Introducing Word Score

Translation services are now essential to everything from international politics to the growth of many businesses.

People have been translating texts from one language to another for thousands of years. But things have changed considerably since the creation of the Rosetta Stone!

These days, translations are rarely carved in stone, inscribed on clay tablets or presented on beautiful, hand-written manuscripts.

Indeed, many translations are completed without any human input at all.

From CAT tools to AI, technological advancements have changed the translation landscape forever.

But translation is an art, not a science. Languages and cultures are diverse and nuanced in ways that at least right now, only humans can fully understand and interpret.

In other words, how translations have been produced really matters.

We are all constantly exposed to content that has been translated but how do we know the extent to which we can trust what we see or what we hear?

No traffic lights for translations

Advancements in technology have largely been designed to make life easier. However, in many ways, they have served to make everything we do more complicated.

Take washing machines, for example. There was a time when washing your clothes involved nothing more than scrubbing them in a tub of water and then hanging them out to dry. During the industrial revolution, rotating drums were invented. They were turned by hand but ensured that the washing of clothes was less arduous. There was only one type of machine and consumers could choose to invest in such a machine – or not.

Electrically powered washing machines appeared at the beginning of the 20th century and changed everything. Over time, washing machines became ever more complex and diverse. It wasn’t long before the machines offered a wide variety of programmes and impressive functionality. Consumers were being presented with much greater choice. But machines varied in their energy efficiency and water consumption which became problematic as awareness of environmental issues grew.

It was relatively easy for consumers to identify which washing machine boasted the functionality they needed. But how could consumers decide which washing machine was the most energy efficient?

Entrance energy efficiency ratings. In Europe, machines are rated on a scale of A to G, with A being the most efficient products, and G being the least efficient. To keeps things clear and simple, the ratings are presented on colour coded charts with shades of green used for the best ratings and shades of red used for the worst ratings.

This type of “traffic light” system keeps things simple and has proved to be extremely helpful. Similar systems are used to identify the energy ratings of home and the nutritional quality of food.

It isn’t hard to see how scoring systems can inform purchasing decisions and therefore the value of goods and properties.

There is currently no such system for identifying the nature of translations. Perhaps there should be!

Red light spells danger

We are all exposed to translated material every day. However, we have no way of knowing whether the translation of the source text was performed by an inexperienced person, an expert or artificial intelligence (AI). Neither can we deduce whether that translation was edited to ensure its accuracy.

For this reason, none of us are in a position to decide the extent to which we can trust what we are reading.

We have previously highlighted the potential dangers of AI and the issues that have already reared their ugly heads regarding translations.

Poorly translated song lyrics or subtitles don’t usually matter that much. It’s a very different story with poorly translated safety guidance or instructions for pharmaceuticals.

Isn’t it time that a traffic light system was applied to translations?

Red labels, if not red lights, could indicate danger!

All about Word Score

At Word Connection we are working to evolve Word Score, a scoring system for translations.

This would indicate how translated content has been created and whether it has been edited. We propose that this would be a self-assessment system but one that should be used honestly and ethically.

Just as with nutritional scores for food, Word Score would help readers assess the credibility of translated text and therefore whether they can rely on its accuracy.

Word Score could also be utilised in the translation industry to assist buyers of translation services to make informed decisions about the suitability and quality of the work they are investing in.

We propose that Word Score should feature ratings from A to F as follows:

A - Fully certified translation by a translation expert and with full human post-editing.

B – Human translation by a less experienced translator with full human post-editing

C - Human translation with no post-editing

D – Machine translation with full human post-editing

E – Machine translation with minimal human post-editing

F - Machine translation with no post-editing

The value of human translation

Word Score would certainly help consumers to evaluate the reliability of content. It could also prove useful to purchasers of translation services.

We hope that the new system will also help to emphasise the importance of human translation skills which are currently undervalued. Translation professionals are often underpaid for their expertise and so many are leaving the industry, taking their skills with them.

If translated content was labelled using Word Score, consumers would soon start to take notice and to prioritise A-rated content. There would be greater demand for that content and its value would surely rise.

Would anyone publish F-rated content if they knew readers would simply ignore it?

Word Score could revolutionise the way both consumers and translation buyers perceive the value of superior translations.