Translation is supported by machine translation (MT) tools more than ever before, and a new trend towards cooperation and standards means this will only become more common.
The world changes quickly, usually without our permission or even our awareness – that’s how you wind up an old person shaking your canes at the kids, outraged at their behaviour that everyone else finds completely normal.
In the translation industry, it seems like things have been static for decades now, frozen in a certain circumstance we can’t seem to muscle our way past, and it all has to do with machine translation (MT).
In the early 21st Century the translation business as a whole seemed to have turned a corner when it came to MT. Translators who had for years regarded MT simultaneously as a threat and a failure came to realise that under the right circumstances MT could be a tremendous help in our work, and the companies selling MT products had finally achieved a minimum of usefulness. The future seemed bright.
Frozen in Time
And yet, here we are and not much has changed. MT still seems poised to become quite useful any moment now, and translation professionals are comfortable with the idea of using it in their daily work, and yet MT hasn’t really grown much. It remains an obscure part of the translation world, for a variety of reasons.
One, there are many, many different MT solutions and platforms. When you take on work from larger companies you are very likely to be handed a proprietary MT system and expected to learn and use it, only to encounter another wholly different system on your next job. And while concepts like Translation Memory have elevated MT to a very useful status, it still isn’t the paradise of automation some thought it would be, because Translation Memory still requires a lot of set up and preparation every time you begin a project.
The Future is Coming
Nothing can stay frozen forever, of course. There are signs that things are going to begin to change in the MT arena very soon.
One of the biggest sea-changes is an extension of the new attitude. Once translation professionals stopped seeing MT as an army of robots coming to take their jobs and more as a collection of tools that will make their work easier and more profitable, the door was opened for true cooperation. Today many of the companies that build MT tools are working directly with the top professionals in the translation business, helping them design and test new MT tools, which means the tools that make it to market are far superior. Just as important, there is a growing pressure from the translation community to make sure these tools adhere to standards so we can slowly move away from the chaos of proprietary tools everywhere.
This process will take some time, and this is as good a moment as any to reiterate the simple fact that machine translation will never replace trained human beings, only support them. Still, the future looks bright for the translation world.
Image courtesy blog.operasolutions.com