Like no manuscript

In his debate with James White, Thomas Ross brought a valid and important point about the cobbled together texts in the UBS and Nestle-Aland.

There are mere handfuls of words hundreds of times in the UBS that look like no manuscript on the face of the Earth, while the TR text type looks like many many manuscripts.

Note here that we are simply speaking of individual lines of text—parts of verses consisting of handfuls of words.  As for whole verses, groups of verses, or larger sections of text, the portion of the UBS/NA text that looks like exactly zero manuscripts on the earth grows exponentially. (Note that the Textus Receptus has MS support in 100% of these passages where the UBS/NA text has 0 MS supporting its reading.)

This fact is difficult to discern from the apparatus, in which “readings have been selected and substituted based upon an inadequate representation of the evidence” and where “the readings and their support are often misleading and/or in error.”  There are many “lines of text in the UBS4 and in Westcott and Hort that have no manuscript support” just in the Gospels of Matthew and Mark and well into the triple-digits of such lines—with no MS on earth that are identical to them, as far as we know—in the UBS/NA text (the vast majority of them without any footnote in the LSB warning readers that their text does not replicate any actual MS on earth, and the vast majority of them difficult or impossible to discern from the UBS apparatus).

Reuben Swanson, editor. New Testament Greek Manuscripts: Variant Readings Arranged in Horizontal Lines Against Codex Vaticanus: Volume I, Matthew (Sheffield, England: Sheffield Academic, 1995), iii, xii.

With 29,648 words in the NA text of Matthew and 138,011 words in the NA text of the NT according to Accordance Bible Software, Matthew-Mark is 21.5% of the NT.  With approximately 41 lines in Matthew-Mark with no textual support in even one witness, there would be approximately 191 instances of lines of text with no MS support in the NA NT.