Translation norms and conventions

What it is

A convention is a regularity a community has settled into: one solution among several that would have served, kept up because each party expects the others to keep it up. A norm is what a convention becomes when that expectation stops being a preference and acquires a binding character. The difference is what a departure costs, and the rest of this section is built on it.

Neither is produced by the language system. The system leaves a space open, several forms fit, and the community narrows the space without writing anything down. Where the narrowing has hardened far enough, an authority may write it down afterwards, and the codification axis measures that writing rather than the narrowing. A convention can therefore exist with no text behind it, and most of the ones a translator meets do.

This record carries the axis and not a compliance. Its own Expectancy value is a statement about its subject matter: for the concept of a norm to change, the field's account of it would have to change, which is the Expectancy test at one remove. A reader who takes the value as a claim about a stretch of text will misread every other record in the section.

Where the expectation comes from

The vocabulary is the field's own, and the three sources behind it were reached very unevenly, which an Expectancy record has to say where it says the rest.

• Hermans (1996) is the only one opened in full, as the author's deposited copy in a university repository. He defines conventions as regularities in behaviour which have emerged as arbitrary but effective solutions to recurrent problems of interpersonal coordination, growing out of precedent and social habit. They imply, in his formulation, a set of mutual expectations: that others expect me to adopt a certain course of action in a given situation, and that I expect them to expect it. A convention turns into a norm when it has solved its coordination problem well enough for long enough that the expectation on all sides grows beyond a mere preference and acquires a binding character.
• Chesterman (1993) supplies the tier's name. His abstract, which is the whole of what this section has read of that article, separates professional norms governing the process, of accountability, communication and target-source relation, from expectancy norms "concerning the form of the translation product, based on the expectations of the prospective readership".
• Nord (1991), also read at abstract level only, supplies the duty that follows: since conventions determine what readers expect of a translation, a translator is responsible for not deceiving users by departing from them without saying so and why. She calls that loyalty.

None of this establishes that any community holds any particular expectation. Two of the three sit behind paid access and were read as their publisher's abstracts; the third is a framework and measures nothing. They give the vocabulary and the test of what would have to change, and they are not evidence for any convention described elsewhere here.

Consequences for translation

The practical value of the distinction is that it tells a reviser what kind of objection is available.

• A departure from a norm is an error and a departure from a convention is an impression. Both are reportable; only the first can be put as a defect, and a report that does not say which has not told the client what to do.
• The two can sit inside one word. Which German address pronoun a text owes is conventional and unwritten; how that pronoun is spelled once chosen is ruled by a code, at paragraphs 65 and 66 of the Amtliches Regelwerk.
• Loyalty is discharged by disclosure, not by compliance. Departing from a target convention is legitimate where the client is told; departing silently is what Nord's principle forbids.
• Neither link field is claimed here, because the record describes what an expectation is rather than a compliance owed by running text.

Examples

EN>DE

• Source: Would you help me?
• Target: Würden Sie mir helfen?
• Comment: The Amtliches Regelwerk's own illustration at paragraph 65, where the rule is that Sie and Ihr take a capital. Nothing written says a stranger is owed Sie: that is the convention, held by the readership. A translator who read the code as ruling the choice would take a spelling rule for a social one.

DE>EN

• Source: Würdest du mir helfen?
• Target: Would you help me?
• Comment: The Regelwerk's illustration at paragraph 66, where du and ihr are written small, with a note allowing the capital in letters. English has one second person form, so what the source encoded conventionally the target cannot encode at all. A reviser who flags the loss is describing the system, not a fault.

References

• Hermans, Theo. "Norms and the Determination of Translation. A Theoretical Framework", in Translation, Power, Subversion, edited by Román Álvarez and M. Carmen-África Vidal, Topics in Translation, Multilingual Matters, Clevedon, 1996, pages 25 to 51. Read as the author's green open access copy deposited in the UCL Discovery repository; no page number inside the chapter survived the conversion, so the chapter is cited by its range. The phrase "convergent mutual expectations" used at that point in the chapter is Ullmann-Margalit's and is not taken up here; David Lewis is discussed there for a technical definition of convention, and Convention: A Philosophical Study was not opened.
• Chesterman, Andrew. "From 'Is' to 'Ought': Laws, Norms and Strategies in Translation Studies", Target 5(1), 1993, pages 1 to 20, DOI 10.1075/target.5.1.02che. Abstract only; the article sits behind paid access and was not read.
• Nord, Christiane. "Scopos, Loyalty, and Translational Conventions", Target 3(1), 1991, pages 91 to 109, DOI 10.1075/target.3.1.06nor. Abstract only, on the same terms. Neither "constitutive" nor "regulative" appears in it.
• Amtliches Regelwerk der deutschen Rechtschreibung, Regeln und Wörterverzeichnis, 2024, binding from 1 July 2024. Paragraphs 65 and 66 (Anredepronomen und Anreden), read in the edition published by the Council's secretariat at the Leibniz-Institut für Deutsche Sprache; the edition and binding statement read at the Rat für deutsche Rechtschreibung.