also authentic?); or as
original texts (too easily confused with source texts).
So the description of the translation profile should also note the similarities
(samenesses and differences) vis-à-vis relevant NTs. In other words, we are
looking at the naturalness relation: how natural is the translation? To what
degree does the translation manifest “textual fit”? (See Chesterman 2004.)
What evidence is there of “translationese”?
Here too, we tend to find asymmetry. Translations into a given target
language are de facto texts of that language, however natural they may or may
not be. At the very least, they are more similar to other target-language texts
58 ANDREW CHESTERMAN

than to the source-language text, in obvious ways. But what scholars have
tended to focus on is not this similarity, but the ways in which the translations
in question are not natural. Here again, we are more fixated on differences. The
reactions of ordinary readers, of course, are usually different. They read
translations primarily as target-language texts, i.e. assuming this similarity,
even though the texts may sometimes sound a bit strange.
What concepts and terms are available for similarity analysis with respect
to non-translated texts? So far, we have the notions of translationese, of
matches and mismatches, or fit and misfit. How could we relate types of
matches and mismatches to types of equivalence and types of shift? Is a
consistent classification possible, or even necessary? We do not seem to have
an agreed set of terms or concepts here, nor even a general cover-term. Could
we talk about kinds of drifts with respect to non-translated texts in the target
language, in contrast to shifts with respect to the source text? Following our
earlier simple classification, we could extend it like this:

Formal drifts e.g. grammatical errors

Semantic drifts odd meanings, neologisms
Stylistic drifts odd style; strange strings
(i.e. odd collocations; cf. Mauranen 2002)
Pragmatic drifts odd message; code-switching?

Returning to our sample data (above), we can for instance note that in the
German there is an unfortunate misspelling (Beute instead of Beutel): a formal
drift. In the English there is also a misspelling (ware instead of wear), and also
some examples of odd or barely acceptable syntax (so that feet: stylistic drift;
adding value: a dangling participle, i.e. another formal drift). The French has
one unusual choice of lexeme (montage, instead of e.g. fabrication). In this
way we can list the observed drifts in addition to the shifts.

5. Other translations
We can also compare a translation (or set of translations) to other translations
(OT), in the same or different languages. Such comparisons take us beyond the
particular case of an individual translation and its source text. One way of
doing this is via the comparative study of intertextual influences, as reported
e.g. in the Göttinger Beiträge series. A more linguistic approach is that of
corpus research on translation universals, pioneered by Baker (1996), Laviosa
(e.g. 2002) and others. Universals are hypothesized generalizations about
features of translation profiles which seem to be widespread, perhaps in fact
universal; they are at least found in translations of different text types and
 SIMILARITY ANALYSIS AND THE TRANSLATION PROFILE 59

between different languages. These generalizations cover both of the textual

relations mentioned in the previous sections, above. Some concern patterns of
difference or similarity in relation to source texts; I have called these S-
universals (Chesterman 2004). Others concern patterns of difference or
similarity in relation to non-translated, comparable texts (NTs) in the target
languages; these I have called T-universals (Chesterman 2004). Examples of S-
universals are “translations involve shifts” (– a relation of difference), or
“translations are marked by interference” (– a relation of similarity). A well-
known T-universal is Toury’s law of growing standardization (Toury 1995).
Statements about potential universals are thus statements about similarities: i.e.
about similar tendencies which are observed across such a wide range of
translations that they are hypothesized to be universal. Comparisons with the
profiles of other translations are complicated, in that they involve similarities at
two levels, as the other profiles themselves are already the sum of those
translations’ similarities with yet other reference texts.
A recent proposal for a new T-universal (or “law”) is put forward by
Kanter et al. (2006). They show that the proportion of word-types that are
unique to a given translation corpus (and do not appear in a matched NT
corpus) seems to remain surprisingly constant across different language pairs,
and is independent of source language. We could call this the translation-
specificity hypothesis. Does this suggest something about the inevitability of
what I have called translation drifts?
This extension of the kinds of reference texts we are interested in, to
include other translations, is also carrying us away from pure description
towards explanation. We can argue as follows. Here, in this translation profile,
I find feature F. Other scholars have also found feature F in different
translation profiles. So perhaps the occurrence of feature F is a general
tendency. Therefore, in some sense, I am not surprised when I find feature F
here: its generality somehow explains its presence in this translation, I
understand why it is there, to some extent at least. This is perhaps a rather
weak sense of “explanation” – but I will return to this point later.
Looking at our sample data again, we can observe for instance that no
translation is quite natural; some contain signs of carelessness; they are perhaps
the work of non-native speakers, or non-professional translators (but evidently
not of machine translation programs). We can also note that all the translators
have taken a good deal of freedom to add, alter or omit details of information,
that the basic layout has been preserved, and that all have preserved the key
term sacchetto, albeit in different ways. These shared features are examples of
similarities across the different translations. They might suggest
generalizations that could be made about all translations of these kinds of texts,
which could then be tested on other texts of the same kind. They may also
60 ANDREW CHESTERMAN

carry implications about the kinds of instructions given to the translator by the
client, and about the kinds of translators selected for the job: here too, there are
hypotheses to be tested.

6. Similarity is a multi-place predicate

But we have been oversimplifying. Similarities depend not just on what
something is being compared to, but also on circumstances, on the purpose of
the similarity judgement, on point of view. Perhaps anything can be judged to
be similar to anything from some point of view, to some extent. So similarity is
always a contingent property. Technically speaking, similarity is a multi-place
predicate. “A is similar to B” means “A is similar to B in respects C according
to comparison process D, relative to some standard E mapped onto judgments
by some function F for some purpose G” (Medin and Goldstone 1995: 106). So
we cannot conceptualize or analyse similarity without the additional notions of
context, relevance and purpose. (See also Goodman 1972.) In TS, this means
that we cannot analyse a translation profile without also referring to the
sociocultural context, the skopoi of the texts concerned, and so on.
So we can ask: is this translation similar (enough) to the ST from the
client’s point of view? The customer’s point of view? (We can also wonder
whether the client will actually read the translation at all.) These points of view
are not the same as that of the translation scholar, of course.
Change the point of comparison, and the similarity judgements change.
Tymoczko (2004: 34) points to the risk of an ethnocentric bias in making
similarity judgements of similarites between translations and their source texts.
This reservation extends to the whole notion of translation itself, which
Tymoczko argues, is a cluster concept based on several family resemblances,
several kinds of similarities. Her conclusion is that “there will be no single set
of criteria that also defines similarity in translation” (2004: 35).

7. The link with quality assessment

Different points of view on similarity also offer a key to understanding
translation quality assessment. Quality has to do with the relation between
similarity expectations and similarity judgements. That is: we have
expectations about the similarity relation a given translation should have with
its source text (is it accurate enough?) and with NTs (is it fluent enough?). In
assessing the quality of a translation, we compare what we observe with what
we expect. “Translation errors” are simply instances where the gap between
expectations and judgements becomes unacceptably wide – and “unacceptably”
 SIMILARITY ANALYSIS AND THE TRANSLATION PROFILE 61

depends of course on the circumstances. In this sense, errors are also multi-
place predicates, and we could set up typologies of errors with this in mind.
The similarity relation with other translations is interesting in this respect.
If our observed profile feature F is one that goes against our similarity
expectations, but also occurs with a high frequency in other translations, is it
somehow “less serious”?
Consider our examples. Does it matter that the translations are not
perfectly natural? Who will worry about spelling mistakes? Who will read the
texts anyway? Perhaps all that is required is “existential equivalence” only –
i.e. that the translations should simply be seen to exist (Koskinen 2000: 83).
What are our expectations of such texts?

8. Divergent and convergent similarity

The notion of similarity is further complicated by the realization that there are
two distinct types of similarity (Sovran 1992, Chesterman 1996, 1998), each
based on a different viewpoint of the origin of the similarity in question. We
are now approaching the explanatory question: why is this similar to that?
Divergent similarity is a one-to-more-than-one relation: you start with one
thing, and produce another that is similar to it in some relevant way. Like a
copying machine; or a composer inventing variations on a theme; or a
translator producing a translation. Like this:

A > A, A’ (, A’’...)

In other words, from an initial situation in which we have A, we move to a

situation in which there is not only A but also A’ etc., which has been created
similar to A.
Convergent similarity, on the other hand, starts with a situation in which
two (or more) entities exist, and the perception (by someone, from a given
point of view) of a similarity between them. This is the kind of similarity you
find in riddles: why is a raven like a writing-desk? Or: this church in Ghent
reminds me of one I saw in Strasbourg recently. This is the similarity used in
contrastive linguistics. I symbolize it like this:

A < ––– > B

Divergent similarity is similarity by design, therefore – a causal explanation –,

whereas convergent similarity is similarity by apparent chance. (True,
underlying this apparent chance there may in fact be some cause.) Divergent
similarity is created; convergent similarity is perceived.
62 ANDREW CHESTERMAN

A translator creates divergent similarity, but a critic or a scholar looks for,

or perceives, convergent similarity. From the scholar’s perspective, we may
conclude that the similarity is inadequate, that the translation is “not natural
enough” or “not close enough to the original”; but the translator may have been
deliberately aiming at a foreignizing translation, or a free one: a different
similarity. Moreover, people looking for convergent similarity may disagree:
you may see a similarity that I do not see at all. I might not see the similarity
that the translator has chosen to create; and she might not create the one I am
expecting.

9. Explanation via generalization: learners’ texts

Consider now the following argument by William Croft about the way in
which generalizations gradually increase their explanatory power. It brings us
back to the point I introduced earlier, about descriptive generalizations
implying a weak explanation. Croft writes:
Instead of using the dichotomy of ‘description’ vs. ‘explanation’, one can
describe grammatical analysis (or any other sort of scientific analysis, for that
matter) with a scalar concept of degrees of generalization. The basic concept is
that a more general linguistic statement can be said to explain a more specific
one, though it may itself be explained by a yet more general statement. Thus,
any given statement is an explanation for a lower-level generalization, but a
description in comparison to a higher-level generalization. (Croft 1990: 246,
cited in Halverson 2003: 230)
Croft suggests various levels of generalization, which we can apply to
translation similarity analysis as follows.

(i) Observation: description of basic facts

(> there are some shifts or drifts here...)
(ii) Internal generalization about these facts
(> similar kinds of shifts / drifts occur elsewhere, under given conditions)
(iii) Higher internal generalization
(> these shifts / drifts may manifest a general / universal tendency)
(iv) External generalization
(> constraints concerning human psychology / biology / sociology...)

So far we have remained with translations and source texts, and with non-
translated target-language texts. All together, these would correspond to
Croft’s levels i-iii. What about external generalizations, level iv? We could add
to our reference texts corpora representing target-language production in a
 SIMILARITY ANALYSIS AND THE TRANSLATION PROFILE 63

different set of special conditions: texts produced by non-native writers. I will

call these LT, for learners’ texts. After all, both translations and learners’ texts
are produced under particular constraints, and it may be that these constraints
have similar effects.
Suppose we now look at the relation between translations and these
learners’ texts, and suppose that we find quite a few interesting similarities. In
one of the first proposals about possible translation universals (explicitation),
Blum-Kulka (1986) drew attention to the importance of making such a
comparison, but as far as I know it has not been taken up much by translation
scholars. There is certainly lots of work on learners’ interlanguage itself,
particularly on simplification in interlanguage, from Selinker (1972) on.
Research in this direction might then lead us to propose generalizations (or
explanations) in terms of general language processing capacity under special
constraints, which might in turn suggest hypotheses about human cognitive
processing more generally. If we ask why feature F occurs in our translation
profile (a basic observation), we could pursue answers back up the scale of
increasing generalization:
• because many translators tend to produce F (under given conditions) –
internal generalization
• because there is a general tendency towards F under extra processing
constraints – higher internal generalization
• because human cognition typically seeks e.g. the most energy-saving
solutions, of which F is an example (see e.g. Halverson 2003) – external
generalization.

Note that we are now beyond explanations couched in terms of the translator’s
intention. A translator may aim at a certain kind of equivalence (divergent
similarity), or a certain kind of target-language usage, but the kinds of relations
we are discussing when we look at similarities with other translations or with
learners’ texts are of a different nature: they are convergent, and therefore
require a different kind of explanation.

10. Conclusion: relating similarities

We could also try to climb up Croft’s scale by looking at similarities between
the very generalizations we are investigating. For instance, many proposals
about universals can be represented in terms of graphs. In some cases, when we
are looking at the relation with the target language (similarities with NT),
translations seem to make more use of the central range of certain items which
have a bell-curve distribution in TL. Examples are levelling, convention-
alization, standardization. What these share is a general feature I have called
64 ANDREW CHESTERMAN

curve-hitching (Chesterman 1997: 72): the bell curve is hitched upwards, so

that it has a higher centre.
Other general tendencies can be observed at each end of the frequency
distribution. Translations seem to have more high-frequency items (Laviosa
2002), and they also have more low-frequency items (Mauranen 2002),
including target-language errors (formal drifts) and other evidence of
translationese. They may also have less frequent “unique items” (Tirkkonen-
Condit 2002). So we could ask: under what circumstances do we find these
general kinds of curve changes? And why under these circumstances rather
than some others?
Doing a full similarity analysis of a translation profile therefore means
looking at both samenesses and differences, and comparing the translation (or
set of translations) at least to the following reference sets: source texts, non-
translated texts (including perhaps non-translated texts by the translators
themselves), other translations, and learners’ texts. But a full analysis may of
course not be the aim of a given project. As we have seen, it is the purpose of a
given similarity analysis that always plays a decisive role. For instance, the
study of pseudo-translations could not use STs at all, but would find similarity
with other translations very relevant.
With respect to non-translated texts, it may also be the case that texts
produced by monolingual native speakers are different from texts produced by
bilinguals (see e.g. Cook 2003). Perhaps we will eventually need even more
categories of reference texts for a still more comprehensive similarity analysis.

References
Baker, M. 1996. “Corpus-based Translation Studies: the challenges that lie ahead”. In
H. Somers (ed.), Terminology, LSP and Translation: Studies in Language
Engineering in Honour of Juan C. Sager. Amsterdam: Benjamins, 175-186.
Blum-Kulka, S. 1986. “Shifts of cohesion and coherence in translation”. In J. House
and S. Blum-Kulka (eds), Interlingual and Intercultural Communication:
Discourse and Cognition in Translation and Second Language Acquisition
Studies. Tübingen: Narr, 17-35.
Catford, J.C. 1965. A Linguistic Theory of Translation. London: Oxford University
Press.
Chesterman, A. 1996. “On similarity”. Target 8(1): 159-164.
Chesterman, A. 1997. Memes of Translation. Amsterdam: Benjamins.
Chesterman, A. 1998. Contrastive Functional Analysis. Amsterdam: Benjamins.
Chesterman, A. 2004. “Hypotheses about translation universals”. In G. Hansen, K.
Malmkjær and D. Gile (eds), Claims, Changes and Challenges in Translation
Studies. Amsterdam: Benjamins, 1-13.
 SIMILARITY ANALYSIS AND THE TRANSLATION PROFILE 65

Chesterman, A. 2005. “Problems with strategies”. In K. Károly and Á. Fóris (eds),

New Trends in Translation Studies. In honour of Kinga Klaudy. Budapest:
Akadémiai Kiadó, 17-28.
Cook, V.J. (ed.). 2003. Effects of the Second Language on the First. Clevedon:
Multilingual Matters.
Croft, W. 1990. Typology and Universals. Cambridge: CUP.
Goodman, N. 1972. “Seven strictures on similarity”. In N. Goodman, Problems and
Projects. Indianapolis: Bobbs-Merrill, 537-447.
Halverson, S. 2003. “The cognitive basis of translation universals”. Target 15(2): 197-
241.
Hermans, T. (ed.). 1985. The Manipulation of Literature. Studies in Literary
Translation. London: Croom Helm.
Kanter, I., Kfir, H., Malkiel, B., and Shlesinger, M. 2006. “Identifying universal laws
of text translation”. Journal of Quantitative Linguistics 13(1): 35-43.
Koller, W. 1979. Einführung in die Übersetzungswissenschaft. Heidelberg: Quelle und
Meyer.
Koskinen, K. 2000. Beyond Ambivalence. Postmodernity and the ethics of translation.
Tampere: Tampere University Press.
Laviosa, S. 2002. Corpus-based Translation Studies: Theory, Findings, Applications.
Amsterdam: Rodopi.
Lecercle, J.-J. 1990. The Violence of Language. London and New York: Routledge.
Mauranen, A. 2002. “Strange strings in translated language A study on corpora”. In
M. Olohan (ed.), Intercultural Faultlines. Research Models in Translation
Studies 1: Textual and Cognitive Aspects. Manchester: St. Jerome Publishing,
119-141.
Medin, D.L. and Goldstone, R.L. 1995. “The predicates of similarity”. In C. Cacciari
(ed.), Similarity in Language, Thought and Perception. Turnhout: Brepols, 83-
110.
Molina, L. and Hurtado Albir, A. 2002. “Translation techniques revisited: a dynamic
and functional approach”. Meta 47(4): 498-512.
Nida, E.A. 1964. Toward a Science of Translating. Leiden: Brill.
Pym, A. 1995. “European Translation Studies, une science qui dérange, and why
equivalence needn’t be a dirty word”. TTR 8(1): 153-176.
Selinker, L. 1972. “Interlanguage”. IRAL 10: 209-231 (Reprinted in J.C. Richards
(ed.) 1974, Error Analysis: Perspectives on Second Language Acquisition.
London: Longman, 31-54.
Sovran, T. 1992. “Between similarity and sameness”. Journal of Pragmatics 18: 329-
344.
Tirkkonen-Condit, S. 2002. “Translationese – a myth or an empirical fact? A study
into the lingustic identifiability of translated language”. Target 14(2): 207-
220.
Toury, G. 1995. Descriptive Translation Studies and Beyond. Amsterdam: Benjamins.
Tymoczko, M. 2004. “Difference in similarity”. In S. Arduini and R. Hodgson (eds),
Similarity and Difference in Translation. Rimini: Guaraldi, 27-43.
van Leuven-Zwart, K. 1989. “Translation and original. Similarities and dissimilarities,
I”. Target 1(2): 151-181
van Leuven-Zwart, K. 1990. “Translation and original. Similarities and dissimilarities,
II”. Target 2(1): 69-95.
66 ANDREW CHESTERMAN

Venuti, L. 2002. “The difference that translation makes: the translator’s unconscious”.
In A. Riccardi (ed.), Translation Studies. Cambridge: CUP, 214-241.
Vermeer, H.J. 1996. A Skopos Theory of Translation. Heidelberg: TEXTconTEXT.
Wendland, E. 2004. “What’s the difference? Similarity (and dissimilarity) from a
cross-cultural perspective: some reflections upon the notion of ‘Acceptability’
in Bible translation”. In S. Arduini and R. Hodgson (eds), Similarity and
Difference in Translation. Rimini: Guaraldi, 329-358.