werner [at]


Translationese describes the conscious or unconscious change of the original sense by introducing ideas that are not found in the original. This ranges from changes in content to formal changes.


The development of the term “translationese,” or “translatorese,” parallels the progression of translation studies as observed by Helge Nika (1999) and Martin Gellerstam (1986; 1996). However, the phenomenon of an “odd, unnatural language that only appears in translations” (Leman 2009)—or, “the fingerprints that one language leaves on another when a text is translated”—is much older, as Silvia Bernardini and Marco Baroni have suggested (2005, 2). The key to its occurrence is the question of translatability, as Anthony Pym and Horst Turk have argued. Thus, one would not be surprised should an awkward translation be produced, if translation adapts to the “principle of necessary sacrifice” (Pym and Turk 2001, 274). Alluding to the intuitive character of translation, this principle implies that not everything in texts and speech is translatable or interpretable. The impreciseness of language as a central constituent in translation, together with the limited human ability to communicate cross-culturally, bring along the possibility of producing translationese as well as the difficulty of bypassing it.

The concept of translationese bears some of the negative connotations that go hand-in-hand with language styles like “journalese,” “officialese,” and “legalese,” but it differs in significant ways from maltranslation and pseudotranslation according to Henry Fowler (1965) and Andrea Rizzi (2008, 153,155). Its occurrence features universals of translation, such as normalization, simplification, greater explicitness, and an overall bias toward conservation. It may be characterized by a smaller inventory of unique words than the text being translated (the “unique items hypothesis” mentioned by Sonja Tirkkonen-Condit 2002, 208, and Sara Laviosa-Braitwaith 1998, 288–91). With the improvement of translation quality assessment introduced by Katharina Reiss (1971), the language of translation became more obvious, and the subject of translationese appeared soon afterward as a separate “dialect” or the “third code” within a language, as William Frawley observed (1984).

Recently, linguistic scholars have begun questioning the assumption that translated texts can be identified as translations when compared with source texts. Empirical studies using questionnaires and computational statistics and support vector machines seem to reveal that a text can seldom be identified as a translation on the basis of textual evidence (Bernardini and Baroni 2005; 2006). On the other hand, Tiina Puurtinen (2003) and Tirkkonen-Condit (2002, 208) noted that research supports the notion of translationese by identifying fixed sets of lexical, syntactical, and textual markers that only occur in translated texts.

In Bible translation, translationese is characterized by strange, literally translated phrases that arise (a) through borrowing phrases from the biblical languages, Hebrew, Aramaic, and koine Greek, and (b) from semantic oddities caused by a literal concordant approach. Examples from the KJV that likely have equivalent renderings in many languages would be the following: the idiom, “Gird up thy (your) loins” (2 Kgs 4:29); the affirmation, “Verily, verily, I say unto you” (John 1:51); and the introduction to a direct quotation, “John answered, saying…” (Luke 3:16).  When a language is reduced to written form for the first time and becomes a vehicle for translated works, a written style is developed. This form may become the norm for literary works, even though it has an awkward or stilted quality. In the case of first Scripture translations, the language of translation may take on an aura of sanctity that will be a constraint for future translators and writers (Nida and Taber 1969, 100, 124–25; Leman 2009). See also Effability of language.