Removing a ‘venerable absurdity’* of spelling: Luke 22.24 and 1 Corinthians 11.16

All editions of the Greek New Testament read φιλονεικία ‘love of victory, contentiousness’ (Luke 22:24) and φιλόνεικος ‘loving victory, contentious’ (1 Corinthians 11:16).

All editions, that is, except one: The Greek New Testament Produced at Tyndale House, Cambridge. As we shall see below, these spellings with <ει> are very curious and have become entrenched in successive New Testament editions without foundation in the earliest manuscripts. This is more than a matter of Koiné Greek, like Modern Greek, having fewer vowels than Classical Greek and its alphabet, a situation, which results in phonetic spellings (misspellings, from a Classical point of view, often called ‘itacisms’) with <ι> and <η> and <ει> confused, or <οι> and <υ>, or <ο> and <ω>, or <ε> and <αι> (cf. the currency of both ‘archeology’ and ‘archaeology’, which are pronounced alike and reflect Greek ἀρχαιολογία). As a result of my research on non-Classical spellings in the earliest witnesses to the New Testament text (the papyri and the early uncials) and, in particular, on <ει> written for etymologically-long <ι>, we printed φιλονικία (Luke 22:24), but kept φιλόνεικος at (1 Corinthians 11:16).

We have here an adjective meaning ‘loving victory, contentious’, which is a compound of the stems seen φίλος ‘dear, friendly’ and νί̄κη ‘victory’, and from that compound we have a derivative – an abstract noun – that means ‘love of victory, contentiousness’. The spelling with <ει> may have been influenced by another noun, νεῖκος ‘strife, contention, quarrel’, which is found, for example, at Proverbs 10:12:

μῖσος ἐγείρει νεῖκος,
πάντας δὲ τοὺς μὴ φιλονεικοῦντας καλύπτει φιλία.
‘Hatred stirs up contention, but love covers over all who are not contentious.’

‘Loving victory, contentious’ and ‘loving strife, contentious’ overlap somewhat in meaning. As a further source of confusion, beside νί̄κη there is νῖκος ‘victory’, as in 1 Corinthians 15:54, 55, and 57, for example.

Before turning to the linguistics and philology of these New Testament words and passages, let us review their treatment in earlier editions (with Luke first in each edition, then 1 Corinthians).

1521 Erasmus (via Bibles-online)

1550 Stephanus
1588 Beza

1707 Mill

(all via The Centre for the Study of New Testament Manuscripts)

Then, there’s Tregelles (Luke 22:24, 1 Corinthians 11:16) and Tischendorf (Luke 22:24, 1 Corinthians 11:16), both with φιλονεικία and φιλόνεικος, and a synopsis, via Biblegateway, of Westcott and Hort, Scrivener, and the 2010 SBL Greek New Testament, all with the <ει> spelling in both words. All of the Nestle(-Aland) and UBS editions that I have been able to consult also have φιλονεικία and φιλόνεικος, as does Alexander Souter’s 1916 Oxford edition. The same <ει> spellings are found also in R.G. Tasker’s 1964 Oxford and Cambridge edition of the Greek text translated as The New English Bible of 1961.

It has long been known that the (etymologically) ‘correct’ spelling of the adjective and the abstract noun has <ι>, not <ει>, in contrast to all the printed editions that I have been able to consult. The Greek-English Lexicon of Liddell and Scott, revised and augmented by Jones (LSJ) ‘with the assistance of McKenzie’, explains the situation s.v. φιλόνικος:

(In codd.
the forms φιλόνικος, -νικέω, -νικία and φιλόνεικος, -νεικέω, -νεικία
occur, without any distn. of meaning, e.g. in Isoc. we find περὶ τῶν
καλλίστων ἐφιλονίκησαν
, but τὰς θεὰς περὶ τοῦ κάλλους φιλονεικού-
; μὴ δύσερις ὢν . . , μηδὲ πρὸς πάντας φιλόνικος 1.31; τῆς
πρὸς ἡμᾶς φιλονικίας
, but φιλονεικία in the same sense, 12.158;
φιλόνῑκος is implied by Arist.Rh.1389a12 (where -νεικ-, though
found in good codd., as also in 1363b1, 1368b21, 1370b33, Phgn.
, must be f.l.), καὶ φιλότιμοι μέν εἰσι [οἱ νέοι], μᾶλλον δὲ φιλόνικοι·
ὑπεροχῆς γὰρ ἐπιθυμεῖ ἡ νεότης· ἡ δὲ νίκη ὑπεροχή τις,
cf. Poll.
, AB315; the compd. of φιλο- and νεῖκος would be *φιλονεικής;
the sense contentious arises naturally from fond of victory; in SIG
(v. φιλονικία sub fin. <Magn. Mae., ii B. C., found in Crete>) we have φιλονικίαν Il.12,36
, and φιλονικίᾳ
in OGI335.7 (Pergam., decree of Pitane, ii B. C.); -νῑκ- is also found
in late documents, as POxy.157.1 (vi A. D.).)

The 1959 publication of P.Oxy. XXV 2432 (1st. c. BCE/1st c. CE), which preserves lyric poetry that may be by Simonides, provided further evidence for the <ι> spelling closer in date to the New Testament period. Here’s line 11 from Simonides 541 PMG, ]..ΘΑΛΟΙΤΕ φιλονικίαι:

(LS8 gave φιλόνεικος as its headword, but included a lengthy discussion. I think that it is less likely that ‘loving strife’ could have a ‘good sense’ (see LS8/LSJ s.v. 2), but very likely that ‘loving victory’ could be positive or negative. The amount of revision here between LS8 and LSJ both in the headword and in the disquisition on spelling and word-formation may reflect the influence of Roderick ‘with the assistance of’ McKenzie, one of my heroes.)

Then, there’s Bauer-Danker-Ardnt-Gingrich (BDAG), which gives φιλονεικία and φιλόνεικος as its headwords, but rightly notes s.v. φιλονεικία that ‘φιλονεκία is the “correct spelling”‘ (it is ironic that there is a typo at this point, with -νεκ- not -νικ-; Georgacas’ review, cited by Danker, has a typo too: θιλό- for φιλό-). BDAG cites its authorities as Blass-Debrunner-Funk, Liddell-Scott-Jones, Peter Walters (né Katz), and Georgacas, who should be known, among other reasons, for his work of compiling the errata to the third edition of Carl Darling Buck’s, The Greek Dialects (Chicago: Chicago University Press, 1955). M.W. Holmes, The Greek New Testament SBL Edition (Atlanta, GA: Society of Biblical Literature, 2010), xii-xiii, followed the orthography of BDAG, but did not follow the correction contained s.v. φιλονεικία. BDAG’s headwords, that is, not orthographic comments, were Holmes’ guide.

It is interesting in itself that editions of the New Testament present a non-Classical spelling, just as some present an etymologically-false aspiration in ἐφ’ ἑλπίδι at Romans 8.20 (e.g. Nestle-Aland 28, with P46 01 03* 06* …) in contrast to Romans 4:18 and 5:2, for example, where, as far as it is reported, the evidence is unanimously in favour of ἐπ’ ἐλπίδι.

The linguistic facts concerning the formation of these two words are as follows. Since the noun νεῖκος -ους (< -εος < *-e(s)os) is part of the s-stem system, it would form an s-stem compound adjective in -ης -ες. Then, an abstract noun from that compound adjective would end in -εια. In other words, we should have νεῖκος, φιλονεικής, and φιλονείκεια, just as we have ἀληθής ‘true’ and ἀλήθεια ‘truth’, with λῆθος ‘forgetfulness’ as their base. Compounds in -νεικής have very little currency as far as we know. The most famous example is the personal name Πολυνείκης ‘Mr Much-Strife’, the name in mythology of Antigone’s brother, whose ‘burial’ is the occasion for Sophocles’ Antigone (which happened to be the Cambridge Greek Play, while I was researching non-Classical spellings in the New Testament). LGPN s.v. reports some historical bearers of this name.

(The authority on s-stem word-formation is T. Meissner, who wrote S-Stem Nouns and Adjectives in Greek and Proto-Indo-European (Oxford: Oxford University Press, 2006), one of my lecturers and now colleagues. On the basis of its Index, I cannot find any other -νεικής compound adjectives. C.D. Buck and W. Petersen, A Reverse Index of Greek Nouns and Adjectives (Chicago: Chicago University Press, 1948), 718, lists four such compound adjectives, including πολυνεικής, and a fifth curiosity: late Lesbian-Aeolic dative plural εἰρονείκεσσι (~ *εἰρονεικαισι), cf. its accusative plural εἰρονείκαις; these are ‘misspellings’ of forms of ἱερονί̄κης ου ‘conqueror in the games’ (ἰρονί̄κᾱς ᾱ in Lesbian-Aeolic).

Instead, we have an o-stem, φιλόνεικος -ον, and its expected derivative in -ίᾱ, φιλονεικίᾱ, not one in -ειᾰ as would be expected from an s-stem compound adjective. Note both the spelling and the position of the accent. A compound adjective built on νίκη would have -νῑκος -ον as its second element and would from φιλονικία as a derived abstract noun.

Now to the philology – the study of the texts as preserved in the papyri and manuscripts themselves: the pre-fifth-century witnesses support the ‘correct’ spelling ΦΙΛΟΝΙΚΙΑ at Luke 22:24. Then, the witnesses that have accents present the correspondingly correct accentuation, φιλονικία.

The pre-fifth-century witnesses are:

Papyrus 75 (III c.; and a negative image);

Codex Sinaiticus (IV c.);

and the original text of Codex Vaticanus (IV c.);

(The re-inker of Codex Vaticanus (9th-10th c. CE ?) inserted the ε to give ΦΙΛΟΝεΙΚΙΑ.)

There is also the fifth-century Greek-Coptic diglot known as 029.

Even as late as 779 (XII c.), we find φιλονικία (but we might wonder whether that is an itacism for φιλονεικία):

Fifth-century (and later) witnesses have ΦΙΛΟΝΙΚΕΙΑ. This spelling either reflects a formation from νῖκος, the s-stem counterpart to νί̄κη, or is a ‘mere itacism’, a spelling that reflects loss of distinctions in the Greek vowel system and the emergence of Modern Greek pronunciation:

Codex Alexandrinus (V c.);

Codex Bezae (V c.);

Codex Washingtonianus (IV/V c., but perhaps later still);

and the uncial known as 022 (VI c.).

(Codex Ephraemi Resciptus is deficient at this passage.)

There are no witnesses for ΦΙΛΟΝΕΙΚΕΙΑ, the etymologically-correct derivative of an adjective composed from φίλος and νεῖκος.

Once manuscripts were accented, we find the ‘incorrect’ spelling φιλονεικία, but the correct accentuation (33 ‘the queen of the minuscules’ is deficient at this point):

8 (XI c.);

27 (X c.);

35 (XI c.);

69 (XV c.);

115 (X c.);

478 (X c.);

688 (XII c.);

689 (XIII c.);

690 (XIV c.), which reads φίλον οἱ κεία (! Perhaps read φίλον οἰκεία ~ οἰκία);

757 (XIII c.);

774 (XI c.);CSNTM Image Id: 254669 CSNTM Image Name: GA_774_0251a.JPG

1418 (XII c.);

1424 (IX/X c.);

2278 (XIV c.);

and 2907 (X c.; the line begins Λονεικία, cf. 8 above).

Here, we have the manuscript evidence for the spelling we have seen in print since Erasmus. Again, there is no evidence for φιλονείκεια, the s-stem system’s abstract noun either in accentuation or in spelling.

1 Corinthians 11:16 is more complex in that there is good early evidence for -νεικος (particularly, the original text of Codex Vaticanus, in contrast to its text at Luke 22:24) and, unlike the noun, accentuation cannot be a guide with the adjective. The etymologically-correct spelling has the support of:

Papyrus 46 (III c.);

Codex Sinaiticus (IV c.);

The original text of the Greek of Codex Claromontanus (VI c.; the corrected text includes the insertion of an <ε>);

Codex Ephraemi Rescriptus (V c.; after a fashion: ΦΕΙΛΟΝΙΚΟΣ helpfully begins in the gap between the two columns of overtext).

On the other hand, the spelling with <ει> has the support of:

The original (and re-inked) text of Codex Vaticanus (IV c.);

Codex Alexandrinus (V c.);

and the corrected text of the Greek of Codex Claromontanus (VII c. or later; see above).

Codex Vaticanus, then shows different spellings for these two related words. This manuscript shows a strong tendency (that is etymologically correct to an astounding extent) to write <ει> for etymologically long <ι>. That means that φιλόνεικος is Vaticanus’ way of writing φιλόνῑκος, but also, more importantly, that its use of the spelling φιλονικία at Luke 22:24 is a remarkable exception to its usual practice and has extra weight as evidence for simple iota, not epsilon-iota.

Our practice was to print non-Classical spellings (or, in this case to depart from the spelling that has become traditional in New Testament editions) when ‘2+’ early (5th c. or earlier) witnesses provided evidence for a non-Classical/non-traditional spelling. At Luke 22:24, the early evidence is unanimously in support of ΦΙΛΟΝΙΚ- (not ΦΙΛΟΝΕΙΚ-; the earliest evidence supports the spelling -ΙΑ, as do the minuscules). All the accented evidence supports -ία, not ´-εια. So, we removed what was known, for linguistic reasons, to be a ‘venerable absurdity’ in the tradition of New Testament editions on the basis of the manuscript evidence.

At 1 Corinthians 11:16, we exercised caution in departing from the traditional (but non-Classical and etymologically-incorrect) spelling, φιλόνεικος. The correctly-formed φιλόνικος is found in Papyrus 46 and Codex Sinaiticus as well as in ΦΕΙΛΟΝΙΚΟΣ, the singular spelling of Codex Ephraemi Rescriptus. That is ‘2+ early witnesses’ for -νικ-. A fourth witness would be the sixth-century original text of Codex Claromontanus, but the fifth century was out cut-off point for early witnesses.

However, the counter-evidence of Codices Vaticanus and Alexandrinus was thought to be stronger. Also, although Codex Vaticanus reliably has <ει> for etymologically long <ι> for many, but not all words (ἡμῖν and ὑμῖν are regularly so written, almost never ἡμεῖν and ὑμεῖν (once and twice respectively), spellings found especially in Codex Bezae), Codex Sinaiticus is variable in its accuracy in this regard (among other points of spelling). In particular, Scribe A wrote <ι> for <ει> 29.5 times per folio (the opposite interchange occurs 1.7 times per folio), while Scribe D wrote <ι> for <ει> 4.1 times per folio (and <ει> for <ι> 2.8 times per folio). Scribe A copied both the Luke passage (see below) and the 1 Corinthians passage.

In the Luke passage (see above), we see a certain instance of <ει> for <ι> in ΜΙΖΩΝ for μείζων at the end of the penultimate line. In the second line, Scribe A wrote εἰς ἑαυτούς, which the corrector ‘ca’ altered to ἐν αὐτοῖς, the reading of all other manuscripts (apparently without exception). In the 1 Corinthians passage (see below), we have another certain instance in ΣΥ|ΝΗΘΙΑΝ, for συνήθειαν, which runs from the end of the third line to the middle of the second. (The abstract noun is to συνηθής and ἦθος ‘ethos’ as *φιλονείκεια would be φιλονεικής and νεῖκος.)

On this basis, the evidence for the ‘correct’ φιλόνικος is Papyrus 46 and a slightly garbled reading in Codex Ephraemi Rescriptus. Was ΦΕΙΛΟΝΙΚΟΣ a visual transcription error for ΦΙΛΟΝΕΙΚΟΣ ?

(On the various scribes of Codex Sinaiticus and their itacistic ‘incorrect’ spellings, see Dirk Jongkind, Scribal Habits of Codex Sinaiticus (Piscataway, NK: Gorgias, 2007), 90-94.)

Let us conclude. This investigation into φιλονεικία and φιλόνεικος has absolutely no impact on the translation or interpretation of the two passages (contrast the spelling variants at Matthew 19:24//Mark 10:25//Luke 18:25 and at 2 Peter 2:4). Indeed, one of the reasons, I suspect, that the manuscript evidence for the ‘correct’ spellings has gone unnoticed for so long, even when the linguistic facts were known, is that such variation is not important enough to report in an apparatus criticus or among variae lectiones minores.

However, this investigation demonstrates the high degree of care that was taken in constituting the New Testament text, even at the level of accentuation and of spellings that make no differences to the meaning. We were not content simply to preserve this ‘venerable absurdity’ from previous editions. Instead, we were committed to following the manuscript evidence, even when such consistent rigor results in ‘inconsistency’ in the presentation of words that involve the same, or similar, elements.

(An extra feature of this post is that it serves as something of an overview of Greek palaeography and early printing as well as a linguistic and philological excursus. I shall leave discussion of each manuscript’s features to the Comments).

This blog post is the result of research which I undertook for the Greek New Testament Produced at Tyndale House, Cambridge, in 2016 as a Research Associate in Greek Palaeography, Papyrology, and Philology and on which I presented a paper at the Indo-European Linguistics Seminar, Faculty of Classics, Cambridge 16 November 2016.

* The phrase ‘venerable absurdity’ comes from J.E. Powell, Herodotus: Book VIII (Cambridge: Cambridge University Press, 1939), xviii, who used it to describe editors’ practice of printing texts of Herodotus with rough breathings (e.g. the opening words of Book I: ροδότου λικαρνησσέος στορίης…), even though we know that Herodotus’ dialect was psilotic (h-dropping) and, as such, there were no rough breathings to mark (as compounds, such as ἀπ-ικομένους (not ἀφ-) in Ι 1.1, demonstrate). That said, the ‘venerable absurdity’ distinguishes Ionic ὤρη ‘regard’ (e.g. Hdt. I 4.2) from ὥρη ‘season, period, hour’ (e.g. Hdt. I 32.3) and ὦν (~ οὖν) from ὧν.