During WWII, the electromechanical cipher machines were used. It was these cipher machines, along with America’s strong desire to be able to read Russian technical documents, which motivated scientists to create computer translation machines. The distrust that fueled the Cold War between the US and Russia produced an environment where translations were needed quickly and often, so much so, that human translators simply could not keep up.
This brief history is the birth of machine translators like Google Translate. Before Google Translate, machines that translated used simple substitution of words in one language for words in another. Google Translate on the other hand, doesn’t bother to translate or follow grammatical rules. It simply scours the web where it uses a statistical method to find the most probable match. Human translators have already translated millions of documents. Google Translate uses these documents to search through to find patterns that will help to determine the best translation. Many times Google does not translate directly from one language into another, but rather translates into English than into the desired language, this can cause significant errors.
Currently, Google Translate, which was launched in 2001, accommodates 103 different languages at various levels and assists over 200 million people daily with their translation needs.
Overall, Google Translate does a good job of getting the principle idea across, but still manages to skip over the many nuances of language and tends to suffer from poor word choices. There many jobs that Google Translate does well, such as: short and simple translations; conversations between friends/family; menus; road signs and anything that just understanding the “gist” of is all that is really needed.
Business, financial, medical, legal, civil or advertising documents that need to be concise and clear should be handled by human translators.