Better Bibles Blog has moved. Read our last post, below, and then
click here if you are not redirected to our new location within 60 seconds.
Please Bookmark our new location and update blogrolls.

Sunday, November 11, 2007

Reading level stats

Suzanne recently posted on Bible version reading levels. Here is a followup guest post by English teacher Tim Carr:
I looked at that chart which the ESV Blog has recently posted. And it has been mentioned on your site also. I'm speaking of Bible Text Stats. I thought I would summarize only some of the info which I find interesting. First the syllables per word category. I was stunned that the KJV, ESV, NASB all had only a 1.3 average. More amazingly the NIrV was in this group. The NLT, NIV, NRSV, HCSB, and NKJ all were in the 1.4 range. ( BTW, I am not sure if the NLT is the 96' or 04' version ).

I'll go from highest to lowest in the next two categories. And I will only reference 8 versions.

This section reflects how many sentences were used.

1) NIrV-- 95,999 --That was no surprise. That version is thick !
2) NLT -- 52,787-- I like that amount of sentences. Not too many and yet not too few.
3)HCSB--46,394-- That's better than the mid to low 30's.
4) NIV --- 39,873-- If it's almost the same as the TNIV, then the number of sentences need to increase.
5) ESV--- 36,457 -- Too few.
6) NKJ -- 34,837 -- Ditto the above.
7) NRSV -34,217 --Same
8) NASB -33,224 -- The rankings of 5-8 are within 3,233 sentences.


1) NIrV -- -873,450 -- That's too many for a person of average or above average reading ability. But, I like some of the wording.
2) HCSB -872,271 -- Is that surprising or what ? It came so close to the range of the NIrV.
3) ESV --804,566 --
4) NKJ --- 803,611 --
5) NASB -784,841 --
6) NLT --782,244 --
7) NRSV -780,660--
8) NIV --741,065 --

I'm not sure how to react to this data. The number of words alone doesn't mirror the reading level. But I like thinner Bibles! Did you notice that the NRSV came in 7th place twice ? We knew that the NIrV would rank as it does.

Words per sentence

1) NASB -- 23.6
2) NKJ --- 23.1
3) NRSV ---22.8
4) ESV -----22.1
5) HCSB----18.8
6) NIV -------18.6
7) NLT -------14.8
8) NIrV --------9.1

I think being in the mid-teen range is ideal. So the NLT gets my vote. Translations 1 through 4 are too wordy. These versions need to reduce the number of words per sentence so texts will read more smoothly.


At Mon Nov 12, 01:25:00 AM, Blogger lingamish said...

Tim, thanks for writing this up. I look forward to reading more of your thoughts on this. These stats are indicative about general reading levels. How do we get past that toward an analysis of comprehension?

At Mon Nov 12, 05:27:00 AM, Blogger John Radcliffe said...

Regarding the HCSB's reported word count, Tim asks: "Is that surprising or what?"

Well I found it so surprising that the HCSB should have 18% more words than the NIV, that I just didn't believe it. So I picked what I thought might be 4 "representative" books, and did a sample comparison. My initial gut feeling was that they should score very similarly, and my figures bear this out. I find it hard to believe that my sample is *so* unrepresentative (and bearing in mind that it does represents about 12% of the total) not to give a reasonable picture of the whole.

[Book of Bible]: [words in HCSB] / [words in NIV] = [HCSB as % of NIV]
Genesis: 34,175 / 35,273 = 97%
Jeremiah: 39,580 / 38,442 = 103%
Mark: 13,955 / 13,823 = 101%
Ephesians: 2,936 / 3,084 = 95%
Totals: 90,646 / 90,622 = 100%

Or as someone is reported to have said: "There are lies, damnable lies, and statistics".

At Mon Nov 12, 10:59:00 AM, Blogger exegete77 said...

My guess is that GW would be in the category of more sentences, with fewer words per sentence. I had seen some statistics not long after it was published, but can't find that document now. I think its reading level is about 5th grade.


At Mon Nov 12, 11:08:00 AM, Blogger exegete77 said...

On the GW site, there is a link to an assessment of reading levels. Search for the link on "Easy Reading Versions of Scripture" by Gail Rice, 1999).

At Mon Nov 12, 02:12:00 PM, Blogger J. K. Gayle said...

Cheers to John Radcliffe for taking time to show us some truth about how there's no monolithic "reading level" for the various people-translators of any 1 published translation. At least, John, you've done the work to show that the summed "level"S of 4 books of 2 translations makes the 2 translations "level," although the 66 books summed together make the "level" of 1 translation 18% less "level" than the other 1. And as clearly the "level"s of any 1 of the 4 books is not consistently less level when the 2 translations are compared. Wow. Yes, 12% of the total is representative of the whole problem here. Moses, Jeremiah, Mark, and Paul should be amazed that their writings have been so "leveled."

Thanks to Tim (and Wayne and Suzanne) for getting us looking at the question of reading level stats.

When so much of the stories, and dialogs, and speeches (and even the telling of the legal stuff) is oral, once upon a time, in the Bible - then to use a single mechanistic literate alphabetic syllabic lexic syntatic measure of "readability" seems quite odd. Never mind that Hebrew orality and Greek orality, and the written versions of these, are very different culturally and linguistically. (Those of us who grew up speaking "monosyllabic" languages - such as 5 and 6 tone Vietnamese - really find it odd to think of "readability" variation. I'd guess everyone writing Chinese would find our English focus on written syllables for readability strange too).

Thanks to Rich Exegete77 for bringing GW back into the discussion. A direct link to Gail Rice's article is here.

At Tue Nov 13, 04:31:00 AM, Blogger John Radcliffe said...

J.K. said … although the 66 books summed together make the "level" of 1 translation 18% less "level" than the other 1.

My point was that I don't believe there *is* an 18% difference between the HCSB and the NIV. I suspect that whoever generated the original stats for the HCSB got it wrong somewhere. (But I'll let someone else do the full recount if they're so inclined.)

Thanks, however, for including the expression "a single mechanistic literate alphabetic syllabic lexic syntatic measure" in a sentence on "readability". It gave me an entirely new slant on the concept, and just goes to show that being able to *read* something (pronounce, or even "understand", the individual words) isn't necessarily the same as being able to *comprehend* its intended meaning!

At Tue Nov 13, 08:04:00 AM, Blogger J. K. Gayle said...

My point was that...

Thanks, however, entirely new slant...just goes to show

? ! :)

touché, John.


(2 easy syllables; 1 readable definition:)


Used to acknowledge a hit in fencing or a successful criticism or an effective point in argument.

[French, from past participle of toucher, to hit or wound in fencing, from Old French touchier, to touch. See touch.]

"touché." The American Heritage® Dictionary of the English Language, Fourth Edition. Houghton Mifflin Company, 2004. 13 Nov. 2007.


Post a Comment

Links to this post:

Create a Link

Subscribe to Post Comments [Atom]

<< Home