Why not just use Google Translate?

A few days ago, I was discussing my work as a translator with a friend of mine. After going through the general details of what I do, he asked: “So, why don’t your clients just use Google Translate?”

It’s a question that I’m sure many translators have faced before, and I doubt it will be the last time I hear it. It wasn’t meant to be an insult to our profession (nor was it taken as one), but rather seems to reflect a common general perception of translation. As it was based on a genuine interest in what I do and highlighted an important issue, I wanted to give a clear, convincing answer.

Naturally, I started out with the typical translator’s response: “It just doesn’t work in many cases, it trips up on verbs, word order, everything really, and it doesn’t take account of the context.” Yet I could see that he wasn’t satisfied, he wanted concrete info.

I continued. “Think of an English word like ‘set’. It can mean so many different things. We can play a set of tennis, we can set up a business, set ourselves up for a long night of translating… Obviously, other languages don’t have the same word for all those meanings and Google Translate can’t always catch the right one.”

Still slightly unconvinced, at this point the conversation drifted onto something else and I went away feeling that I didn’t make my point quite as clearly as I could’ve done. As such, I decided to see if I could come up with a few examples here to demonstrate a few of the limitations of Google Translate.

I’m not going to go to the extremes of using examples that have gone catastrophically wrong, there are plenty of posts that do that. Instead, I want to show that, despite its many strengths and its rapid improvement in recent years, Google Translate won’t be replacing professional translators any time soon.

First up, how about a nice French proverb like “le monde appartient à ceux qui se lèvent tôt” [literally: the world belongs to those who get up early]?

Google gets this literal rendition spot on. But what does that actually mean? Though French speakers will be familiar with the world belonging to them when they get up nice and early, it won’t strike a chord with an English ear. However, if you were to say “the early bird catches the worm”, your Anglophone audience would certainly catch your drift.

Or what about the lovely (if slightly obscure) French phrase “c’est fromage et dessert” [literally: it’s cheese and dessert] that I recently had to tackle in translation?

It seems simple at first glance, it’s just a pair of common nouns. But this wasn’t a translation of a menu or something similar as you might expect. Instead, it was an article on university education in France. In this context, the phrase was used as a play on the old menu choice of “fromage ou dessert” [cheese or dessert] that invites diners to choose one or the other at the end of a meal.

In using “et” [and] instead of “ou” [or], it highlights that, whereas normally you have to make a choice between the two, in this situation you can enjoy each of two normally opposed options.

Putting the phrase through Google Translate, we’re left clueless as to its meaning in this context with a literal rendering of “it is cheese and dessert”. 

When translating the phrase myself I opted for an English version reading “it’s a case of having your cake and eating it”, recreating both this idea of combining two seemingly opposing choices as well as the culinary allusions contained in the French.

Both of these examples highlight a huge problem with Google Translate, it simply can’t handle idiomatic language. The same can be said of rhetoric, style, humour, and many other important facets of language that are tough even for a professional translator.

Finally, here’s an example from another text I worked on a while back. This time I thought I’d go all out and give an excerpt from the French text, a Google-translated version and the published English version to show the discrepancies.

The article in question offered a round-up of the sprinting events at the Rio 2016 Olympic Games and the excerpt used is taken from the opening paragraph, which summarises a few of the highlights from the competition.

I’ve deliberately chosen an example where it was possible for the published version to remain very close to the French to show that even when a fairly literal rendering would suffice in English, Google Translate still isn’t up to the task (despite faring pretty well).

The problems are fairly clear to see, but I’ve highlighted a few personal favourites.

French text

Avec un nouveau triplé 100 m-200 m-4 x 100 m, Usain Bolt a définitivement cimenté à Rio sa légende de plus grand sprinter de tous les temps. Le Sud-Africain Wayde van Niekerk a fait sensation en battant un record du monde vieux de 17 ans sur 400 m, Elaine Thompson, la compatriote d’Usain Bolt, a réalisé le doublé 100 m-200 m, tandis que l’Américaine Allyson Felix est devenue la femme la plus titrée de l’histoire en athlétisme.

Google’s version

With a new triple 100 m-200 m-4 x 100 m, Usain Bolt has definitely cemented in Rio its legend of the biggest sprinter of all time. The South African Wayde van Niekerk made a sensation by beating a 17-year-old world record over 400m, Usain Bolt’s compatriot Elaine Thompson scored 100m-200m, while American Allyson Felix became the most titled woman in history in athletics.

Published translation

In the men’s sprinting events in Rio, Usain Bolt claimed three more golds to cement his status as the greatest sprinter of all time and South Africa’s Wayde van Niekerk broke a 17-year-old world record in the 400m. In the women’s competitions, Bolt’s compatriot Elaine Thompson took a sprint double in the 100m and 200m while Allyson Felix of the United States became the most-decorated female athlete of all time.

Ultimately, Google Translate is a great resource for certain purposes. It works pretty well in many cases and you can often get the gist in the examples above.

If all you need is that gist, then that’s fine. However, when you’re looking for a flowing, polished translation that always makes sense, don’t just assume that Google will do the job, get yourself a pro!

All that remains now is for me to print off this post and make it into a neat little handout to give to the next person who asks “Why not just use Google Translate?”



Getting to the Heart of Medical Texts

After all the fun of Fawlty Towers last time out, I thought I would tackle a topic that is more concretely tied to the profession this time around and give an insight into some of the issues I’ve found in my work as a medical translator.

There are many key characteristics of medical texts that make them challenging and rewarding to work with and I want to highlight some of the most commonly encountered issues before suggest potential methods of addressing them.

Where to start?

Medical Translation by Montalt & Gonzalez Davis sets out the basic requirements of a medical translation as follows: coherent with source text, coherent internally, truthful or accurate, readable, clear, grammatically and syntactically correct and adequate – stylistically and rhetorically in keeping with the communicative situation and context.

While this only serves to further complicate an already daunting task as many of the points listed apply to translation in general rather than medical translation specifically, the points can be easily broken down so that areas of focus can be prioritised. In original articles, for example, the decisive point to address is ‘accuracy[,] so that the experiments can be repeated and that the argumentation can be followed in detail’.

In aid of this maxim of accuracy, editing the source text in both terms of content and style is regarded as standard: enhancing the author’s ideas and predicting his intended meaning if it is unclear are both considered common practice while the prevailing attitude among medical translation specialists is that ‘ideally a translation should read like an original text, that is, the readers should not be aware that they are reading a translation’ (Montalt & Gonzalez Davis again) and, although this may go against the thought’s of many scholars in translation studies, it is an inescapable requirement of the field and one which entails a thorough understanding of how medical texts function and what to look for when translating.

There is some help at hand

Fortunately, there do exist certain features of medical texts which require little time to become familiar with; among the most notable of these for example is the highly standardised structure of the texts. As Henry Fischbach describes: ‘[m]edical translation is the most universal and oldest field of scientific translation because of the homogeneous ubiquity of the human body and the venerable history of medicine’ and this summary alludes to several key features of medical writing as well as the major advantage of the translator in their work with medical texts: that is, the universal nature of the material which suggests that ready equivalents will be available in other languages. (While this is definitely true of physical attributes and many procedures, treatments etc. you will often find that many new terms, techniques or other discoveries will NOT have an equivalent in all languages – but it’s a start!)

Furthermore, given that ‘the value of the text often lies in its factual content’ (Montalt & Gonzalez Davis 2007:155), elements of meaning are explicit, and rhetorical devices – allusions or other persuasive techniques, for example, which are frequently found in texts of other genres – are rarely used.


Yet still standing in the way of us achieving the desired end product is the not-insignificant issue of terminology – with terminological issues among the most commonly found and most pressing issues in medical translation .

Looking a bit more closely, some clear features can be distinguished. As Fischbach suggests in the citation above, medical terminology is steeped in a long, historical tradition, with a mostly Greco-Latin parentage that is common between languages such as English and French. This terminology also hints at the high register that is maintained throughout most medical writing (although products such as leaflets and information sheets are specifically designed to have a low register and be accessible to a wider audience).

Meanwhile, beyond archaic-rooted or arcane terminology which can be challenging enough to hunt down, the presence of terms which adopt a specific meaning unique to the context of medical writing present an additional challenge for the translator. Further still, the presence of acronyms is extremely regular (such as the fairly common LCR, TDM and ETF in French) and this requires the translator to enquire into the existence of recognised equivalents in the target language.

One example demonstrating the difficulty of accurately addressing terminological issues is the translation of the French term ‘fenêtre thérapeutique’. Upon initial enquiry the term ‘therapeutic window’ seems an obvious, literal equivalent but is actually incorrect (an example of a false friend), referring to a wholly different phenomenon in English medical terminology. Meanwhile the Grand Dictionnaire Terminologique’s suggested translation of ‘drug holiday’ (itself a seemingly curious translation) is actually a correct rendering of the term, albeit one that consultation of parallel texts shows is generally indicative of a lower register in English, more akin to medical fact sheets than journal articles. Fortunately, there exist several accepted synonyms for the phrase in English and ‘structured treatment interruption’ would be chosen as the most fitting in the translation of a medical journal article.

Finally regarding terminology – and with thanks to this excellent termcoord article for the example – eponyms can present a major problem. For example, according to Dermatology Therapy: A-Z Essentials, “Infantile Scurvy” has the following synonyms: “Barlow’s disease”; “Möller-Barlow disease”; “Barlow’s syndrome”; “Cheadle-Möller-Barlow syndrome”; “Moeller’s disease”; “vitamin C deficiency syndrome” and ‘choosing between an eponym and another term would depend on which is more common in the target culture.’

In aid of this difficult task, I would recommend the use of ongoing online searching throughout the translation process to quickly and efficiently isolate terms and phrases in both the source and target languages – have a read of my post on the best places to hunt down terms online if you’re stuck!

Poorly written texts

However, even when these considerable problems are all dealt with, one undesirable feature that is common to many medical texts (and indeed technical texts in general) is the prevalence of many typographical, stylistic, grammatical and formatting errors which serve to extend the translator’s role to that of proof-reader and editor. These errors are indicative of the lack of linguistic competence of many medical writers, for whom it is not at all a prerequisite as technical authors in any language are ‘often chosen for what they know, not how well they write, and many write very badly’ (Wright 1993).

Examples of formatting and stylistic errors can include the incorrectly placed spaces, the unnecessary usage of capitals as well as the inconsistent or unnecessary usage of full stops and commas and – last but not least – between using numerals or their full, written versions.

In addition to these examples the numerous instances of typographical and grammatical errors even include simple misspellings and, while a few such mistakes could perhaps be forgiven, their frequency and severity demonstrate that errors go beyond a few careless slips on the part of the author and all of this ultimately adds weight to the importance of the translator’s extended role as editor and proof-reader in their work, highlighting the tremendous importance of a comprehensive understanding of technical writing norms in both the source and target cultures.

Overall, beyond all linguistic issues, one key to producing a high quality translation lies in accurately and thoroughly identifying the requirements of the target audience in terms of stylistic conventions, register and terminological concerns: a clear idea of the expected end product will in turn aid the translation process, guiding certain decisions while allowing the translator to reflect upon the required approach with greater perspective. Hope that helps!

Talking Translation – Reflecting on Reflective Practice

I was going to say that this post will be a little bit different but, as I seem to say that at the start of most of my posts, I guess that this could be entirely expected.

This post is based around the link below (click the Soundcloud image) which heads over to a podcast from ‘Transcast’, a collection of podcasts and recordings relating to different aspects of the field of translation. This particular podcast involves a discussion between Dr. Sarah Maitland, Dr. Fruela Fernandez – Translation Studies professors at the University of Hull – and myself, centring on the importance of reflective practice in translation training and professional work, and I just wanted to quickly run through a few key ideas here before leaving you to listen to the podcast.

Reflective practice, for me, represents one of the best working methods available for the practicing translator as well as providing a rare opportunity to integrate theory and practice, which remains a key issue in the discipline. Reflecting on your own work as you translate forces you to question and to justify your translatorial decisions in relation to an end goal (a translation brief in this case) and this in turn allows a deeper understanding of your own working process and a better grasp of your areas of strength and weakness.

But beyond this added certainty in your choices, the act of questioning and justification also leads to a greater level of replicability in your work: I may be able to unwittingly produce the best translation in the world, but without an understanding of the processes leading to it, chances are that I won’t be able to reach this same high standard on a consistent basis, something of key importance to the practicing translator.

On a different level it also stands as an extremely useful pedagogical tool in the way that it allows assessors to gain a valuable insight into the creative process that a student follows in their work as well as their overall understanding of the discipline.

The reflection does, however, need to be based upon a solid knowledge of the underlying principles involved in translation in order to soundly justify decisions. This knowledge can be drawn from professional experience, translation theory, hermeneutics or many other fields.

Ultimately, if you can amply justify a decision to yourself, then you can be much more certain of the quality of your work and can assuredly justify that same decision to a client. Anyway, listen to the podcast and reflect on considering reflecting in your own work. Ciao for now!