Are you searching for the best option to have your text translated?
What did you find when you typed in translation or translator?
If you're like me, you most likely got results for:
As tempted as you might be to click on one of these links to translate your 5,000 word website for free, maybe just wait a second.
Although you may already be able to guess my position on machine translation, I'm not here to condemn you.
Well, if I were then you could always stop reading this article and move on.
No matter what you do on the internet these days, you should always try to make an educated and well-researched decision - this includes translating your texts.
Here, I'll share with you my knowledge on machine translation, when it could actually be ideal for you (and when it's not) as well as some considerations you should keep in mind when using free translation software.
A (Very Short) History of Machine Translation
Machine translation can be traced back to the 1950s where the idea for it was first articulated and thoroughly researched.
The first translation broadcasted was carried out by IBM in which Russian sentences were automatically transferred into English with machine.
Alas, the quality was not so great. A report by ALPAC (Automatic Language Processing Advisory Committee) concluded that machines could not compete with human translators in terms of quality.
However, like in many instances in human history, some people could just not take no for an answer.
A handful of researchers continued on their search to build higher quality translation machines - at this time using a rule-based method (the machine's translations are based on inputted rules).
Today, neural machine translation is the most common and most natural sounding because it learns as it goes - think Google Translate.
When Some Think You Should Use Automated Translations
Now at this point, you might be thinking - you put natural sounding and Google Translate in the same sentence. That must mean Google Translate is the way to go!
Not so fast.
Like I said, there is in fact, a time and place for machine translation, but that still doesn't mean Google Translate is the best option.
When should you take advantage of free online translations?
Various translation agencies or language service providers will give you different answers.
content that doesn't produce high revenue or "straight forward" content (Smartling)
only your "most unimportant documents" (LanguageWire)
internal communication, documentation/instructions, short-lived texts (Acolad)
But is that a good idea?
Now, as great as it sounds to have a list of "dos" and "don'ts" when it comes to machine translation, it's really not that black and white.
First of all, let's discuss LanguageWire's suggestion of only automatically translating items that are deemed "not very important." How do you define "not important?" It most likely depends on your perspective - in other words, it's subjective.
Additionally, why would someone bother to have “unimportant” texts translated? Doesn’t it seem like a waste of time and resources?
How about Acolad's suggestion to use MT to translate internal communication, instructions and "short-lived" texts?
I don't know about you, but I'm not sure I want my colleagues to read translated messages from me that have been machine translated.
Just consider this for a moment:
We want to present ourselves professionally in the workplace. This means we carefully choose our words when writing emails or communicating with others. We may even use cultural references or slang to build rapport with team members. Have you ever tried to enter slang into Google Translate?
Documentation and instructions are created for a reason - I assume no one writes these types of texts for fun. They're intended to improve efficiency and reduce misunderstandings. Confusion or mistakes as a result of misunderstandings can increase costs in the end.
Lastly, how should we understand Smartling’s suggestion of only automatically translating that is “straight forward” or that “doesn’t produce high revenue?”
Although Smartling’s tip on when to apply machine translation sounds the most logical, I still wonder how a company should decide what kind of text “doesn’t produce high revenue.”
My point is, like in a team, where (you hope) each component functions optimally, you most likely want your texts both for internal and external use to function optimally. Because we know what happens when one element of the building blocks is weak: the whole structure comes tumbling down.
A Better Idea
Instead of making your important what-to-translate and what-not-to-translate decisions based on other people’s rules, why not actually think through your own strategy?
What is the purpose of translating the given text?
What does a successful translation look like for your company/team?
What would a failed translation mean for your company/business/team?
What would a successful translation mean for your company/business/team?
What would you do if the translation failed or didn’t produce the results you expected?
How is your budget allocated? Do you plan on investing more money upfront or in smaller amounts along the way?
These are some questions you and your teams should be asking yourselves because unfortunately, no one can answer these questions for you.
Free in What Sense?
Finally, I would like to add a note on the notion of “free machine translations.”
My question for you: Do you believe online translations are truly free?
If your answer is yes, then you might want to rethink your answer.
If your answer is no, then do you know why?
As mentioned previously, online automated translations are not, in essence, free because of the time investment as well as potential cost investment to fix errors.
However, have you ever considered that you pay for these translation services with your data?
According to an article from ATA, machine translation tools available to the general public at no cost (Google, Microsoft, DeepL) often save your texts to improve their own translations.
Now, if you pay for a premium version or the translation tools are accessed through a translation system, the data is not stored.
Either way, by using these automated tools, you pay in some way.
Whether you decide to go the machine translation route is entirely up to you. Just remember that there is no clear-cut answer as to when you should and shouldn't use automated systems. The answer depends on your own needs and your company's goals.
Talk with your language service provider and your team. Decide what makes the most sense for you after doing thorough research.
Hopefully this article helped you to get one step closer to finding the best solution for you.