Language Translation

Introduction

From time to time, we receive a trickle of enquiries, (and I do mean a very small number), from people who would like to use our products in languages other than English. I have to say first of all that I think it is wonderful, and it is very gratifying when, having invested a lot of time and effort into creating something, a person who does not speak English wants to use it.

I am sure over time we will indeed have some of our products translated into some widely used languages. But I thought it may be helpful if people gained an understanding of the enormity of such a project because I would imagine that some of what I am going to describe perhaps would not have occurred to you.

What Exactly is Localised Support?

Every so often, I see Emails or Twitter posts from developers of JAWS scripts stating that they support a particularl language. But what does support actually mean? How do you define it?

In a number of cases, it means translating various English prompts spoken by a screen-reader into the desired language and perhaps a small amount of documentation. Is that "supporting" people in the true sense of the word?

I believe support goes way beyond that. From the time a person picks up the telephone or sends an Email to make an initial enquiry, through to the purchase (and usage of) the product, and ongoing technical support with possible training, that should all be delivered using the person's mother tongue. That is true holistic support. The person would be paying the same monetary value for the product, perhaps a little more in fact, and so he or she is perfectly entitled to receive the same level of assistance and overall quality as someone who speaks English and nothing less.

To provide that high level of assistance, there would have to be a very cohesive relationship between ourselves and a partner organisation in the project, both of whom would need to invest financially and having due consideration as to how it was going to be sold and supported. Unless the product sells, there is no value. Clearly from our point of view, there would need to be complete trust placed in the partner, and fortunately, Freedom Scientific do have distributors who are used to the localising process. Undertaking market research would also have to be done to determine the need for the product in the country concerned to ensure financial viability.

So what exactly would be involved in localising a product on the scale of those we develop?

Understanding language and culture.

Clearly from the outset, the partner would need to understand each and every one of the many concepts of our products. But in turn, we need to understand the culture of the language speaker. I suggest localisation isn't just about a literal product translation. There is much more to it.

As a very basic example, with our J-Say product, one phrase to activate dragon's microphone is, "Listen to Me". But that phrase may not sit well with someone who speaks another language, such as French or Spanish? It may be easier for a person to use an alternative phrase which is not a direct translation of the words. It has to be "natural language".

Other important issues would be to consider whether the localised access technologies on which the products were based contained the same functionality as in English releases. Usually they do, but not always, or there may be subtle differences. Keyboard changes are also very worthy of consideration.

Translating the Product.

Next comes the sheer volume of time involved in translating the product. That means translating (or to be more precise localising) every aspect of it, rephrasing the help utilities, making high quality audio recordings of prompts in some cases and finding the most appropriate voice talent, translating the documentation (anywhere from 100 to 500 pages per product), and many other items not listed. For example, two of our products contain a radio player. The stations need to be chosen reflecting the given language, both for technical quality and suitability. Of course they also need to be maintained to ensure continuity.

When the product is completed, of course it needs to be tested thoroughly. This would involve the recruiting and joint management of a beta testing team. The partner organisation would play a key role here, since obviously there could well be language differences when trying to communicate a problem with the software.

Selling the Product.

So our product is all beautifully translated and ready to go. Staff training may need to occur if the sales, training and support personnel have not been involved with the conception of it.

Then, the marketing needs to take place. Remember, in man hours alone, thousands of dollars have already been invested to bring the product to market so now it is time for the return on the investment. So the people doing the marketing need to be very aware of publications in which to advertise, podcasts on which to promote it, ensuring the company's website contains lots of information to keep people interested, not to mention meeting people and talking to them. In short, the whole point of this entire process has been to ensure that the product can potentially reach as many people as it is able to, so this is where the hard work begins.

What Happens Next?

People now begin to use the product and probably report problems, although hopefully not too many. We should also bare in mind that some of the people potentially calling in will not be conversant with computers, and so those providing the support will need to spend time with them, again to ensure they are getting a good experience.

What should also be happening is that a similar process to that described above begins again with the adaptation of new features which may be part of the next release. And remember, we're just talking about one product here out of a number.

Summary.

As you can see, there is a good deal more to the process of localising a product than perhaps meets the eye, and I do applaud the screen-reader product translators particularly who have a lot of work to do. But I hope this blog post has given you a little understanding of what potentially could be involved. Absolutely every part of the above is essential in my view, and a good deal more. Able computer users may be content with a product where the localisation only exists in the program itself, and that is absolutely fine. But for some individuals, I really do not believe that is true and certainly if a product is going to sell and reach the people for whom it is intended, I would suggest all of those strategies need to be in place.

I come back to the starting point of this post. I am sure in time language translation is something we will think about. But we would need to be convinced that there was a market for doing so. If we were going to do this, it would be done properly and would be well thought out. As I hope you can tell, I have already given the process a good deal of consideration.