While there is little doubt that Dean John William Burgon was one of the most learned, meticulous, and pious textual critics of all time, the thesis of this review is that Burgon's arguments in Revision Revised, while strong in some areas, are weak in others, and even illogical. Burgon's own canons (or rules) for the practice of textual criticism are clear. 1. Any reading that has consistent ecclesiastical testimony is preferable (xxv). Specifically, this involves the testimony of the manuscripts, translations (or versions), and the early church fathers (xxvii). 2. The earliest reading is preferable (this is not concurrent with the earliest manuscripts necessarily, 339). 3. The largest number of manuscripts, versions, and patristic citations has a better claim to be original (339). Later on, he will directly contradict Westcott and Hort's theory (manuscripts are to be weighed, not numbered) by claiming that the number of manuscripts is weight (455). 4. The widest geographical distribution of the readings in canon 3 has a greater likelihood of authenticity. 5. The reading that can explain all the other readings is also preferable (340). 6. External grounds weigh heavier than internal considerations (96), or what he calls “postulates of the Imagination.” While much of what Burgon affirms here is clearly sensible, there are some caveats that need attention. Concerning the first canon, the question arises of what constitutes consistent ecclesiastical testimony. What if two readings, for instance, enjoy fairly equal numbers of attestation by manuscripts, versions, and fathers? Secondly, the third canon is difficult to maintain consistently. If number is weight, then why does Burgon not once mention how much agreement there is between א and B, on the one hand, and the Textus Receptus (hereafter TR) on the other? This huge agreement does not prevent Burgon's complete rejection of the two manuscripts as two of the most corrupt manuscripts of all time. The fourth canon is not one that affects Burgon's arguments much. He argues far more from majority than from geographical distribution. While canon 5 is one that almost all textual critics agree with, Burgon oftentimes forgets to address alternate possible explanations of how a reading might have arisen. Concerning canon 6, while it is relatively strong to assert that patristic readings ought to be given more weight than they usually are, since they often attest to a reading that predates any extant manuscripts, much more caution is necessary with regard to the patristic evidence. Determining actual citations is notoriously difficult, since some fathers quoted from memory, and some biblical passages are very similar to others, thus casting doubt on which biblical passage is actually cited. Burgon shows little hesitation concerning the church father citations, which shows an over-confidence on his part concerning the evidence. In other words, external evidence is not always as clear as Burgon thinks it is. Burgon's strongest arguments in the book are in reaction to Bishop Ellicott's articles (367-520). This response basically constitutes one long argument for the TR reading of Θεὸς as the correct reading over against the relative pronoun ὃς in 1 Timothy 3:16. These arguments are immensely detailed, and generally convincing. However, here or there one may quibble. For instance, the detailed argument concerning the cross stroke of the Θ and the overstroke indicating abbreviation in Codex Alexandrinus does not take into account the fact that it is equally possible that they were added by a corrector. Later textual criticism (particularly Metzger) refers the relative pronoun reading to the original reading of A, and the cross-stroke to the corrector of A (see NA28). That being said, it is very instructive to note Burgon's comments on how manuscripts can fade over time, and how the evidence then “changes.” It is also difficult to argue with his dictum that all readings are old (245). This fact does not receive enough attention among modern textual critics. Most modern critics uncritically assume that the oldest reading is always (or usually) in the oldest manuscript. However, if a tenth century manuscript is copied from a first generation manuscript, there might be fewer generations of manuscripts in between the later tenth century manuscript and the autograph than between a fourth century manuscript that has more generations between it and the autograph. Of course, it is exceptionally difficult to determine genealogical relationships of any sort among the manuscripts. Even affinity among manuscripts is not proof of common ancestry. This is true because it is not usually evident from what exemplar a manuscript might have been corrected. This allows the possibility of much cross-pollination among the so-called families. This leads to the first major problem of Burgon's book. On the one hand, he rejects the genealogical method in general, a method which he calls “purely arbitrary” (20). His objection is clear: that if no actual steps in the genealogical process are evident, then scholars cannot speak of genealogical evidence (256). However, this sets up a massive tu quoque: he uses the very same genealogical argument with regard to the relationship between א and B (12, 28, 227, 255-257, and 318)! Just because they have similar readings does not prove they have a common ancestor. To add a further problem: Burgon actually grants that “groupings” can be valid (339), though he rejects using such an idea as a foundational principle. It is difficult to avoid the conclusion that Burgon rejects the idea of grouping if it disadvantages the TR, but is perfectly willing to use the argument if it disadvantages א and B. Burgon is therefore overly prejudiced against א and B. The second major problem of the book is that, although Burgon posits the TR as a standard for comparison only, not a standard for excellence (xviii-xix, 75), he does not avoid using the TR as a standard for excellence as well. Deviations from the TR are even labeled “depraved” (xxx). Furthermore, since the TR is assumed to be correct, any minuses from the TR are labeled as “omissions.” This begs the question of whether the material in א and B (for example) is shorter by intention or accident, versus the material in TR possibly being an expansion. This problem is one reason why most modern textual critics prefer the non-value terms “plus” and “minus” instead of “addition” or “omission,” terms which imply an a priori conclusion. Burgon does not hold back from judging the motivations of the scribes of א and B as being intentionally sinister (16, 245). The third major problem with Burgon's book is the illogicality of many of his arguments. For instance, he argues that the reason why א and B have survived is because they were not of average purity (319, 325). In other words, the purer documents were in use and thus wore out, and the less pure documents were unused, thus being preserved. This argument is common among defenders of the TR. It only works, however, if there are no other possible explanations of why such older manuscripts might have been preserved. All the extant manuscripts have been preserved. Is it completely out of the realm of possibility that the reason א and B have survived is because scribes thought it might be good to produce two excellent manuscripts and then put them away so that they might be brought out later to correct other manuscripts? Why is it impossible for God's providence to have worked in this way? Burgon's argument is so bizarre, that, if taken to its logical conclusion, it would cast doubt on the reliability of any ancient manuscript of any book. Burgon's view would imply that the older a manuscript is, the less likely it is to be a good manuscript. Where would one draw the line between a date that ensures it was a well-used copy versus a date that proves it was a poor copy and thus put away? So there are two main problems with this argument. First, it falls prey to the argument of the beard (what date is the cut-off?). Secondly, it has a too-narrow view of the providence of God. All of the extant manuscripts have been preserved by God's providence. That God's providence looks unusual in the case of some manuscripts should not be a strike against them. This leads to the fourth major problem with Burgon's book. It has quite a large number of logical fallacies. Burgon commits the following fallacies: 1. the fallacy of the beard—arguing that manuscripts have differences in every verse does not negate their usefulness (31). Where would he draw the line with regard to how many verses the manuscripts would have to offer consistent testimony on in order to be credible? 2. Begging the question—Burgon's use of statistics in the number of words in Luke's gospel (19,941) is begging the question, assuming the TR count. This is evident also in the way Burgon uses statistics generally, which never manages to point out the agreement of א and B with the TR. He only points out the differences. As a result, he greatly exaggerates the differences. 3. Non sequitur—it does not follow that manuscripts quoting the LXX are therefore self-condemned (56). It does not follow that Scripture could not have recorded something so foul as Herod being sexually aroused by his own daughter, and that this is therefore grounds for rejecting a reading (68). Genesis 19 comes to mind. It does not follow that because Burgon might be right in some instances of analysis, that therefore he is correct in all (107). 4. Poisoned well fallacy—just because there might be a “thicket” of incorrect readings does not make a reading in the midst of them untrue (96). Similarly, the often-used poisoned-well fallacies of the discovery of א in a waste-paper basket, and the long housing of B in the Vatican find their way to Burgon's pen (343). Joseph came out of a prison. Does this mean he was guilty or wrong in something? We do not know the links in the chain of God's providence as to how א got to be in a monastery's waste-paper basket, or how B got to be housed in the Vatican library. It is highly unwise to speculate on such things, let alone cast doubt on the manuscripts' worth by such means. 5. Ad hominem fallacy—the doctrinal position of a textual critic on the deity of Jesus Christ has nothing to do with how well he might be able to do his job as a textual critic (witness the care of Jewish scribes!). Yet Burgon casts doubt on everything Westcott and Hort did simply by attacking their association (and thus committing simultaneously the guilt by association fallacy) with some people who deny the deity of Christ (344). Burgon himself asked for help from Roman Catholic scholars. Should this fact cast doubt on the validity of Burgon's own conclusions? 6. False dichotomy—Burgon commits a false dichotomy in describing א and B, claiming that they are either the very purest or the very foulest of all manuscripts (365). These two manuscripts have far more agreement with the TR than disagreement. How is Burgon so certain that there is no mediating position? Surely every manuscript in existence has a greater or lesser degree of purity. To conclude, Burgon's book has some very strong arguments in certain sections, especially when he is dealing with particular text-critical problems. At the very least, Burgon should be a caution against espousing Westcott and Hort's theories blindly and fanatically. Westcott and Hort were certainly too slavish in their devotion to א and B. It is also certain that Westcott and Hort downplayed the Byzantine manuscripts far too much. The dictum that says manuscripts should be weighed not counted is a false dichotomy. Surely some manuscripts are better than others (though none have any kind of absolute sway). And yet it is also true that numbers count. Furthermore, the genealogical method has many flaws that make conclusions based on it far more tenuous than most proponents of the method would like to believe. It is far too difficult to discern real genealogical relationships when the historical data is so thin (all we really have in most cases is the evidence of the manuscripts themselves), and when agreement and difference can be due to a fair number of different factors. Affinity among manuscripts can be a factor, however, even if it has to be chastened, just as many of the canons of textual criticism require adjustment.