12.12.12.: Numbers need translating 2

A little more than two years ago, on 10 October 2010, I posted a piece about numbers and the problems they can cause for translators. It seemed to me that the difficulty of numbers, and their importance for the translator, was underestimated. See, for instance, what Henri Louis van Kooten has to say on the subject of numbers and localisation here (scroll down to section 12) and what Steve Vitek has to say about calendars and counting systems here.

Since it's 12 December 2012 today, it seems a good moment to post some further thoughts on the topic.

Question: When is it OK to translate...

...118 as 999?

Answer: when it's an emergency number.

...41 as 8? 

Answer: when it's a shoe size. In this case, Jean Gabin's shoe size in a subtitled version of Quai des brumes. (The expression 'act your age, not your shoe size' might also be a casualty of translation...)

...1776 as 1789?  

Answer: when it's a year of political turmoil. This particular example is reportedly from a French subtitled version of The Band Wagon. The Declaration of Independence is transposed into the French Revolution.

...12.20 as 5.15?

Answer: when it's a time which is culturally determined as 'late'. In Georges Perec's La disparition, Anton Voyl is suffering from insomnia and is still awake at twenty past midnight. But is 'lateness' in across cultures always equivalent? And does the fact that the source text is a lipogram - a text with no 'e's in it - complicate matters?  

Here are three translations of the French: 

Son Jaz marquait minuit vingt. (p.17) [His Jaz [watch] said twenty past midnight]

English: 

According to his watch it's only 12.20. 

(A Void, trans. Adair, p.3)

Italian: 

Il suo Jaz indicava quasi l'una. [His Jaz said nearly one].

(La scomparsa, trans. Falchetta, p.13)

Spanish: 

Miró el reloj: cinco y quince.      [He looked at his watch: a quarter past five] 

(El secuestro, trans.

Arbués, Burrel, Parayre, Salceda & Vega, p.25) The Spanish text could have used numbers closer to 12.20, like the Italian, but chose not to; the choice of 'cinco y quince' seems at least partly related to differing cultural perceptions of whether 12.20 at night constitutes 'late' or not! (More on this translation here, by the way.)

Adair clearly cheats slightly by using numbers when the corresponding words (twelve twenty) would not be allowed because of the 'e's. Adair is taken to task by another translator, Ian Monk, who argues that where Adair translates '10' similarly literally (into '10'), at another point in the novel, the translation makes no sense; but if you translate the 10 into 6 does the passage work, for the reasons given here.

It's not just that numbers need translating, but they need translating by the principle of relevance. Sometimes they need to be quantitatively identical but expressed in a different format (localisation) and sometimes they need recalibrating, as in recipes. 

How would you translate 'a stick of butter'; 'a cup of flour'? B.J. Epstein has reflected on this at length in a lovely post on recipe translation which I hope she will not mind my quoting at some length:

Cups or grams? Tablespoons or ounces? As is well known, there are different measurement systems around the world and it is not enough to, say, go to http://www.onlineconversion.com/, type in the numbers from the source text and write down what the website has offered you. If you did that, 2 cups would be 4.7317 dl, and when have you ever seen a recipe that calls for 4.7317 dl flour? In cases where measurements have to be changed, there are two major possible strategies. The first is that the publisher simply retains the measurements and then offers a conversion table at the back of the book. This can be quite irritating for a reader, however, because then she or he has to keep flipping from the recipe to the table. If the cookbook is more of the coffee table type, however, which is to say one that people read and look at, but don't really plan to cook from, this solution is fine. But for a cookbook that is meant for real use, it is just not practical. In this situation, new measurements based on the target culture's system must be used. This can be done either via complete replacement or replacement and retention. Complete replacement means that either the translator or another expert tests all the recipes and shifts the measurements so that instead of 4.7317 dl flour, the recipe calls for 5 dl flour. The translator must be careful here to ensure that all the new measurements make sense in the context of the recipe and that all have been converted. A recipe may not work if even one measurement is off, especially for baked goods. Replacement and retention is a combination strategy that means both changing the recipe so it reads 5 dl flour and also keeping 2 cups flour in parenthesis. This can, however, confuse readers, so it is a rare book that will use this strategy. [...]

For further interesting examples of how numbers and mathematics can cause problems for translators, see Brian Mossop's recent article on this topic.

Lastly,  as a reminder of how numbers are culturally inflected, I'm going to leave you with a visual example of how culturally charged counting can be:



Happy soundcheck day.

(c) Carol O'Sullivan 2012