First of all, raise awareness in your students that GT is probably better than they think.
You can try out two experiments.
The fi rst experiment is to get them to write an email directly into GT. For example, if they are Portuguese speakers, suggest that they write directly into the Portuguese box using English words and expressions if they know them, and any words or expressions that they don't know they can write in Portuguese. This means that they
will write in a mix of Portuguese and English which will be rendered just in English in the English box. Below is an example:
The above example produces a pretty good English version but inexplicably returns a lower case i ( i could ).
Of course, if there is a mistake in the English that the Portuguese writer types in, then Google will not miraculously correct (so to hear doesn't change to hearing ).
Below is a version with a straight Portuguese / English translation with a major mistake: if you can get rather than if I can get . However, this is the kind of mistake that is relatively easy for the student to identify - as is the i in i could which would be picked up by a spell checker.
The next step in the same exercise is to get the students to check the individual parts of the English versions of their email on google.com. For example, if they type in "I look forward to hear from you" (note that the searched for phrase must be typed in with inverted commas before and after), they will get this return:
Note that although the phrase gets 118,000,000 results, the fi rst three are a clear indication that there is something particular about the phrase "I look forward to hear / hearing from you". In fact, the fi rst returns are all from grammar and vocabu- lary sites rather than real examples in use.
So the next step is to verify the authenticity of the sites. The fi rst is a famous inter- national site (wordreference) and can therefore be trusted. The third and fourth also look reliable. But, at least to a native-speaker's eyes, the second looks a bit dodgy - 'jakubmarian' doesn't sound English at all: in fact, Jakub turns out to be a Czech living in Germany! Jakub does actually get it right in his explanation but students do need to learn to distinguish native from non-native sites.
Now let's try the phrase "I await your return"
These fi rst three returns should sound alarm bells and hopefully will alert the Portuguese student that the Google translation is not correct.
And so on … the student could also look up "thank you for your attention" and discover that, while this sentence is correct English, it tends to be used at the end of a presentation, rather than at the end of an email.
So the student then has to decide what to do next, i.e. to fi nd an equivalent to "I await your return" or simply delete the phrase. Finding an alternative is not diffi cult if they use context.reverso.net or linguee.com.
The next example regards the translation of an abstract.
Follow this procedure:
1. Ask students to prepare an abstract in their own language (or use an existing one). Then they should translate it into English noting how many minutes it takes them.
2. Students go through their version (or the version of a fellow student with the same native language), identifying as many mistakes in their abstract as possible.
3. Students submit the original version to GT and correct the resulting version again, noting how many minutes it takes them.
4. Finally, you need to count the number of mistakes in i) their original manual version ii) the GT version iii) the GT version corrected by them.
This exercise should either prove or not prove (there is no certain result either way) one or more of the following:
• GT is as good at translating - if not better - than the student
• It's easier to spot a mistake made by GT than one's own mistakes. In fact, the human eye tends to be more alert to other people's mistakes. We don't actually seem to be able to 'see' some of our own mistakes
• it's generally a lot quicker to do a GT and then revise it than translate 'manu- ally' (and then still have to correct it)
• GT throws up ways of translating phrases that you may not have even thought of and which may even be better than your own ideas