102 Andrew Wilson critical apparatuses used are so inaccurate in their reporting of readings that their value is compromised32. On the contrary, the use of multiple apparatuses is good protection against errors in one. Rather, the problem with critical apparatuses is that they are less helpful in their reporting of corrections. Some apparatuses (like Tischendorf and von Soden) did not appear to report all corrections, and others (like Swanson) were not as accurate in their reporting of corrections as base text. This would appear to be due to the difficulty of reading corrections from photographs33; however, while this problem is real, the correction of omissions and ad- ditions would appear to be the least likely category to suffer from mis- interpretation. Further examination of the actual manuscripts or their original editions would therefore appear to be an area in which the results of this study could be enhanced. Correction Among corrections of singular additions and omissions in the critical apparatuses, the results show that correctors heavily tended to remedy omissions. Thus, among the 431 singular additions or omissions later corrected, there were 133 additions and 298 omissions corrected (31:69). Among the corrections which were themselves singular readings34, there were 40 additions and 27 omissions (a ratio of 60:40). Despite the need for caution in relation to the reporting of corrections in apparatuses, the general conclusion is clear and unlikely to greatly change35: correctors heavily tended to fix omissions. It will not have escaped the notice of textual critics that these results immediately suggest a possible explanation for the expansion of the text 32 Von Soden’s reputation for inaccuracy, for example, is undeserved. See J.R. Royse, “Von Soden’s Accuracy”, JTS 30 (1979) 166-171. 33 Thus, Stephen Carlson, after collating Swanson in Galatians against the original editions of certain manuscripts, writes: “Because corrections can be difficult to discern from photographs, this probably explains why Swanson’s accuracy seems to fall short again mostly when corrections are involved” (http://hypotyposeis.org/weblog/2010/08/errata-for- swansons-collation-of-f-in-galatians.html, accessed 6th September 2011). “As is turning out to be typical for Swanson, his transcription of the base text is very accurate (though not 100%) but gets quite confused when he has to deal with manuscript corrections” (http:// hypotyposeis.org/weblog/2010/08/errata-for-swansons-collation-of-1424-in-galatians.html, accessed 6th September 2011). 34 Not necessarily all corrections of singular readings, although some fell into this cat- egory. 35 After nearly thirty years of extensively checking MSS, Royse’s add/omit results in percentage terms only shifted 1%: 2008 (29:71), 1981 (28:72).