Subscribe to FutureBible (futurebible.ethnolinks.com/feed/rss/)

Existing competitors

A full Bible has huge symbolic power and authority. It symbolizes a person’s membership in a church and signals a certain status in larger society. Therefore, anyone from a minority ethnic group who joins the church is going to want a full Bible. It doesn’t matter what language it is in. And particularly I believe a person will want the Bible that is used by pastors and other church leaders and that is normally the Bible of a large regional or national language.

Print is dead

Bible societies are finding it increasingly difficult to fund the publishing of existing Bibles. New Testaments for small language populations are essentially boutique products in a mass market. The more marginalized the group the more likely that other factors such as illiteracy and lack of financial power are going to contribute toward the perception of a New Testament as a bad investment. Bible publishing is costly and heavily subsidized. And printing Bibles for small ethnic groups is a losing proposition for already cash-strapped Bible agencies.

Orality

The current orality craze is pulling energy and resources away from traditional New Testament projects. Print Bibles are being portrayed as inappropriate for oral cultures. There is much to like about Orality and I support it but it does signal one more reason why minority Bibles are going to be a hard sell in the near future.

Electronic and audio alternatives

At least here in Africa, the rate of mobile phone adoption has been staggering and users continue to stretch the boundaries of what a phone can be used for. Mobile phone Bibles are free (if you can get around the download) and importantly they bring Scripture into a device that is thoroughly literate and social. Audio Scripture is also a huge medium that will take a chunk out of the print Bible market. By the time a New Testament hits the street in print it can potentially have been on audio devices for months or even years.

Conclusion

In summary, I think those wishing to reach a certain saturation point for distribution of print Bibles and New Testaments in minority languages are going to face frustration and ultimately failure. A New Testament dedication has for decades been perceived as the crowning achievement of years of hard work. I expect we need to readjust our expectations and look to audio and electronic versions as achievements worthy of celebration and in fact more effective in reaching the most people possible with the Good News.

Another aspect of this that is interesting is how much Bible reading is done by the average person every week. The most recent Barna report for the US shows relatively low rates of weekly Scripture reading. It’s possible that pastors and other church leaders assume that people read the Bible more often than they actually do. If so, then it might be good to adjust the amount of Scripture read and also the amount of references to Scripture that the speaker assumes will be understood by those listening.

Methodology

I’ve been using a pretty low-tech system. When the service officially starts, I take a picture of the person speaking. Each time there is an “official” Scripture reading, I film it, trying to capture the projection on the screen as well. Then at the end of the service I take a picture when the person up front dismisses everyone. (I’m using a camera that takes pictures and video) It’s not a perfect system but it does allow me to go back and see what was happening at any point in a Scripture reading. Ideally the entire service would be recorded. I might try to do that at least once.

Scripture is used extensively through the services, sometimes in a formal reading. Other times a pastor or speaker might simply reference a verse and then cite it in the middle of a longer speech.

To give you an idea of how this method plays out in real life, here’s an example from yesterday (5 September, 2011)

The service was one hour and twenty minutes long. During that time there were three Scripture readings:

1. A collection of verses about repentance (46 seconds)

2. Psalm 95 read congregationally (80 seconds)

3. Gospel reading from John 20 (96 seconds)

The second and third readings were from the NIV version and the overhead projection matched that although it didn’t always display what was being read in the third reading.

Because the reading was videoed I can check the audio against what is displayed on the screen and also calculate words per minute and errors.

Conclusions

In this 80-minute church service there were approximately 3.7 minutes of Scripture reading (4.6%). That’s an interesting number to consider. I’ve only recorded data for three services so far and the amount of Scripture has been relatively constant although the total length of the service has ranged from less than an hour to more than two hours.

I’m planning to add a component of interviews as well as asking congregates to read Scripture in order to determine their rate of reading and errors.

How much time in your church service is comprised of Scripture reading?

Here’s a graph of the results from a quick test I did that shows reading fluency purely by measuring words per minute.

The three translations tested were CEV, NIV and ESV and for each translation I read the same Gospel passage and an Epistle passage.

Here are the numbers, if you prefer to look at a spreadsheet:

As we can see from the graph and spreadsheet, the numbers are pretty consistent across the three translations. The noticeable difference is between the genres of Gospel and Epistle. Words per minute was considerably higher across all translations for Gospel than for Epistle.

This was only for one reader (me!) and a much larger sample would be needed to see if these trends are consistent.

Before I close, I want to tie this to two other studies that are currently being done. Tim Bulkeley is trying to look at the orality and the KJV/AV Bibles:

My thought is to address the well-known aural/oral qualities of the KJV/AV and relate that to the possibilities of various oralities/new oralities introduced by the move to electronically mediated communications.

I have sometimes noticed that I have more trouble reading easier translations like the CEV because of my long familiarity with the NIV. So even though the language of an idiomatic translation might be closer to the diction of modern speech in English, I’m actually a faster reader when using the NIV since it is the translation I have used most over the last thirty years. Our family only started using the CEV in 1999 and despite our love for that translation, my wife and I still have a lot of NIV in our heads (and NASB and NKJV for that matter).

This phenomenon is explained by the concept of “chunking” (Yamashita and Ichikawa, 2010) which allows fluent readers to “fill in the blanks” while not having to read every word. This is easier with a well-known translation, but possibly less so with a unfamiliar translation even if it more closely follows natural English syntax.

The second study I heard about at the Pan African Reading for All conference in Gaborone earlier this month. Agatha van Ginkel presented on her study of fluency of L1 Sabaot and L2 English speakers. The study was interesting because the results were counterintuitive. Despite the readers being L2 speakers of English, they were faster readers of English than their native Sabaot. One possible explanation is that Sabaot, a Nilotoc language has really long words. So your word per minute count is necessarily going to be lower. Another explanation proposed by van Ginkel is that Sabaot is agglutinative and has challenging features like vowel harmony that mean that distinct occurrences of any particular word form are likely to be very low.

I’m hoping to apply van Ginkel’s study to a research project at our international church here in South Africa. Her study shows that it’s not easy comparing fluency rates between languages, and it’s further complicated when dealing with a very familiar text like the Bible. My particular interest is in how the use of overhead projections of the Scripture help or hinder the comprehension of Scripture readings during the liturgy. If the reader from the pulpit is reading at a rate faster than that of the listeners, do they follow the reader or rely on the written text being projected? And what happens when the reader and the projection don’t match, that is, for example NIV is being read but ESV is being displayed. Another factor to look at would be whether worshipers are using their own Bible, possibly in another language to scaffold the reading experience.

Our meeting was held at the Lovane Boutique Winery in beautiful Stellenbosch

Yesterday I attended an interesting meeting held in one of the most beautiful places I’ve ever seen. The region between Stellenbosch and the ocean is scenic in the extreme with rolling hills covered in autumn vineyards and craggy mountains on every side. And the weather was gorgeous and sunny.

At the meeting were representatives of Global Recordings, Bybel-Media, MegaVoice and Davar Partners. I was representing Wycliffe SA Regional Translation Services. These organizations play vital roles in supplying Scripture materials in audio format throughout Southern Africa. Activities range from securing licenses to distribute recordings, to archiving of audio, to distribution and maintenance or gadgets for people to listen to the Bible. Those in attendance were vastly knowledgeable about the task and also visionaries looking ahead to what the future might hold for Scripture distribution in this region.

There are huge challenges in setting up and maintaining supply lines for getting materials from resource-rich South Africa into the surrounding countries. I’m optimistic about the chances of these organizations succeeding in long-term sustainable distribution of Scripture in non-print formats. Their openness and collaboration is admirable but so is their expertise. In fact, I believe we’re on the verge of a revolution in distribution and resourcing for African languages. It’s an exciting time to be working in this part of Africa.

Note: See my post at Kanyimbe for more links to the study.

Below is a graph I created based on the data just for teens for the section on “What You Can’t Live Without”:

We could spend a lot of time deconstructing this data. For example, “parents” rate lower than “cell phone.” But the top score goes to “Family.” So there’s some fuzziness in these results. But if we broadly look at what answers make up the top ten there are some interesting results:

• Family
• Cell phone
• Money
• Parents
• MXit
• Friends
• School
• Music
• Bible
• Love

I find the “Bible” answer to be a surprise. It ranks higher than “Love” even! Those who distribute Bibles in South Africa should consider following up this survey and trying to find out why “Bible” ranked so highly and in what way exactly do teens feel they “can’t live without it.”

Another thing to consider is how Bibles might be distributed using some of the other items on this list. Bible Society of South Africa already distributes the Bible in a format for cell phones. But MXit is far more pervasive than a Java App format and also far more inculcated in the day-to-day experience of South African teens. A Bible “channel” similar to Yoza on MXit if handled correctly could in fact be a big hit for those who are looking for answers to managing the other things on their list, i.e. money, love, and school. A Horoscope contact is already available and might offer a good model to follow.

Some of the types of research questions I’ll be asking are:

1. The Bible tradition of individual members: What various Bibles have they used over the course of their lives.
2. Comprehension in the LWC: Using fluency testing as a relative indicator of comprehension, members will be asked to read Bible passages in various English translations.
3. Personal sacred language: What languages do they sing and pray in when they are not in church.

One thing I won’t be doing is asking comprehension questions of the sermon of the day. There are too many variables to control, such as how awake was the participant? How difficult/unfamiliar was the passage? How prepared/skilled was the preacher.

I’m hoping to do a similar survey later this year in Mozambique.

In this post I’d like to simply collect a list of questions and invite readers to suggest questions as well.

Finally, if you know of someone who would be interesting to hear from on this topic please let me know in the comments or by email.

These questions are meant to be provocative rather than comprehensive.

1. Does the Bible have a future in Africa?
2. Is the Bible’s role in Africa symbolic or pragmatic?
3. How do we resolve the mismatch between oral cultures and a written tradition?
4. You’ve been given the task of distributing Bibles in the Democratic Republic of Congo (215 languages and 58 million people). How would you do it?
5. The urban church in Africa is multilingual, mobile and growing explosively. How does a pastor meet the spiritual needs of a congregation with limited command of the language of wider communication?

OK, that’s a good start. What other questions should we be asking?

The methodology is based on the Oral Reading Fluency test in which readers are given a text and read for one minute then errors are counted and based on that a relative level of fluency can be determined. What I’ve done here is actually used the tool to look not at the speakers but at two different translations.

One would expect that the more formal Almeida would cause more difficulties than the O Livro version which was originally a translation in the style of the English Living Bible.

This is obviously experimental and a small sample. Much larger samples would be needed for conclusive results.

Methodology

The text I chose was Genesis 11:1-9, the story of the Tower of Babel. I removed all verse numbers and formatting and then I broke the texts in half and mixed them. So, David read verses 1-4 in O Livro and 5-9 in Almeida, while Hilary read verses 1-4 in Almeida and 5-9 in O Livro.

The texts were given to each person without time to look at them and then they were asked to read it.

Results for the two tests were:

Hilary: (191 words) 1 min 32 seconds

• Almeida, 3 errors
• O Livro, 3 errors

David (219 words) 1 min 46 seconds

• Almeida, 3 errors
• O Livro, 4  errors

Both of us read at a rate of approximately 2 words per second.

Preliminary results:

Contrary to expectation, David actually made more errors reading the “easy” version.

The types of errors aren’t in focus here but for Bible translators it would be interesting to look at what types of errors were made. I’ll list them below:

Hilary:

1. servirem instead of serviram

2. cume – uncertainty of pronunciation of uncommon word

3. façamo-nos – lengthy pause for unusual grammar

4. limites – accent on wrong syllable

5. ousaram instead of ousarem

6. espalhou-os – hesitation in pronunciation

David:

1. toda a humanidade – stuttered this twice before getting it out

2. planicie – uncertainty as to pronunciation

3. impediria instead of impedirá

4. pela terra – stumble on words

5. começem instead of começam

6. intentaram instead of intentarem

7. uma instead of um a

Conclusions

Contrary to my hypothesis the “easier” translation caused more errors for David than the archaic translation. Hilary made an equal number of errors on both texts. A wider sample of texts and readers would be needed before any definite conclusions could be made about the fluency and comprehension.

In the future, I might try having the reader begin with a short text just to warm up. Neither or us had read Portuguese out loud in many months so our tongues were not very limber!

However, most automated metrics are very prone to giving false results. Wayne Leman showed this with regard to the Flesch-Kincaid reading scale. It is meant to measure text difficulty and grade-level by measuring sentence and word length. However, he shows that analyzing the poem Jabberwocky results in a grade-level of 4.4 (In the U.S. that would be a child around age 10) but the poem is filled with nonsense and largely unintelligible (although very fun to read aloud). Wayne then goes on to discuss the case of Cheyenne. In this North American language, words can often reach up to 50 letters in length! Using the Flesch-Kincaid scale would show just about any Cheyenne text as being very difficult because the very structure of the language is different.

Suzanne McCarthy also looks at reading levels on the same blog. In an email, Suzanne  told me about the Lexile Analyzer which is slightly more sophisticated (it measures word repetition) but still prone to incorrectly representing the difficulty of a text.

At ChristianBook.com, there is a chart of English Bibles and their reading levels. I’m unsure where they got these figures or how they were determined.

Another frequently cited scale was done by Zondervan several years ago but it is no longer available on their website. (Please let me know if you can find it)

Whatever the value of these scales, they do seem to agree with our intuitive idea of how difficult the various translations are (in fact these might just be figures based on someone’s intuition!).

In Africa, where few people speak English as their mother-tongue, church leaders should carefully consider the impact of using a translation with a high reading level. I’ve been very pleased to see the Bible Societies in Africa distributing the Contemporary English Version in bookstores and other retail outlets. Such a translation is going to be helpful for non-native English speakers. The same is true for Portuguese in Lusophone Africa. Mozambican and Angolan Bible societies would do a great service to the populace by distributing an easy to read Portuguese Bible in addition to the default Almeida which is written in very difficult Portuguese.

Two issues that would tend to undermine decisions about choosing an easy to read version in one of the languages of wider communication are: tradition and licensing. The older translations are more familiar and well-loved even while they are often unintelligible. And newer Bibles are often subject to the payment of royalties or other types of licenses which make them prohibitive for Bible agencies who want to publish Bibles but can’t afford to give readers readable options.

The example of the Almeida in Mozambique is particularly sobering where despite the fact that this well-loved translation has been continuously revised and updated, the most commonly available version for sale is the “Corrigida and Fiel” (ACF) which is an extremely old translation based on an outdated manuscript tradition. The only conceivable reason for the continued publication of this version is that it is public domain and can be printed by Bible agencies without paying the royalties required by the very tightly held modern versions of the Almeida. As a footnote, I should add that because the Corrigida e Fiel has been consistently distributed it has become the preferred version of the Almeida. When I surveyed a classroom of Bible students from all over Mozambique, they universally preferred their ACF even when they had been given a free copy of the latest version of the Almeida.

Looking at the urban African church, we should realize that most church services will make use of a language of wider communication which is often not understood well by the members of the church. In these cases, pastors should consider what options might help them to better transmit their message.

Since data projectors are so common these days, pastors can display the text while it is being read. In those cases where a projector isn’t available, printouts of Scripture passages might be effective and also help reinforce the message by giving church attendees something to take home. Pastors can also choose a translation that uses simpler language. And they can consider limiting the amount of text they try to cover in a single sermon or teaching session as well as avoiding jumping around from verse to verse in the Bible.

The multilingual urban church is a reality in Africa. New technologies and increased choices in Bible translations are available to aid pastors in building up their congregations.

A simple test like the Oral Reading Fluency test (ORF) would be an easy way to establish the reading-level of members of a church.

More information:

“Oral Reading Fluency Norms: A Valuable Assessment Tool for Reading Teachers” (citation) (pdf)