543
date, and because corruption may be rapid in one line of transmission, slow in another. The only adequate criterion of authority for an individual document, apart from its affinity to other documents, is the character of its text, as ascertained by the fullest possible comparison of its different readings; the variations in which internal evidence is of such exceptional clearness as to be provisionally decisive being taken as tests of the general characteristics of the text throughout, and thus shewing how far it is likely to have preserved genuine readings in the more numerous variations in which internal evidence is more or less ambiguous. Criticism resting on this basis, the basis of 'internal evidence of documents' as distinguished from the preceding 'internal evidence of readings', involves not a single but a threefold process; tentative examination of readings, examination of the texts of documents by means of the materials thus collected, and final decision upon readings. It thus makes all variations contribute to the interpretation of each. Its principle may be expressed in the single proposition, Knowledge of documents should precede final judgement upon readings.
The use of 'internal evidence of documents' in the New Testament is however impeded by various exceptions to the homogeneousness of texts, especially by the difficulty of applying it to a plurality of documents in places where the better documents are ranged on different sides, and by the fusion of two or more independent texts in one. This fusion or mixture would arise in several different ways. Sometimes two exemplars would be used together in transcription: sometimes a scribe would consciously or unconsciously intermingle reminiscences of another MS with the text which he was copying: sometimes variant readings noted in the margin of the exemplar, or inserted as corrections of it, would be substituted for the corresponding readings of the exemplar itself. Now, since almost every important document combines readings from more than one ancient source, the nature and therefore ultimately the value of its testimony in any particular case must vary accordingly; and there is no possibility of discriminating the readings derived from the several sources except by observing what the other documents are with which in each case it is associated. When therefore each document is treated as a constant unit of authority, so that the attestation of each reading becomes merely the sum of such units there is no way of arriving at a decision except by resolving the comparison of total authority
