Was Timothy God’s ‘fellow-worker’? A Text-Critical and Grammatical Examination of 1 Thessalonians 3.2

October 31, 2004

First Thessalonians 3.2, in the NET Bible, reads as follows: “We sent Timothy, our brother and fellow worker for God in the gospel of Christ, to strengthen you and encourage you about your faith.” Paul is referring to Timothy’s coming to Thessalonica after Paul and Silas had visited there briefly in 50 CE. But there is a textual problem in the text; the reading followed in the NET Bible is a minority reading, found largely in the Western text. Timovqeon, toVn ajdelfoVn hJmw'n kaiV sunergoVn tou' qeou' fits the Granville Sharp rule. This reading is supported by D* 33 it ^d,e,mon* Ambrosiaster Pelagius Ps-Jerome. There are a host of variant readings (UBS lists seven more, though they are looking at a larger chunk of text), the most important of which are: (1) toVn ajdelfoVn hJmw'n kaiV diavkonon tou' qeou' (an Alexandrian reading, supported by A P Y 0278 6 81 629* 1241 1739 1881 2464 pauci lat Coptic Basil), and (2) toVn ajdelfoVn hJmw'n kaiV diavkonon tou' qeou' kaiV sunergoVn hJmw'n (Byzantine reading, supported by D² vg^mss sy^{p, h}**). Although the external evidence for the text reading is not very significant (e.g., only two Greek MSS, one which has been corrected and the other a minuscule!), it does seem best to explain the rise of the other readings. “In order to remove the objectionable character which the bold designation sunergoV" tou' qeou' appeared to have, some copyists deleted the words tou' qeou' (B 1962) or transferred them to qualify tou' eujaggelivou (arm), while others substituted diavkonon for sunergovn … Still later are the conflate readings….” (Metzger, Textual Commentary², 563).

What is of interest here is that the Byzantine reading seems to be a conflation of the Western and Alexandrian. That is, the Byzantine reading combines the two predominant readings found in the Western texttype and the Alexandrian texttype. The architect of the Byzantine texttype could only do this if he knew of the other two readings; this conflation thus implies the lateness and general inferiority of the Byzantine texttype. We can lay these readings out to illustrate this, showing at the same time “the genealogy of descent” of the readings:

(1) toVn ajdelfoVn hJmw'n kaiV sunergoVn tou' qeou' (Western)

(2) toVn ajdelfoVn hJmw'n kaiV diavkonon tou' qeou' (Alexandrian)

(3) toVn ajdelfoVn hJmw'n kaiV diavkonon tou' qeou' kaiV sunergoVn hJmw'n (Byzantine)

It can readily be seen that, given the assumption that sunergoVn tou' qeou' was an objectionable phrase, the key word (sunergovn) would have been replaced by a less problematic term (hence, creating the Alexandrian reading). Later on in the transmissional process, the Byzantine editor(s) desired to preserve every drop of the textual tradition, yet at the same time smooth out the text. Hence, (t)he(y) retained both words (sunergovn, diavkonon), though keeping tou' qeou' only with diavkonon and adding hJmw'n at the end to smooth out the text. Though Hort only produced eight Byzantine conflations, four from Mark and four from Luke, here is another one that could be added to the list.¹

A final comment about the construction, this time grammatical in nature: if the Western reading is what Paul wrote (as I believe it is), this would not necessarily imply that Timothy is here called “a fellow-worker with God.” However, the very fact that it could be taken this way seems to have been the motivation for scribes to alter the text.² That is, tou' qeou' does not have to be taken as a genitive of association. It is quite possible that hJmw'n functions as possessive-associative and qeou' functions simply as a possessive. Thus, “our brother and fellow worker for God.” Timothy is an associate of Paul and Silas, and all three are working for God. But is it legitimate to take a genitive after a sun- prefixed noun as other than associative? What is needed are clear examples of such a usage, especially in Paul. Consider the following:

Rom 11.17—suV…sugkoinwnoV" th'" rJivzh": ‘you are fellow-participants [with Israel] in the root.’ In this text, as is often the case, there is an implied genitive of association while the genitive in the text bears a different relation to the head noun.

Rom 11.34—tiv" gaVr e[gnw nou'n kurivouV h] tiv" suvmboulo" aujtou' ejgevnetoV ‘for who has known the mind of the Lord? or who has become his couselor?’ Although etymologically possible, the usage of suvmboulo" in both classical and Koine Greek meant simply “advisor,” not “fellow advisor.” The gen. must be taken then as objective (“who has counseled God”), the thought being all the more pernicious, for the hypothetical counselor would not be in league with God, but above him.

1 Cor 1.20—pou' suzhththV" tou' aijw'no" touvtouV ‘where is the debater of this age?’ Obviously, the point is not that there is one who debates this age!

1 Cor 9.23—pavnta deV poiw' diaV toV eujaggevlion, i{na sugkoinwnoV" aujtou' gevnwmai: ‘I do all things for the sake of the gospel, in order that I might become a fellow-participant [with others] of it.’ Here again, the true genitive of association is only implied.

Eph 3.6—ei ai taV e[qnh sugklhronovma kaiV suvsswma kaiV summevtoca th'" ejpaggeliva": ‘[that] the gentiles might be fellow-heirs and fellow-body members and fellow-partakers of the gospel.’ Obviously, the gentiles are not fellow-heirs, etc. with the gospel but with Jewish believers, the implied genitive of association.

1 Peter 3.7—OiJ a[ndre" oJmoivw", sunoikou'nte" kataV gnw'sin…ajponevmonte" timhVn wJ" kaiV sugklhronovmoi" cavrito" zwh'": ‘Likewise, you husbands, live with your wives according to knowledge… showing them honor as fellow-heirs of grace.’ Again, the genitive of association is implied (gunaikw'n) while the genitive found in the text (cavrito") is of a different kind.

Cf. also 1 Cor 10.33; 2 Cor 1.24; Eph 4.3; Phil 1.7;³ Phil 3.17 (possibly);⁴ Col 3.14; Smyr. 5.1; Tral. 3.1; Papias 3.15.

In sum, the reading that is most likely the original in 1 Thess 3.2 is Timovqeon, toVn ajdelfoVn hJmw'n kaiV sunergoVn tou' qeou'. This would have been changed by scribes because of its possible offensiveness, if the qeou' were taken as a genitive of association. At the same time, it was shown that qeou' does not need to be taken as an associative genitive in this construction; in many such instances the sun-prefixed noun takes an implicit genitive of association while the genitive that is in collocation with it bears a different function. In collocation with sunergov", qeou'—in Paul’s thought at least—would especially bear other than an associative idea. The force of the phrase thus seems to be “Timothy, our brother and fellow worker for God.”

¹ It should be noted that the minimal number of conflations produced by Hort is no argument on behalf of Byzantine superiority. First, in Hort’s opinion, any Byzantine conflations proved that the Byzantine text, as a texttype, was secondary. Of course, this can only be a relative statement as long as the actual MSS that the Byzantine Vorlage utilized are unknown to us. Nevertheless, that Hort found texttype conflations only in the Byzantine, rather than the Alexandrian or Western texts, argues strongly for the generally secondary nature of the Byzantine text. Second, conflations are difficult to come by. A look at the apparatus of any page in the NA²⁷ text will reveal how difficult it is to produce a conflation. Any textual variants in which one reading is an addition or omission in relation to the other reading cannot be utilitized by a later editor into a conflated reading: adding the wording of one of the readings to the other does not produce a new variant; it simply repeats the longer reading. Likewise, transpositions cannot be turned into conflations. The only kind of textual variant, with few exceptions, that is susceptible to conflation is substitution. Yet even here only select readings can be conflated. How would a scribe conflate iwannh" with iwanh"? Or ihsou" cristo" with cristo" ihsou"? As a concrete illustration of this matter, consider p. 534 of the eighth printing of NA²⁷, which includes 1 Thess 2.19b-3.13.a. There are a dozen textual problems listed here. The substitutions are in 3.1, 3.2 (two different problems), 3.3, 3.8, 3.9 (two problems), and 3.13 (two problems). But the only substitution that could possibly be conflated is the first problem of 3.2, the topic of this paper. In 3.1, dioti substitutes for dio. Obviously, one cannot conflate this into dio dioti or dioti dio! Similarly, in 3.2b peri is a substitute for uper; in 3.3, mhdena siainesqai is a substitute for mhdena sainesqai; in 3.8 sthkhte is a substitute for sthkete; in 3.9a kuriw is a substitute for qew; in 3.9b kuriou is a substitute for qeou; in 3.13a amemptw" is a substitute for amemptou"; in 3.13b dikaiosunh is a substitute for agiwsunh. Only in 3.2a and 3.13b is there any theoretical possibility of conflation. In 3.13b, since dikaiosunh occurs only in A pc, it is doubtful that the reading had an opportunity to present itself to the editor(s) of the Byzantine text. This brief sampling illustrates a couple of important points: First, conflation can normally occur only in textual problems involving substitution, and only in such problems where the substitutes are sufficiently distinguishable to warrant inclusion in a new reading and yet sufficiently similar to not be contradictory, or at least not perceived to be contradictory (e.g., eulogounte" ton qeon/ainounte" ton qeon/ainounte" kai eulogounte" ton qeon Luke 24.53; cf. also Luke 9.10). Second, even such restricted conflations can only occur if the two base readings were sufficiently early and widespread to present themselves as viable options to a later editor. In 3.2a, this second criterion is apparently met; in 3.13b it apparently is not.

² This, of course, raises the question as to why they did not alter the text in 1 Cor 3.9. One or two reasons seem likely: the context there is crystal clear that Paul and Apollos are not in league with God but are his servants (see D. B. Wallace, Greek Grammar beyond the Basics: An Exegetical Syntax of the New Testament [Grand Rapids: Zondervan, 1996] 130 for discussion). Second, there Paul the apostle is called a sunergo" while in 1 Thess 3.2 Timothy, Paul’s assistant, is called a sunergo". Scribal veneration of St. Paul could well explain why 1 Cor 3.9 remains undisturbed textually.

³ There are two genitives with sugkoinwnouv": mou and cavrito". mou is a genitive of association, while cavrito" is an objective genitive.

⁴ summimhtaiv mou could mean ‘fellow imitators of me’ (so NET, RSV, NRSV, NIV, NLT, ESV, KJV, ASV) or ‘fellow imitators with me.’ The fact that very few translations render the genitive as associative shows that translators intuitively sense its inappropriateness here, even though there is no other genitive in the construction and the genitive is personal. Why do they fail to see this in 1 Cor 3.9 or 1 Thess 3.2?