Jamaican Psalms Community-sourced Pilot Project

Extracted text from Jamaican_Psalms_Community-sourced_Pilot_Project.pdf:

The Jamaican Psalms Bible Translation Community-style Gilles Gravelle

Four Bible translation agencies got together and discussed the questions: How can we broaden and accelerate community feedback on a new translation project? How can we produce better quality with greater community acceptance and use? How can the translation spread faster across society? They pondered these questions because the translation process leading to the completion of the Jamaican New Testament had not produced the sort of results the Bible agency had hoped for.1

Traditional translation methods are more office-based. A small team of trained translators spend much time apart from the community working on the translation for the community. They apply their skills as best they can to produce the best translation draft possible before testing it in the community. With this approach, community testing is an occasional step in a linear approach to translations.

Draft

Internal Checks

Community Test

Refine

Consultant Check

Publish and Distribute

When people from the community know the amount of work translators have put into a first draft, they assume major changes are no longer possible. At best, they can provide small suggestions here and there.

The amount of time translators can spend on individual and group interviews limits the amount of feedback they can gather. When they get to the community testing stage, traditional review practices can make the interviewees feel like they are indeed being tested. They may think they lack enough Bible knowledge to answer correctly, or they think they lack theological perspective. Gathering feedback from a broad spectrum of society is daunting, if doable at all. A better way to gather community feedback was needed.

This project sought to build early and ongoing broad feedback from a diverse community of Jamaican Creole speakers. To accomplish this, the translators would spend more time in the community than perhaps in the translation office. A custom-designed website would allow more people to follow translation progress, ask questions and post comments. Participants could also engage in on-line discussions with each other and the translators. They could freely invite friends to participate in the translation work through web-based discussions.

This new translation strategy was an important step because Jamaican Creole has been primarily a spoken language. Indeed, colonial influence left a negative imprint on Jamaican

Creole as an illegitimate spoken or written language, and that impression remains with many Jamaicans to this day. People are not used to reading the Jamaican New Testament or anything written in Jamaican. Many find it challenging. Yet 95% of Jamaicans speak Jamaican Creole, so with a community approach, oral processing and listening would benefit the translation work. The book of Psalms was chosen because of its importance in Jamaican society. The genres oral/poetic style fits well because it was already very Jamaican.

Consumer or Co-producer? Treating end-users as potential translation consumers has not generated a sense of pride and ownership of a translation in the past. In too many cases this has resulted in low use. Low use also means slow spreading. People dont usually promote what they dont use. Talking about group diversity, James Surowiecki (2005) identifies a problem that seems to describe Bible translation. That is, a small team of experts are good at doing only a small set of things well. The lack of information diversity is limiting. Research shows that aggregating more information diversity through crowdsourcing coupled with the work of experts produces better results than the experts working alone.2

Lately, sociologists and phycologists say open innovation rather than closed innovation is needed. This means work can be done better through group collaboration rather than hierarchy. This also suggests the end-user of a translation project should also be an active participant rather than a passive consumer.3

Could broad community discussions of the Psalms, both off-line and on-line, help the translators produce better first drafts? It probably would, but this notion still has a consumer-oriented feel to it. So an additional thought was proposed. Could diverse pre- draft community input groups help the translators produced a better first-draft? And would the participants view this first draft as the fruit of their efforts, too, creating a sense of pride and ownership? Finally, would subsequent face-to-face meetings with a wide variety of people generate growth in the number of people providing new and ongoing feedback on the Jamaican Psalms Website, accelerating acceptance and use by more people? This approach would work more like this:

Input Improvement Feedback

Feedback

Translator's Translation Draft

Improvement

Input

Input

Improvement

Feedback

Community Drafting

To produce pre-draft input on a Psalm, community groups were subdivided into smaller groups, each assigned a different portion of the Psalm. They were also presented with translation notes.

Each sub-group discussed the Psalm among themselves. They were also assigned verses to draft.

Next, the groups came together and combined their Psalm verses with the other groups verses. They read their translations out loud and then the full group discussed what they had produced. In a sense, the pre-draft group produced a rough draft for group discussion with the translators and each other. This process provided important insights for the translators as they began work on the first draft.

After finishing the first draft, the translators posted it on the Jamaican Psalms Website designed for conversations around the Psalms translation. The people who helped produce the first draft could log-in and see the results of their work. They could also post additional comments, question, and recommendations, providing further improvements. This was iterative translation improvement continually applied as the translators came across new, helpful feedback. Perhaps the notion of first and second draft became blurred by this time. It was just ever-improving.

Ongoing Feedback As mentioned earlier, traditional testing utilizes an interview approach in translation testing. It asks who, where, what, and why sorts of questions. These things are important, but because they are not framed within a group conversational setting, it feels more like a test. People want to perform well on tests, so they may try to satisfy the person who is asking the questions. According to a West Indies Bible translation advisor, people in that region see the test questions as condescending, difficult, and it would not be inviting for further input.

It prevented us, the translators, from acting like arbitrators who impose their own opinions on the Jamaican language

Gathering feedback through open group conversations, where people know their opinion matters, can produce more feedback than traditional methods. Significantly, it produces more diversity in that feedback. For example, the chart below shows the sorts of things the face-to-face and on-line groups commented on.

Audio dislike Audio like Reading comprehension difficult

Reading comprehension moderate Reading comprehension easy General impression (dislike) General impression (like) Improved wording, phrase, or sentence Positive comment on something or response to a Negative comment on something or response to Word choice/meaning questioned Spelling disagreement Grammar / style

5% 10% 15% 20% 25%

0%

Information Diversity

Aggregating a lot of input and feedback through diverse communities and on-line conversations provided a more accurate appraisal of the translations strengths and weaknesses, and importantly, its acceptability. Feedback varied according to the Psalm under discussion.

In addition to items listed in the chart, topics included literal versus meaning-based translation and what was considered good Jamaican. For example, some people thought there were direct borrowings from English. They also discussed the Hebrew meaning in the

text and the best way to convey that in Jamaican. Discussions around YHWH/Yahweh or other terms for God, such as Jehovah or Lord, were important. They discussed whether certain varieties of Jamaican, such as Rastafarian Jamaican, should be represented in the translation. Opinions varied.

As one of the translators put it, The community group provides diversity rather than large numbers of the same kinds of people. I thought of it as You could see how different peoples gifts functioned.

something useful; now A particular seminary instructor was not good at drafting, but I think of it as good at hermeneutics. Some people were more poetic, and some necessary were better at theology. Others were good at drafting. One seminary professor was good at making the Psalm translation more poetic. The youth groups brought short-term colloquialisms. They said they dont talk like the Jamaican New Testament. They provided a balance with the translation team in the use of phrases. It helped the team explore the use of variations. It prevented, them, the translators, from acting like arbitrators who impose their own opinions on the Jamaican language. Make is Social Including a Website designed for conversations around the translation increased the amount and diversity of feedback. Participation was unlimited. People enlisted to help could invite friends and family to contribute, as well. The idea was to produce more ongoing feedback, rather than only seeking feedback at project intervals.

The project target was for 500 registered users on the Website, but in the end only 212 had registered, and only 101of those were active participants. Project planners assumed individuals would participate on the Website. However, they soon discovered that Jamaicans like to work in groups so church groups used single group logins.

There were about 22 church groups using group logins. Each church had about four groups. Each group had about 10 members. So by the end of the project there were approximately 750 people who likely contributed to the project just from the church groups using the Website. Adding the 101 active individual (non-church) logins means about 851 people likely contributed feedback. However, even the 101 individuals worked with friends at times, too, so the number of individuals involved is likely to be greater than 851. The on- line community produced over 1,362 kinds of feedback in the form of answers, comments, questions, and suggestions for improvement. Participants commonly offered examples of how they would improve something in the translation.

 The Website provided an important platform to enable people from the community, and even some Jamaican speakers living in the U.S and the U.K, to follow progress, contribute feedback, and participate as desired.

Monitoring Crowdsourcing experts claim when it comes to quality control, self-monitoring for accuracy seems to be a natural occurrence with open source and crowdsourcing collaboration. When people work together on something they highly value, they develop a greater sense of ownership.4

There is no indication in the data that an individual or group tried to force a view, control the discussion, or impose a denominational or doctrinal views on the on-line community. Website moderators never needed to censor comments or block contributors. Data reveals the on-line community showed mutual respect, with no apparent hostility exhibited.

Scaling Diverse Feedback One hundred community pre-draft and feedback groups participated, totaling about 1,500 people. Many of these same people produced the 1,362 sorts of comments on the Website. In terms of diversity, groups included men, women, seminary professors, linguistics students, pastors, poets, musicians, church members, young, middle-aged, older, people from varying social classes, within rural and urban locations. Awareness and Acceptance In the beginning, some people argued that community group feedback meetings would not make the translation more acceptable, but less acceptable when they see the problems of translation. However, that was not the case. People appreciated the work of translation and the value of the translation. The community group method naturally involved gatekeepers, and even if they didnt agree they lowered their resistance.

The translation team observed that speed can be achieved if resources are there to support speed. What did speed up with the community-based approach was awareness and acceptance of the translation. With this process, translators could directly address peoples concerns and opinions about the translation. This affects acceptability and support, which was greater with the community pre-draft method. Resistance was reduced because of the pre-draft input meetings.

What did the community think?

Crowdsourcing theory states that people will volunteer their time when asked if they are motivated to do so. About 1,500 people provided pre-draft and first-draft feedback in group settings. Sessions nearly always went over time. People engaged and were eager to share. As one of the translators put it, The quality of discussions was good. The people went deep. They Ordinary Jamaicans have a really engaged in exegetical and hermeneutical chance to participate in Bible issues, and they did this using their best Jamaican Creole. They freely shared their thoughts and translation. Yu no niid no big opinions with vigorous discussions. I like this method digrii!

because it creates more acceptability in the project earlier.

One community group member put it this way, Ordinary Jamaicans have a chance to participate in Bible translation. Yu no niid no big digrii!

Comments by other participants

Translation is not as easy as one would think. There are a lot of things to take into consideration. Everyones viewpoint is also important. It has allowed me to view the Jamaican creole in a more positive light. I had no exposure to Jamaican Creole before visiting this Website; the Psalms and other users' comments were enlightening. It has helped me read Creole a lot better. It truly takes time and commitment. I learned that Bible translation can be very tedious. Also, among other things, I learned about the history of the development of the Jamaican Creole language. Better appreciation for our Jamaican Language.

It has proven to me that the Bible translation into Jamaican Creole can be appreciated and put to good use in the church.

The Psalms make more sense. This is a great way to get input from a lot of people and a variety of people, especially geographic diversity.

I thought it was a needed project. The input of the community before drafting was helpful. It was good. We felt a part of the process.

Conclusions The speed of translation was slower than the office-based approach because the two translators spent a lot of time organizing community groups. They also spent time with all the groups discussing the Psalms chapters they were going to translate or had recently translated. According to the translators, translation quality surpassed that of the Jamaican New Testament, produced using the office-based low community involvement approach.

Given the struggle that Jamaican has had to be a legitimate language of Scripture and Church, the pilot project went a long way in reducing community resistance to its use. Acceptance of a new translation is always challenging, especially for Creole languages. It can take years. However, the community-based approached accelerated acceptance among a wide variety of churches and institutions. Knowledge of the new translation spread faster as well. Certainly, helping the translators to draft and improve the translation instilled a sense of accomplishment and pride. It should, since they were co-producers and not just consumers of something done for them.

Resources and Notes

1 2

Bible Society of the West Indies, United Bible Societies, American Bible Society, Seed Company.

James Surowiecki. 2005. The Wisdom of Crowds. Anchor Books, Kindle Edition, chapter 4, location 1250. 3 Anthony Pym. 2005. The Moving Text: localization, translation, and distribution, p. 47. John Benjamins.

4 Clay Shirky. 2010. Cognitive Surplus; Creativity and Generosity in a Connected Age. The Penguin Press