character. This limitation is deliberate, because at this stage of the investigation our purpose must be simply to determine which reading in each case has the best claim to be regarded as original, apart from any preconceived theory as to the weight to be attached to the authorities by which it is attested. Some minds are, no doubt, constitution- ally impatient of this class of considerations, and profoundly sceptical of any conclusions which are based on them. And, no doubt, there would be far less room for difference of opinion, and far less need of patient study and careful and exact scholarship in Textual Criticism, if we could start with some external standard, and go dispense with internal considerations altogether. Nothing, for instance, can be easier, if one may assume that the mass of authorities must always be right, than to prove that a numerically insignificant group of dissentients must be worthless, just be- cause the evidence of ‘the many’ can always, ex hypothesi, be described as ‘ overwhelming.’ But when the precise question at issue is the relative weight to be attached to the rival groups, no amount of erudition can conceal the fact that a demonstration constructed on these lines has no logical value; it does ‘beg the question.’ It is well, therefore, to realize from the outset that the element of personal judgment can never be elimi- nated from the processes of Textual Criticism. A clear realization of this fact is necessary if we are to understand the importance of a careful study of the laws which must regulate the use of the critical faculty, and of the different methods which other workers in the same field have found useful as safeguards to minimize the dangers arising from unconscious caprice or personal idiosynerasy. The criteria for testing the internal evidence of Keadings are of two kinds: Intrinsic Probability, or ‘the consideration of what an author is likely to have written,’ and Transcriptional Probability, or ‘the consideration of what a copyist is likely to have made him seem to have written.’ No doubt, taken separately, they are, as Mr. Miller calls them,* ‘weak pillars.’ But, when they combine in favour of any variant, their testimony is overwhelming. Such cases are indeed comparatively rare. They are numerous enough, however, to enable us to form, first a provisional, and then a more carefully balanced estimate of the characteristic excellences and defects of each authority with which we have to deal. They enable us— that is, in cases where the internal evidence of the readings is ambiguous —to appeal to the internal evidence of the Documents by which the different variants are attested. But even this is not enough. The same document may be of very different value in different parts.
30. We have therefore still to inquire what methods are available when, as in the case of most of the MSS of the NT, whether uncial or cursive, the documents are of a very mixed character, and considerations derived from internal evidence alone are in consequence unusually precarious.
Itis at this point that the real importance of the
rinciple of Genealogy comes full into view. It is ese on the obvious fact that our documents, to quote the words of Dr. Hort to which Mr. Miller has called special attention, ‘are all fragments, usually casual and scattered fragments, of a genealogical tree of transmission, sometimes of vast extent and intricacy.’ It is true, as Mr,
* Trad. Text, p. 238 ; cf. Hort’s Summary, ed. minor, p. 543.
+ The mixed character of the text in the uncials will be ob. vious from the study of any App. Crit. Tf any one wishes to realize the mixed character of the text even in the cursives, he cannot do better than study Mr. Hoskier’s admirably thorough examination of the codex 604. The only surprising thing is that he should imagine that the facts he has observed disturb any of the results at which Dr. Hort arrived. Compare also
the introduction to Scrivener’s collation of 20 MSS with Tregelles’ remarks upon it (Horne, Inti. p. 145).
TEXTUAL CRITICISM (OF NT)
TEXTUAL CRITICISM (OF NT)
bo LS ise)
the value of an argument is measured by its im- mediate effectiveness rather than ‘the single- hearted lover of truth,’ who knows that there is no influence against which he must guard more resolutely than the influence of the popular preju- dices which tend to warp his own judgment, and ely respond most readily to a rhetorical ap- peal.
33. Dr. Salmon suggests in another place (p. 43) that the dogmatic tone of WH’s Introduction is due to the influence of the established Cambridge method of mathematical teaching. A closer parallel would seem to be provided by ordinary text-books in any department of Natural Science. We expect to find in them a description of the methods, and a classified record of the results, of an investigation into a series of phenomena which the student is no doubt expected to take on trust, but only until he has repeated the experiment and verified the result by his own observation. The extraordinary insight and skill in classifica- tion which the Introduction reveals, reflect the expert botanist more than the mathematician.
34, The last consequence of the acceptance of the principle of ‘Genealogy’ to which we wish to call attention, is the light that it throws on the radical unsoundness of any system of Textual Criticism which bases itself directly on a numerical caleu- lation of the attesting documents, before the significance of the numbers has been checked and interpreted by descent. It is, no doubt, a remark- able fact that one of the types of text which were current side by side in the fourth century is repre- sented to-day in extant MSS by a progeny ie the stars of heaven in multitude’;* while the representatives of the others are few and for the most part fragmentary. But the principle of Genealogy reminds us that, however numerous the progeny of any MS may be, their united value can never be higher than that of their common original. And it has yet to be seen whether that common original can, in the case of distine- tively Syrian readings, be traced back beyond the 4th century. The facts which we have already noticed in the history of the text of one of the Pauline Epistles prove that the answer to that question cannot be taken for granted. We must not forget that, if ‘identity of reading’ implies ‘identity of origin,’ identity in a demonstrably wrong reading, except in the case of a primitive error, implies a common original later than the autograph. And in such cases it becomes of primary importance to determine as precisely as possible the date of the common original. 4