One of the four traditional transcriptional canons - that is, rules for determining the correct reading, based on the ways that scribes are supposed to have altered the text - is that we should Prefer the Shorter Reading. In practical terms, this means that we should prefer manuscripts that omit words and verses from our Bibles rather than manuscripts that add them in. This rule is sometimes given in Latin form, Lectio Brevior Potior.
The thinking behind this rule is that scribes would be more likely to add to the text rather than to omit material. It is partly on the basis of this rule that most English Bibles produced in the last hundred years are significantly shorter than the older King James Version. Thus, if you look down at the footnotes in an NIV or ESV, you will often find that these Bibles have left some words (or even whole verses) out of the passage, and instead relegate these deleted words to the footnotes with something like, 'Some manuscripts read ...'
Over the last few decades there have been a number of studies conducted to test whether scribes tended to add to the text or omit material from it. These studies have been based on singular readings and they all show that, contrary to the tradition rule, scribes in fact tended to omit much more material than they added.
Here are the results:
|Studies||Type of Manuscript and Study||Additions||Omissions|
|Royse (1981)||Major Papyri||130 (28%)||337 (72%)|
|Robinson (1982)||Papyri, Majuscules and Minuscules in the Book of Revelation||451 (40%)||678 (60%)|
|Head (1996 and 2004)||Minor Papyri in the Gospels||10 (34%)||19 (66%)|
|Hernandez (2006)||Majuscules in the Book of Revelation||57 (40%)||87 (60%)|
|Royse (2008)||Reworked Update of his study of Major Papyri||127 (29%)||312 (71%)|
|Wilson (2011)||Papyri, Majuscules, Minuscules and Lectionaries in 10% of the entire NT text||1088 (39%)||1712 (61%)|
These results show that scribes tended to omit rather than add to the text, and that the traditional preference for shorter readings (i.e. leaving words and verses out of our Bibles) is without foundation.
The other problem with this rule, of course, is that it flies in the face of common sense: the easiest mistake for a scribe to make would be to omit something as he copied.
Our current critical Greek NT text (UBS4/NA27) consistently prefers shorter readings, because Prefer the Shorter Reading was one of the principles upon which it was edited and also because the editors closely followed the readings of certain manuscripts which consistently omit. UBS4 prefers a shorter reading about 60% of the time, in line with the rule, Prefer the Harder Reading. However, the rule is wrong, and we should prefer the shorter reading only about 40% of the time. Thus, the current critical text is far shorter than it should be.
Despite thirty years of studies showing that this rule is wrong, not only has there been no attempt to rectify this problem, but the recent SBL (2010) Greek New Testament is an even shorter text (as can be seen from the graph below). Likewise, the new NA28 text has made no attempt to correct this systemic error.
The current critical text (NA27/28) needs to be re-edited, and hundreds of places of textual variation need to be reconsidered. Hundreds of readings that have been relegated to the critical apparatus need to be reinstated into the mainline text, particularly in cases of short omissions (one or two words in length, which constitute the vast majority of cases) to reflect the reality that scribes tended to omit rather than add.
We may conclude by quoting James Royse's words in response to reviews of his monograph at SBL Boston in 2009:
James Royse: (It is my) conviction that the preference for the shorter reading is fundamentally mistaken