Finally, even in Wikipedia:
Attempts to 'de-Judify' the Bible
"
"Some churches remained led by German Christians until 1945. In 1939 with the approval of 75% of the German Protestant churches the Eisenacher "Institute for Research and the Elimination of Jewish influence on German Church Life" was founded, led by Walter Grundmann. One of its main tasks was to compile a "People's Testament" ("Volkstestaments") in the sense of what Alfred Rosenberg called a "Fifth Gospel", to announce the myth of the "Aryan Jesus". It became clear in 1994 that the Testament's poetic text was written by the famous ballad-poet and proprietor of the Eugen-Diederichs-Verlages , Lulu von Strauß und Torney. Despite broad church support for it (even many Confessing Christians advocated such an approach, in the hope that the disaffiliation of 1937 to 1940 could be curbed), the first edition of the text did not meet with the expected enthusiastic response.
On the original data, I promised some time ago in this blog to continue with some data analysis that would go beyond mere graphical representation of the attrition of Jewish mention in the New translations. Before outlining tests of statistical significance though, there are a few reminders essential to the understanding of what happened to our Bible, including a summary of points made previously:
1. Data collection was via online search engines, intra-site engines, which almost entirely pull up VERSE COUNT vs WORD COUNT. That means that the totals you see graphically represented in this blog/research are total # of verses which mention the term of interest, e.g. The term "Jews" occurs 257 times in 244 verses in the KJV, in the search process of Blueletterbible.com or .org. While some may feel it is an inaccurate representation, since we are consistent in using this method across translations, it essentially 'factors out' and allows a reasonable comparison. Since exact word counts are only available for some translations, and original RV translations rarely have online counts at all, I have chosen this method as a means of obtaining comparative data. Sources include Bible.com, Crosswalk.com [Bible tools section]; Blueletterbible.com/.org; and several others to check accuracy, including the few singular online translation sites with intra-searches. There was some variation among search engines, although rarely by more than 1 or 2, and when this occurred, I took the most frequently represented count.
2. Outliers-Contrary to the new book by that title, for those not used to Statistical jargon, an 'outlier' technically is a data point or value which lies far outside of most data values, and may skew statistical significance and/or 'slope' or 'gradient' of a relationship between variables in any direction which does not represent the 'obvious' relationship. For example, if a class takes an exam and 99.9% of the students make between 78 and 82, and one student makes 100, that student's test score would probably in a one time testing be considered an 'outlier'. If you took the average of tests with the high score the mean would be [just an example based upon the range], 86.7*, but without the '100' would be '80'. Most researchers report the data both ways and let the reader decide what to make of it. This is the convention I have used also---for example, in the term 'HOLY ONE OF ISRAEL', a term used for Messiah,
you may note that the relationship among the data would look very different if it were not for the TEV count---that could skew for example a test of statistical significance to look greater than if that translation had been left out the count. It occurs in the opposite vein also.
3. One and two tailed tests. I cannot give a course in Statistics in this little blog even to scholars in Bible and theology who may not have taken it in college, but briefly, I am reporting both one and two tailed t tests of significance, although as with most 'counts' which make logical sense, the two tailed is more logical. Still, though depending on the question, the one tailed is included for perusal. If you are a theologian, ask someone from your Math dept to give you a quick explanation, or several very efficient ones can be found on the net. I have mostly used Student's T
to compare a number of terms, and occasionally a more conservative measure of Welch's T, and each gives the mean, the 'variation' around the mean, the 'standard deviation' (95% of data occurs within 2 SD of the mean), degrees of freedom which will make sense to your engineer and math friends, and the 'probability' that given the data observed, that the versions differ on the term SIGNIFICANTLY. Please remember that statistics can be used to skew data in many directions, and 'truth' does not rely necessarily on statistical significance though we feel comfortable with it. CRITICAL NOTE: Since the Bible is within its pages declared to be inerrant and perfect, even a small change in raw count can be significant in terms of the veracity of modern texts, and a great change would have to indicate something gone very wrong. And, in short, something did go very very wrong, as the effect of the 'Nazi Church' the Deutsche Christen and its philosophies and scholarship did indeed have an effect on the modern translations as great or greater than the sedition of texts by Westcott and Hort based upon the fallacious finds of the late 18th century antiquities called the Vaticanus, Alexandrinus and the Sinaiticus.
RESULTS FOR SEVERAL JEWISH TERMS
Looking across several translations including the Douay, The KJV, the ASV, The RSV, the NASV, NIV, TEV, Living, Holman, (see previous articles) etc we find the following:
ISRAEL (Pre-Reich vs Post Reich}
Mean: 2348 pre 2084 post
SD: 149.93 275.91post
N: 4 8
Welch's T: two tailed: p=.0608
Confidence Interval: 95% CI
One tailed : .05 STATISTICALLY SIGNIFICANT
(note that the 2 tailed approaches significance and would be at 90%, so it is counted.)
-------------------------------------------------------
CHILDREN OF ISRAEL- across translations, pre and post Shoah
t=4.97
Mean: pre 607 post: 77.1
SD 23.3 208
CI: 95%
p= .001 STATISTICALLY SIGNIFICANT
with outlier removed,n two tailed,one tailed: p=.000281
---------------------------------------------------------------
JEWS
unpaired t test
t=3.28
SE=50.8
df=10
p=.0083
pre:
Mean=314
SD=88.4
range 234-399
..............
post:
mean=212
SD=18.5
range 188-233
95% CI: p=.0083 STATISTICALLY SIGNIFICANT
---------------------------------------------------------
JEWISH
t=-1.96
sd= 33.4
df=10
p=.078 95% CI [prob would fall within range on 90%]
APPROACHING SIGNIFICANCE AT 95% CI
-------------------------------------------------
JEWESS
t=2.18
SD=.935
df=10
p=.054
STATISTICALLY SIGNIFICANT
-----------------------------------------------
ZION
T=.855
SD=34.4
DF=10
P=.412
note: THIS IS NOT SIGNIFICANT, AT 95% CI, BUT NOTE DIFFS IN STANDARD DEVIATIONS:
PRE-SHOAH: 1.26 AND POST: 41.1. TWO TAILED.
------------------------------------------------------
"JEWRY"
T=1.00
DF=3
PAIRED Student's i
p=.391 Note: this is b/c the Douay and Darby are in the pre 1901 and 1st ASV.
-----------------------
TRIBE
t=-.890
sd=31.7
df=10
p=.395
95% CI. (needs a 90% CI run)
_----------------------------------
TRIBES
t=-.798
SD=25.3
DF=10
p=.444
mean : pre 114, sd=4.36 vs post 127, sd=30.1
Note that while the power of this test may not detect at 95% a statistical difference, it does note a great change in deviation around mean (more variation], and may be significant in either a one tailed or 90% CI test.
---------------------------------------------------
"DAUGHTER OF JUDAH"
T=1.83
SD=1.34
DF=10
P=.098 AT TWO TAILED 95% BUT: AT ONE TAILED 95% P=.049
APPROACHING OR STATISTICALLY SIGNIFICANT
-----------------------------------------------
MORE TERMS WILL FOLLOW BUT PART OF DATA WAS STOLEN
____________
_
This is an invented mean based on the high and low score, not on what would be a number of data values.
0 comments:
Post a Comment