A colleague and I were looking at Luke 6.4 in the NA27:
[ὡς] εἰσῆλθεν εἰς τὸν οἶκον τοῦ θεοῦ καὶ τοὺς ἄρτους τῆς προθέσεως λαβὼν ἔφαγεν καὶ ἔδωκεν τοῖς μετʼ αὐτοῦ, οὓς οὐκ ἔξεστιν φαγεῖν εἰ μὴ μόνους τοὺς ἱερεῖς;(Lk 6:4, NA27)
Specifically, the first word ὡς, which has a possible variant of πως. Here's the NA27 apparatus:
[replace] πως 012 L Θ f 1.13 33. 700. 1241. 1424 pc co | – P4 B D syp | txt 01* A C W Ψ M
Nestle, E., Nestle, E., Aland, K., Aland, B., & Universität Münster. Institut für Neutestamentliche Textforschung. (1993, c1979). Novum Testamentum Graece. At head of title: Nestle-Aland. (27. Aufl., rev.) (171). Stuttgart: Deutsche Bibelstiftung.
Reading the apparatus, you really don't know what to think. The variant πως is likeable because it is cleaner and reads easier. But according to one of the so-called "canons" of textual criticism, we're to prefer the harder reading — which is just what the NA27 editors have done. But why?
In this case, we can view some of the original MSS, particularly Sinaiticus (01). As noted in the apparatus, both readings are witnessed in Sinaiticus. The original reading is ὡς, a corrector has revised to πως. So, what does it look like?
(kudos to zhubert.com and csntm.org for lookup and graphic)
So it's fairly easy to tell the π is slipped in by a corrector because the style is different (compare other π in same snippet, fifth letter on top line, fifth letter on second line) and because it is on the margin. So it wasn't in the original pass.
What does this prove? Not much. But the initially attractive variant (and still attractive depending on how you measure it) looks a little less attractive because we can see the nature of the addition. Perhaps a well-meaning scribe also appreciated that πως would be the easier read here and slipped the variant in. Either way, we've confirmed it wasn't done by the original scribe, whatever you may think of his work.
Of course, you should ask other questions at this point because perhaps the variant is a true correction. While the "rule" about preferring the more difficult reading makes sense at times, one has to account for grammatically incompetent or perhaps near-sighted scribes. Maybe even hard-of-hearing scribes.
For instance, the variant under discussion here (ὡς vs. πως) could be the result of a mispronunciation (was that an aspiration or a pi?) from a lector. Or perhaps the previous word ended with a closed syllable (particularly if closed on a labial plosive) and the following aspirated syllable was mis-heard and thus mis-copied. Or perhaps the scribe mis-read the original line for similar reasons when he copied it. But in this case that isn't likely, the previous syllable is open, which we've confirmed by examining the source.
Another reason why apparatuses are helpful, but examining the actual MSS can be more useful.
What did my colleague and I decide? I guess we like sticking with NA27 because of the 'harder reading' argument and also because the original hand of Sinaiticus wrote it that way; along with the confluence of other witnesses of the reading. C'est la vie.
Update (2007-01-09): Note Stephen C. Carlson's (Hypotyposeis) comment. Thanks, Stephen. I hadn't noted the possible harmonization to Mt 12.4 nor the previous variant in Lu 6.3; thanks for supplying the info. Much better info for evaluating the variant. Yet another thing that a bottom-of-the-page apparatus doesn't handle nicely (or at all).