With the help of AI (or artificial intelligence), machine translation is getting better and better. It’s scary how accurate the free new technology is becoming. Free, online machine translation tools, such as Google Translate, DeepL, Microsoft Translator, and Apple Translate have great potential to help you improve your second language skills. On the other hand, and most unfortunately, these tools might also prevent you from learning. Should you use machine translation, then? Yes, I think so, but only if you use it properly. What are some bad ways and good ways to use these tools when trying to improve your English? Let’s find out.
Let’s start with a couple of ways you SHOULD NOT use machine translation.
Don’t use it like a dictionary.
Well, many people use tools such as DeepL like they would a dictionary: Instead of a definition, they’re looking for a shortcut to understanding the meaning or a word or phrase through translation. So, they type the word in their native language and get an instant translation of the word into English, or whatever language they’re learning. This can work, but often, it doesn’t. This is because you aren’t giving the system any context or cultural cues (or hints). Machine translation works best if you give it some context. (Context refers to the parts of something written or spoken that come right before or after the word you type in that help to make the word’s meaning clear.)
Here’s an example of a translation query that failed. I go to DeepL, and after selecting French as the origin language, I write the word “avocat” because maybe I’m a native French speaker and I want to know how to say or write “avocat” in English. DeepL gives me the word “lawyer” or “attorney.” I can scroll down the page and even get more information on possible translations. However, the software is only taking one possible meaning of the word here. Is this the word I’m really looking for? Let’s see what happens now if I give more context to DeepL by using the word in a sentence: “J’ai mangé un avocat.” Oh dear. We can see that the English word I was actually looking for is “avocado.” I don’t recommend that you say, “I ate a lawyer.”
Don’t use it to translate FROM your native language (as is).
Another way machine translation is not useful is if you copy a chunk (or a piece) of text you’ve written in your native language, paste it into Google Translate or DeepL, and just use the resulting text in your homework assignment, for example. Well, this is not good for two reasons.
Academic Dishonesty
First, as you’ve probably already guessed, this is a form (or a type) of academic dishonesty (or cheating). You made the translator do the hard work instead of doing it yourself. The machine is not a person, but it’s still dishonest to get it to do your work for you. Furthermore, some learners might not even write the original text by themselves. They might copy and paste text they find on the Internet in their own language and put it into the translation system. Any maybe they don’t even cite (or reveal) where they originally found the information. This is plagiarism, another type of academic dishonesty. I have two videos on the topic of plagiarism which I suggest you check out. I’ll link to them in the description box below, and up here somewhere.
Not Learning That Much!
A second reason why it’s not good to machine translate text from your first language into English is that you won’t be doing much to improve your English skills. Learning a foreign language requires struggle – a degree of pain and suffering. There’s not really any escape from this. By passing on (or giving) the struggle to Google, DeepL or some other translation tool, you’re skipping a crucial (or essential or very, very important) step in learning English. Even if the result is faster and more accurate than you could have done by yourself, you’ll likely have wasted time. And you’re also wasting your teacher’s time, by the way. Even when the software is amazing, many teachers can identify when you use machine translation, because the results are either strange and/or don’t match your real level of English.
Now, are there times when you can use machine translation in this way? Sure! I plan to use it for this video, because I want to make subtitles in different languages so that people who don’t understand my English can get some help trying to understand my message. But, let me be clear: my goal in this situation is NOT to learn those languages, and I don’t have to submit it to a teacher for homework.
OK, now let’s move onto how you can use machine translation to help you learn English. First, you should know that translating between languages (with or without software) is actually useful because it can reveal structural differences between the two languages, as well as many similarities they might share, such as vocabulary or word order. Using a translation tool is therefore useful for helping you to understand how English expresses or conveys a message compared to your native language. Machine translation to get feedback on your written and spoken language is a great way to use the technology. You might develop higher levels of linguistic awareness. How can you do this? Well, let me give you one example.
Use it to translate TO your native language!
Use the tool in the opposite translation direction. For example, if you’re studying English, which I assume you are since you’re watching this video, try to write some original text in English – perhaps a few sentences or a paragraph. You could also limit yourself by writing English for a set (or decided) amount of time or for a set number of words. When you’ve finished, review what you have written to do an initial check to see if you can identify any of your own mistakes, or to see if you can write some parts in a better way. Next, copy your English text and paste it into the machine translator with English as the set language of origin. Set the tool so that you translate what you have written into your native language and translate it. How good is the translation? In some tools, such as Google translate, the accuracy of the translation is open to public feedback. In this image, you can see that several people have approved the translation and offered alternatives, which can be very helpful to you when you’re judging how accurate the translation is. Next, based on what you see in your native language, make corrections to the English until it’s better. You can also look at alternative suggestions made by the translator. Back translate to make sure it all makes sense. (Back translate just means reverse the translation from your language into English.) Finally, edit further if you need to. And if you want to help yourself, remember what you’ve learned from doing this activity, you can take notes on points to avoid or focus on for next time, as well as any new vocabulary you may have learned.
Clearly, if used properly, machine translation can be an amazing tool help to you in your language learning. It’s already an everyday technology that helps professional translators do their jobs more efficiently. But remember that no matter how fantastic translation software continues to become, it will probably never fully replace human translation, at least not for a while longer. This is because because each act (or each bit) of communication is as unique as the person that first spoke or wrote it, and as unique as the cultural context it came from. A trained human translator looks at the true meaning behind words. A machine, on the other hand, treats language as a science based on data – only data.