Skip to main content

Deepl versus Google Translate



Online translators are the curse of language teachers, but of huge benefit to most of the world. You may not have heard of Deepl yet, but some say it translates better than Google, so I thought I'd give it a try on a piece of advanced French into English, then some advanced English to French. So here is my French passage from France Info today (12th April 2019):

"Il compte parmi les pires zoos au monde. Situé à Rafah dans la bande de Gaza, le zoo est impacté par les conflits réguliers qui ont lieu dans les zones environnantes. Certains animaux ont succombé à des explosions d'obus mais aussi à la malnutrition et aux maladies. En effet, les éleveurs ont des pratiques particulièrement répréhensibles à l'égard des animaux. Ceux-ci vivent en captivité dans de petites cages insalubres et sont victimes de maltraitances. En février 2019, la vidéo d'une jeune lionne se faisant arracher les ongles avec une pince avait été diffusée, alarmant les internautes.

Après des années de combats, c'est une victoire pour l'ONG Four Paws. Les membres de l'organisation ont souvent été offusqués par cette brutalité et ils ont enfin obtenu gain de cause. Les autorités israéliennes et palestiniennes ont donné l'autorisation de libérer les lions, les loups ou encore les autruches de l'établissement. "Dès que la frontière sera ouverte, nous partirons et évacuerons les animaux du zoo de Rafah", explique Amir Khalil qui est en charge de cette opération complexe.
Après avoir chargé les animaux dans un camion, l'ONG a été escortée par l'armée israélienne pour faire cap vers la Jordanie. Les animaux ont ensuite été déposés dans un sanctuaire où ils devraient passer le reste de leurs jours. Au total, 47 animaux ont été secourus : cette opération est à ce jour la plus grosse jamais menée par Four Paws."

Here is Google's effort - I've highlighted problem or slightly suspect areas

It is one of the worst zoos in the world. Located in Rafah in the Gaza Strip, the zoo is impacted by the regular conflicts that take place in the surrounding areas. Some animals have succumbed to shell explosions but also to malnutrition and disease. In fact, breeders have particularly reprehensible practices with regard to animals. They live in captivity in small, unhealthy cages and are abused. In February 2019, the video of a young lioness having her nails pulled off with a forceps had been broadcast, alarming internet users.

After years of fighting, it's a victory for the NGO Four Paws. The members of the organization were often offended by this brutality and finally won their case. Israeli and Palestinian authorities have given permission to release lions, wolves or ostriches from the facility. "As soon as the border opens, we will leave and evacuate the animals from the Rafah Zoo," explains Amir Khalil who is in charge of this complex operation.

After loading the animals in a truck, the NGO was escorted by the Israeli army to head for Jordan. The animals were then deposited in a sanctuary where they would spend the rest of their lives. A total of 47 animals were rescued: this operation is currently the largest ever conducted by Four Paws.

Now Deepl's version:

It is one of the worst zoos in the world. Located in Rafah in the Gaza Strip, the zoo is impacted by regular conflicts in the surrounding areas. Some animals have died from shell explosions but also from malnutrition and disease. Indeed, breeders have particularly reprehensible practices towards animals. They live in captivity in small unhealthy cages and are victims of abuse. In February 2019, the video of a young lioness having her nails pulled out with a pair of pliers was broadcast, alarming Internet users.

After years of fighting, it is a victory for the NGO Four Paws. Members of the organization were often offended by this brutality and finally won. The Israeli and Palestinian authorities have given permission to release lions, wolves and ostriches from the facility. "As soon as the border opens, we will leave and evacuate the animals from Rafah Zoo," says Amir Khalil, who is in charge of this complex operation.

After loading the animals into a truck, the NGO was escorted by the Israeli army to Jordan. The animals were then placed in a sanctuary where they would have to spend the rest of their lives. A total of 47 animals were rescued: this operation is the largest ever carried out by Four Paws.

So both translators do a very good job on this piece of journalistic text, with a slight advantage to Deepl, I would say.


Let's try it the other way round with a piece of English journalism from The Guardian:


The Daily Telegraph has been forced to correct a column by Boris Johnson after the Brexiter MP and potential Tory leadership candidate falsely claimed a no-deal Brexit was the most popular option among the British public.

The claim was made in a column (£) published in January but has since been removed from the online version after a complaint by a member of the public to the press regulator Ipso.
Boris Johnson under fire for failing to declare house in Somerset
Read more

An online correction said: “In fact, no poll clearly showed that a no-deal Brexit was more popular than the other options. This correction is being published following a complaint upheld by the Independent Press Standards Organisation.”

In its defence the Telegraph said Johnson was “entitled to make sweeping generalisations based on his opinions”.

It also suggested that claims in Johnson’s column should not be taken seriously as the piece “was clearly comically polemical, and could not be reasonably read as a serious, empirical, in-depth analysis of hard factual matters”.

Mitchell Stirling, a statistician from Reading, said he made the complaint because he felt “a potential prime minister shouldn’t be able to make things up in a weekly column”.

“I complained because Johnson’s article went beyond him referring to anecdotal evidence around the popularity of no deal and saying it was becoming more popular (which it was) but to saying that polling was showing it was the most popular opinion by some clear margin. Both of which were not backed up by any polling done by a BPC [British Polling Council] member.

Here is Google's version, with issues in bold::

"Le Daily Telegraph a été contraint de corriger une colonne de Boris Johnson après le député de Brexiter et un candidat potentiel à la direction du parti conservateur a faussement prétendu qu'un Brexit sans accord était l'option la plus populaire parmi le public britannique.

La réclamation a été faite dans une colonne (£) publiée en janvier, mais a depuis été retirée de la version en ligne à la suite d'une plainte déposée par un membre du public auprès du régulateur de la presse Ipso.

Une correction en ligne a déclaré: «En fait, aucun sondage n'a clairement montré qu'un Brexit sans transaction était plus populaire que les autres options. Cette correction est en cours de publication à la suite d'une plainte confirmée par l'Independent Press Standards Organization. ”

Dans sa défense, le Telegraph a déclaré que Johnson était «en droit de faire des généralisations radicales basées sur ses opinions».

Il a également suggéré de ne pas prendre au sérieux les affirmations contenues dans la chronique de Johnson, dans la mesure où cette phrase "était clairement polémique au comique et ne pouvait raisonnablement être interprétée comme une analyse sérieuse, empirique et approfondie de faits factuels dures".

Mitchell Stirling, un statisticien de Reading, a déclaré avoir formulé cette plainte, estimant qu '"un premier ministre potentiel ne devrait pas être en mesure d'inventer des choses dans une chronique hebdomadaire".

«Je me suis plaint parce que l’article de Johnson l’avait dépassé, évoquant des preuves anecdotiques sur la popularité du non-accord et affirmant qu’il devenait de plus en plus populaire (ce qui était le cas), mais que les sondages montraient qu’il était l’opinion la plus populaire de loin. Aucun des deux n’ayant été étayé par un sondage effectué par un membre du BPC [British Polling Council]."

And now Deepl's:

"Le Daily Telegraph a été contraint de corriger une chronique de Boris Johnson après que le député de Brexiter et candidat potentiel à la direction du Parti conservateur eut faussement prétendu que Brexit était l'option la plus populaire parmi le public britannique.

La plainte a été formulée dans une chronique (£) publiée en janvier mais a depuis été retirée de la version en ligne à la suite d'une plainte déposée par un membre du public auprès du régulateur de la presse Ipso.

Une correction en ligne a dit : "En fait, aucun sondage n'a clairement démontré qu'un Brexit sans transaction était plus populaire que les autres options. Cette correction est publiée à la suite d'une plainte accueillie par l'Independent Press Standards Organisation."

Dans sa défense, le Telegraph a déclaré que Johnson avait "le droit de faire des généralisations radicales basées sur ses opinions".

Il a également suggéré que les affirmations de la chronique de Johnson ne devraient pas être prises au sérieux, car l'article " Ã©tait clairement polémique et ne pouvait raisonnablement être lu comme une analyse sérieuse, empirique et approfondie de faits concrets ".

Mitchell Stirling, un statisticien de Reading, a déclaré qu'il avait déposé cette plainte parce qu'il estimait qu'"un premier ministre potentiel ne devrait pas pouvoir inventer des choses dans une chronique hebdomadaire".

"Je me suis plaint parce que l'article de Johnson allait plus loin que lui en faisant référence à des preuves anecdotiques concernant la popularité de l'absence d'accord et en disant qu'il devenait de plus en plus populaire (ce qui était le cas), mais dire que les sondages montraient que c'était l'opinion la plus populaire par une marge claire. Tous deux n'ont pas été étayés par un sondage effectué par un membre du BPC[British Polling Council]."

Once again, both versions are good, only seriously being defeated by the phrase: "Johnson’s article went beyond him referring to anecdotal evidence"which needed a circumlocution of some sort, something like "il évoquait des aspects qui allaient au-delà de l'anecdote" (?).

Conclusion? Well, both are good with French at journalistic language. If you want a more scientific comparison try this:

https://www.at-languagesolutions.com/en/atblog/google-translate-vs-deepl/

I've always been in the habit of using Google, but will seriously consider Deepl in the future.


Comments

  1. Les outils de traduction automatique ont vraiment fait du chemin ! Je me demande quand ils seront parfaits

    ReplyDelete

Post a Comment

Popular posts from this blog

What is the natural order hypothesis?

The natural order hypothesis states that all learners acquire the grammatical structures of a language in roughly the same order. This applies to both first and second language acquisition. This order is not dependent on the ease with which a particular language feature can be taught; in English, some features, such as third-person "-s" ("he runs") are easy to teach in a classroom setting, but are not typically fully acquired until the later stages of language acquisition. The hypothesis was based on morpheme studies by Heidi Dulay and Marina Burt, which found that certain morphemes were predictably learned before others during the course of second language acquisition. The hypothesis was picked up by Stephen Krashen who incorporated it in his very well known input model of second language learning. Furthermore, according to the natural order hypothesis, the order of acquisition remains the same regardless of the teacher's explicit instruction; in other words,

What is skill acquisition theory?

For this post, I am drawing on a section from the excellent book by Rod Ellis and Natsuko Shintani called Exploring Language Pedagogy through Second Language Acquisition Research (Routledge, 2014). Skill acquisition is one of several competing theories of how we learn new languages. It’s a theory based on the idea that skilled behaviour in any area can become routinised and even automatic under certain conditions through repeated pairing of stimuli and responses. When put like that, it looks a bit like the behaviourist view of stimulus-response learning which went out of fashion from the late 1950s. Skill acquisition draws on John Anderson’s ACT theory, which he called a cognitivist stimulus-response theory. ACT stands for Adaptive Control of Thought.  ACT theory distinguishes declarative knowledge (knowledge of facts and concepts, such as the fact that adjectives agree) from procedural knowledge (knowing how to do things in certain situations, such as understand and speak a language).

12 principles of second language teaching

This is a short, adapted extract from our book The Language Teacher Toolkit . "We could not possibly recommend a single overall method for second language teaching, but the growing body of research we now have points to certain provisional broad principles which might guide teachers. Canadian professors Patsy Lightbown and Nina Spada (2013), after reviewing a number of studies over the years to see whether it is better to just use meaning-based approaches or to include elements of explicit grammar teaching and practice, conclude: Classroom data from a number of studies offer support for the view that form-focused instruction and corrective feedback provided within the context of communicative and content-based programmes are more effective in promoting second language learning than programmes that are limited to a virtually exclusive emphasis on comprehension. As teachers Gianfranco and I would go along with that general view and would like to suggest our own set of g