Achieving the Hope and Promise of Using Technology for Bible Translation While Avoiding the Pitfalls
The Word of God is what changes lives and makes the growth of the Church possible in every people. Yet there are still thousands of languages that do not have a complete Bible and thousands more that have no access to Scripture or only just a few portions. As recently as the year 1999, Wycliffe Bible Translators estimated that it would take eight generations or around 150 years to get the Bible into the hands of all the peoples who need one. Imagine being able to complete a New Testament in a language with no Scripture in less than a year when it used to take 7 to 10 years. Imagine being able to complete the initial task of getting the Bible translated into every language group that still needs one in just the next 10 years, not 150. That is the potential when mixing new machine translation technologies like the Sovee “Smart Engine” Bible software with the efforts of the people who will actually be using the translation. But is it possible for this dream of rapid Bible translation to take place and still maintain high quality translations? What are the pitfalls of these new technologies and how can we avoid them?
Some form of “machine translation technology” has been around for decades. Why are today’s results any different than what has been produced in the past (which was considered inadequate)?
The big pitfalls have been centered on lack of quality (we want God’s Word to be accurate and correct) and lack of work flow process to ensure that quality.
Doing translations of the Bible is a sacred honor and should not be taken lightly. On the other hand, we should not oversimplify a process simply because we can do it faster now while achieving the same level of quality output.
More and more global Christian communities are feeling that they know their language better than anyone else and have the theological understanding to make good translations possible. The use of technology (in any form) is only there to help make individuals more efficient, not to replace those individuals. If anyone suggests anything different, they are falling into one of the biggest pitfalls.
To put it bluntly, machine translation technology should not replace ANY individuals but rather make those same individuals who will be responsible for making quality decisions (regardless of the process used) more efficient. When used in this fashion, we should embrace any technology that helps us accomplish the task of seeing the Word of God in everyone’s language within our lifetime (preferably the next 6-10 years).
Imagine in less than a decade having a Bible in every person’s language (both written and oral). This is the very real possibility that Bible translation technology allows us to consider.
What does a viable work flow process look like?
The biggest pitfall to address is the quality issue. Once this is addressed, then scaling or rapid production can be looked at while having the same underlying core value of quality.
Historically, Bible translation was done by a team of people that then had a designated group or person checking for the quality and accuracy of that translation.
Nothing really changes that in the context of using Bible machine translation technology to help make recommendations to the initial translators so the first draft (NOT finished product but draft for review) can be accomplished. It is just done faster than by doing the recommendations manually.
Research has shown that proof reading and editing a suggestion can be between 5 and 50 times more efficient than translation from scratch with no recommendations. With today’s modern Smart Engine technology, the “Smart Engine” has the ability to learn words, phrases, meanings and context. This allows for suggestions to improve very quickly even in languages that have very little translation work previously completed.
In order to avoid pitfalls, it is critically important to understand that the Bible translation process using a tool like the Sovee Bible App combines human and machine translation, stressing human involvement at all levels.
This method takes a more iterative approach to community production of a translation. That is, trained theologians interact with the Sovee generated drafts; then mother tongue speakers improve the drafts. Finally the local church reviews and improves the drafts.
As a result, more people from the community are familiar with the translation and contribute to its successful completion. It becomes their translation! The Sovee translation engine becomes more intelligent with each improvement. Additional Sovee-generated translations produce better rough drafts with less improvement needed. This method is not product-oriented, but rather community process-oriented.
Consultants review the final drafts when all parties believe it is ready for wider distribution.
The biggest pitfall of using machine translation technology with no checks and balances is avoided using this kind of community based and expert based work flow process.
Does it really work?
We now have examples of many languages that have finished the New Testament in 6-9 months. These languages have sometimes had some verses or chapters already completed which were utilized to train the Smart Engine to help make better suggestions and in other cases had no known scriptures already completed and the whole project started from ground zero.
There is no question that utilizing this technology with a good work flow process to maintain the quality output can and is producing New Testaments and complete Bibles in much shorter time periods than in the past.
Does this mean that the old ways were a waste of time or not good stewardship of resources? The answer is emphatically NO!!
All the technology has done is looked at different steps of already proven processes and determined if there were ways to speed them up without compromising the quality output. The ability for the current technology to be created would have been an impossibility if the models for Bible translation and good work flow processes had not already been established.
Quality vs. Quantity in Bible Translation – Is it possible to have both?
One of the big myths about using machine translation is that it must by default compromise quality. How can you speed up the process without taking short cuts?
Just as computers have helped accelerate many of the tasks we work on, so too can the software that helps us with Bible translation technology.
As stated in the work flow process above, the machine output has nothing to do with the final quality control checks. This is and always will be reserved for those who have the theological training necessary to ensure that the Word of God is not marginalized through a translation process.
The technology just helps us get drafts into the hands of the final verification people much faster. There are many verses in the Bible that really are not hard to translate. So and so begat so and so is pretty easy and doesn’t lend itself to much interpretation. Using a Smart Engine to do a draft of the entire Bible can take as little as 11 minutes. That is the quantity part. Doing all of the careful proof reading, checking, editing and so forth is what takes the 6-9 months to get a New Testament completed. That is the quality part.
A big pitfall would be to believe that a 100% accurate Bible can be translated in about 11 minutes versus the reality of needing that 6-9 month time period for peer and theological review to get to a final community approved version.
However, using this technology, we can see a gospel (like maybe John or Luke) start to be available in about a month to an indigenous church for preaching and teaching. Imagine what it would be like if you were in a language that never had a Bible before and the idea of getting the Word of God to preach from in your language could start to happen that fast. Imagine also that you have been waiting your whole life to see this happen and the joy that you would feel to finally see it come to reality.
One pitfall to avoid would be to let the temptation of rushing the translation out the door before these final proof readings, edits and checks were done. How do you avoid that? I think Timothy 2:2 says it best by entrusting this work to faithful men and women. While I have not been involved in every country of the world, I have worked with Christian leaders in at least 150 nations. And based on that experience, I can tell you that there are faithful men and women in every country who take serious responsibility for a quality output of God’s Word. God has anointed leaders all over the world who already know their own language and are more than capable of ensuring the quality of any translation.
So, where can we go from here?
- First, we never compromise the work flow process to ensure quality.
- Secondly, we only work with trusted men and women who have been given the abilities to work in this area.
- Thirdly, we make the technology available on a massive scale at a cost that assures that anyone who meets the first two steps above can participate.
- Fourthly, when a new version is completed, it is made available freely to the whole Body of Christ in that language so everyone, of every class, kind and condition in that language can access the Word of God.
Let me speak briefly as to the third step of making the technology available. Many groups have over the years, decades and centuries been involved in this process of Bible translation. We have usually implemented whatever technology (printing presses, computers, software, etc.) comes along that helps us accomplish the task without compromising our quality standards. Often, it seems that financial resources have limited how quickly we can adopt new technology or benefit from it.
In an effort to see all the remaining languages have the opportunity to be completed in the next decade, Sovee has intentionally decided to fund the technology, as long as people utilize our Bible translation application.That does not mean we fund the indigenous translators or the editing / checking process. We look at this as a Body of Christ partnership.
The indigenous community provides the manpower and we provide the technology engine. There is no charge to use the Sovee engine for Bible translation. It is not wide open though for anyone to use. We are looking for those trusted men and women who have the capability and understanding to be able to produce a quality translation for a language group. We know who they are because they always have the backing of the Christian leaders from that community and often involve the main Christian leaders themselves.
To have full disclosure, while they are doing the Bible translation they are also “training” the Smart Engine so it can be used to translate other materials with topics like evangelism, church planting, discipleship, stewardship, children’s ministries and so forth. There is a small charge for these additional translations. These charges range from $99 for a month of 30,000 words translated to $1450 to translate a million words.
Even if any group never translates anything else using the Smart Engine, we will not change the free access to the Bible app. Our motivation is not financial compensation, but rather seeing a Bible in another language that has never had it before...all in less than a year.
We simply wanted this information included so people wouldn’t say or feel that there must be some other reasons why Sovee is doing this. The acceleration of the Great Commission is reason enough. God has blessed so many of us in so many ways that all of us should be looking at how we can use our time, treasure and talents in ways that help see lost souls saved for an eternity through a saving relationship with Jesus Christ.
All of us have only so many years on this earth. I hope we can look at all technology and processes through the lens of whether it will help us complete what we understand the Great Commission to be in our lifetime. If it does, we should certainly consider it.