[Transliteration note: ö represents omega; ë represents eta.]

In this passage, several textual variants occur; verse 1 contains the most plausible variants.

Verse 1: Instead of ouk autën ("not itself"), the miniscule 1908 and sy^p (the Peshitta) have ouk autön (genitive plural, "not of themselves"), apparently in erroneous conformity to the preceeding two words, mellontön agathön. Miniscule 69 has ou kata ("not according to"), perhaps a mistake in listening.

P⁴⁶ has kai at this point, changing the meaning of the sentence: The law has a shadow of good things to come, and the image of the realities. Thus "image" is parallel to "shadow" instead of realities. Bruce comments on NT usage: "Eikön normally has a more substantial meaning than this [i.e., more positive], and the construction of the sentence indicates that here is it contrasted with skia, not practically synonymous with it."¹ Since skia is emphasized by being first in the sentence, a contrast is implied. The P⁴⁶ scribe apparently made the sentence conform to a Platonic sense of eikon, i.e., similar in meaning to skia. The other possibility, less plausible, is that P⁴⁶ is original and a scribe antecedent to all other copies assumed from the sentence structure that a contrast was necessary and resolved the problem by reversing the meaning of the second clause.² Either way, however, the meaning of the sentence is the same: The law has a shadow of the good realities. Whether the eikon fits on the shadow side of the sentence or on the good realities side does not change the overall meaning.

Some manuscripts, including À and P, insert the word autön after autais thusiais, meaning "their same sacrifices," which fits the sentence but is cumbersome and unnecessary.³

The accusative relative pronoun has (found in P^46c, À , C, D², and most others) is omitted in some manuscripts (P^46*, A) and is dative, hais, in others (D*, H, L). Prospherousin needs an object, and has makes better sense here. The dative form may have been an imitation of the previous word, thusiais, a dative of means. Moffatt reports a speculative emendation of Hort, which has some support in Syriac and Armenian, but it has an awkward nominative absolute, discussed below.⁴ The NA²⁷ choice seems to be most satisfactory.

The more seriously considered variant in this verse occurs near the end of the verse. NA²⁷ follows P⁴⁶, D* and others with dynatai; the important uncials À , A, C, D^b and others have dynantai (plural instead of singular). However, if the plural form is original, then nomos, a nominative form in the first part of the verse, has only a participle, not an active verb. A nominative absolute is not impossible, but it is grammatically awkward and not in the author's polished style. Dynantai may be more easily explained as scribes (who read a few words at a time and may be less attentive to the larger context than other readers are) assumed that the subject must be the more proximate thusiais, i.e., that the sacrifices cannot perfect the worshippers, which makes such good sense that the scribes would not back up to question the meaning.⁵ This would be reinforced by v. 11, (the only other verse in Heb. using the word oudepote) which uses the phrase oudepote dynantai. Again, the meaning of the sentence is the same no matter which variant is original; the chief problem is in trying to discern how the syntax works to convey that meaning.

Verse 2: Instead of ouk an, H* has ou; P⁴⁶ has kan, and some miniscules have only an. Attridge seems correct in explaining the omission of the negative as a failure to recognize the "otherwise" sense of epei,⁶ turning the rhetorical question into a dry comment: "they would have ceased to be offered."⁷

Verse 4: Some versions have tragön kai taurön instead of taurön kai tragön, perhaps in imitation of 9:13.⁸

Verse 5: The Greek manuscripts agree in this verse, but the Massoretic text is significantly different. I will deal with that below.

Verse 6: P⁴⁶ and D have holokautöma, singular, perhaps in conformity with hamartias, or with some LXX manuscripts of Ps. 40 (39), or with thusian and prosphoran in v. 5. The meaning is the same, since they are generic nouns: "The exposition in v. 8 shows that he is quite able to interpret these singulars correctly as generic, paraphrasing them as plurals."⁹

Verse 7: P⁴⁶ and D_*.2 add gar after gegraptai. Gar follows gegraptai in some other NT verses, but it is out of place here.

Verse 8: Some manuscripts (À ², D², I) have singular thusian kai prosphoran, seemingly in conformity to v. 5.

Two manuscripts (D, P) have the article ton in front of nomon. In our author's argument, there is little difference — if something is done legally, it is done according to the law, and vice versa.¹⁰

Verse 9: A few manuscripts (À ² etc.) add ho theos in conformity with v. 7.

Verse 10: Some manuscripts (D² and the Byzantine family) add the article oi in front of dia,¹¹ turning the last phrase into a substantive: we are those being sanctified by the offering... As Attridge says, the "phrasing is extremely awkward and the omission of the article is well attested."¹²

One manuscript (D*) reads haimatos instead of sömatos. This maverick reading may have come from a scribe's theological bias or from the prominence of blood in chapter 9, but was corrected even in D.

In summary, the only serious variants are in verse 1, and there I concur with NA²⁷, as well as noting that these variants affect only the syntax, but not the meaning.