Human-aided machine translation

Alikberova Maftuna 2301-80

Human-aided machine translation

Today, computers are used in all fields, and even almost every field has its own software packages. Using computers to translate text from one language to another is called machine translation. It is  common knowledge that machine translation software can translate texts very quickly. Machine translation is not a perfect translation tool, nor is it a valid translation method. Let's consider at a few of the disadvantages of machine translation from different languages to English.

Firstly, translation is not only a linguistic but also a cultural act. The quality of machine translation continues to improve, and the idea of using machine translation instead of human translators is becoming increasingly attractive, but without human translators, machine translation cannot capture the cultural subtleties of translation. For example, a Russian translation of a story by French writer Jean Laffitte - "Nous retournerons cueillir les jonquilles" - with machine translation would read as follows: "We will return for the jonquilles". The concept of "jonquilles" is unknown to the general reader, so a human translator would use an adaptation here and replace "jonquilles" with "snowdrops", giving the title an emotional meaning. For Russians, a snowdrop is a symbol of hope for the best, for the arrival of spring, and with it, new joy and happiness. After all, without human help, machine translation couldn't to understand and replace the cultural subtleties in the text.

Secondly, there are a lot of polysemous words in English. On the one hand, this is convenient – you need to remember fewer words to convey the same amount of information. But on the other hand, it is confusing for machine translation, because sometimes it is hard to know which of the multiple meanings applies. Machine translation cannot take double meanings into account, for example, when translating the sentence “Today I saw my Honey” with Google Translate, you get “Я пилил мой мёд сегодня”, but in this case “saw” is past tense of verb “see” which translated as “видеть” and “Honey” in this case is a reference to a loved one, the correct translation of this sentence would be: “Сегодня я виделся с моей милой”. To summarise, the concept of ambiguity makes machine translation very difficult. Moreover, an incorrect translation can distort the original meaning of a sentence and even the whole text.

Thirdly, when translating slang, the specific context in which it is used must be taken into account. Slang is a colloquial, non-literal, form of colloquial speech that has a pronounced emotional and evaluative colouring, humour, language play and trendy neologisms. Machine translation cannot translate slang in a way that is correct for users. An example would be: “Hey, man. How’s it going?”. In most cases machine translation uses the technique of colloquialism.  Colloquialism is the literal translation of a phrase, and in this case machine translation will translate “man” as “мужчина”  rather than “чувак”, so you would get “Привет, чувак, как дела?”. “Bob’s your uncle” literally translates to “Боб твой дядя”, although the slang has a different meaning. This slang means something like “Вуаля!” Let’s look at the dialogue as an example: – How did you make this cake? It’s delicious! (Как ты испек этот пирог? Он восхитителен!) – Well, I just mixed in the batter thoroughly, poured it into a cake pan, baked it for 30 minutes and Bob’s your uncle! (Ну, я просто хорошенько смешал тесто, вылил его в форму, выпекал 30 минут – и вуаля!). Thus, poor knowledge of slang means incompetence compared to translation by a professional translator, a loss of meaning in the text.