O n f e r e n c e

machine translation
experience students are supposed 
to prepare their own translation of a short professionally-oriented text from English into 
Russian (Kalinicheva & Popova, 2014). Then they are encouraged to “communicate” with 
automatic translators 
Promt
, 
Translate
, etc. and prepare the second machine-translated 
variant of the same text for further comparison of the two versions. For the comparison 
of their own translation version with the machine one, students can use at least two 
algorithms of their learning activities, with the first one being proper for a conscientious 
student and the second one fit for a “lazy” guy: 
1.
Creating one’s own version of the translation in electronic form, then getting 
machine translation (MT) of the initial text and printing out both variants leaving 
large spaces between the lines. Reading both versions carefully and identifying 
the differences, students are supposed to correct grammatical, lexical and stylistic 
mistakes in machine translation. Correcting the printed version of the machine 
translated text is the final stage of completing this task.
2.
Getting several variants of machine translation of the initial text, selection of the 
most appropriate one from the point of view of the student, then creation of his 
"own" MT-based version, and underlining the differences. 
In either case, the student is prepared for interactive pair work in class, which, we 
believe, is supposed, to some extent, to counter-balance the difference in the fair and 
simplified home preparation.

---

-204- 
2014 CALL Conference 
LINGUAPOLIS
www.antwerpcall.be 
During pair discussion of exchanged independently produced translation options any 
student can tell the difference between the translation versions, and strong students are 
usually able to comment on it.
"Friendly" rating or peer mark for the presented translations is given drawing on pre-
defined teacher’s criteria taking into account grammatical, lexical and stylistic errors. 
Students have to get used to the idea of the most “significant” lexical errors, while 
grammatical and stylistic errors made by the computer should be treated more leniently. 
The skill of error gradation according to their importance for the translation of a special 
text is, as manifested in peer communication, an important shred of knowledge, which 
should contribute to the develop
ment of learners’ language competence. 
Alternative approach principle may be also illustrated drawing on