At present there is no recent manual eclectic GNT that seeks to present the earliest form of the text alongside variants selected on the basis of likelihood that they were in the original. The UBS editions seek to present variants that are relevant to Bible translators; the NA editions seek 'to provide the reader with a critical appreciation of the whole textual tradition' (NA27 p. 45*). Consequently, major variants about spelling are dropped in favour of semantically important variants that virtually no one thinks are original (or even in favour of semantic information which does not involve variant Greek texts, as in Matthew 6:11).
There is a need for a manual edition that is more focussed on the question of the original text. This would begin to highlight where true uncertainty exists and would allow researchers to focus attention on those parts.
Despite the level of abbreviation that takes place in the apparatus of NA27 it still generally considers variants as whole words. I should like a system that is more willing to pinpoint the variation by letters. Thus, if ° were allowed to apply to single letters within a word,
could signal the variation in John 20:31 most succinctly, provided the ° could be placed in a way that did not interrupt the flow of the text. Nor is there any particular reason why ╭ ╮ should not occur in the middle of a word.