There has been a lot of recent discussions and bold claims about whether or not machine translation will make human translation obsolete. There are strong arguments in the case of human translation vs machine translation as to why human translators could never be outdone but even if it were possible, machine translation capabilities are nowhere near that level of ability nor will they be in the foreseeable future.
There are approximately 6,500 spoken languages in this world, each with different dialects, cultural influences, alphabets, grammatical rules, and many other nuances that make the variables of translation innumerable. Linguists and programmers are tasked with not only identifying all of these variables but also quantifying them into usable data for machine translation algorithms. It certainly seems like an impossible feat.
Human Translation vs Machine Translation
In addition to technical accuracy, another fundamental aspect of translations is conveying style and tone. How something is said is as important as what is said, which relies heavily on word choice and cultural understanding. Machine translation trails far behind human understanding of context and their ability to express ideas in a specific manner. There are several conundra programmers and computers have difficulty overcoming.
1. Translating words with multiple meanings
Many words have multiple meanings depending on the context they are used. Whereas humans can naturally infer which meaning of a word is being referenced based on the topic and words around it, computers struggle to do the same. For instance, the word “pool” can refer to a small body of water, to a group of people or vehicles, or the act of collecting resources or efforts. Clearly, one word can have greatly varying definitions. Furthermore, machine translation software that uses verbal input can easily “hear” the word incorrectly, misidentifying “pool” as “pull” or even “pole”. A person is far less likely to make a similar mistake.
2. Translating style and tone in human translation vs machine translation
Conveying the proper tone in translation is very difficult but it is vital for the translation to be accurate and complete. Translation software is designed to replace the provided words with the most commonly used vocabulary. The problem is, using the most common vocabulary is not always the best fit. For example, there is a tonal difference between synonyms like “angry” and “annoyed” or “excited” and “agitated”. Machines do not yet know how to discern between these nuances and maybe never will, which may be the biggest defining difference between machine translators vs human translators.
3. Literal translations vs figurative translations
One of the most humorous shortcomings of machine translations is the literal conversion of phrases that result in a mistranslation. For example, the literal translation of the Cantonese words “syut gwaih” is “snow cupboard”, but the proper translation is “refrigerator”. These errors make us chuckle unless, of course, it is your translated work that becomes the butt of a joke.
4. Cultural differences and their effect on translations
There are some words and phrases that have harsher meanings in other languages. For instance, in English, the word “stupid” is often used offhandedly; however, the direct translation in Japanese, “baka”, can be considered highly offensive. A human translator can help you express your thoughts and ideas without crossing any cultural lines that may lead you to inadvertently insulting your audience.
5. Casual vs professional translations
In most languages, there is a big difference in the way peers communicate with each other compared to superiors and subordinates or among business professionals. You do not want to come across overly formal in a casual interaction nor disrespectful in a professional sphere. Machine translators do not navigate these waters well; It is much better to put a person at the helm.
6. Time vs the machine
Languages are changing every day, constantly evolving with new jargon and new cultural inputs. New terms like “buzzy” and “on-brand” are often widely used well before being officially recognised as words. People catch on to current language trends and advances much faster than can be officially tracked and inputted into machines, meaning a machine translation database can be considered obsolete to a human’s vocabulary.
Costs and Benefits of Human Translation vs Machine Translation
1. Speed of delivery
Computer-driven translation software is often marketed in a way that highlights the speed of delivery, asserting human translations will take much more time to complete. There is truth behind the statement but the extent of the time requirement is greatly exaggerated. Easy-to-use online human translation services have greatly reduced the turnaround time from an order being placed to a translation been delivered. Furthermore, the margin of error is so much smaller that it saves the customer time being wasted on corrections and more importantly, embarrassment of using weak or incorrect translations. Time is valuable, but not at the expense of quality.
2. Cost of translation
Many believe free google translations or other free services will have to suffice because human translations are too expensive; however, those who research it further are surprised at how affordable translation services can be. Online translation services have greatly cut down on overhead and upfront costs, allowing professional translators to charge a lot less for the same great quality product. True value is found with human translations.
3. Certified translations
Certified translations are commonly required when submitting translated documents for official purposes. A certified translation service provides a stamp of approval and statement declaring the document is a true and accurate translation of the original. Machine translations are not able to certify translations.
Will Human Translators Become Obsolete?
It would be arrogant to say machine translation will never be able to do what humans can but that day, if it ever comes, is far into the future. The arguments of human translations vs machine translations still land in favour of humans. Computers today simply cannot produce a comprehensive translation like a real-life person can that considers culture, tone, diction nuances, and more. For the time being, there is no better alternative to the thinking and feeling human translator.