Published online by Cambridge University Press: 24 June 2015
This paper explores the meaning of ἐφ’ ἑλπίδι in Rom 8.20c. This phrase has been variously understood as denoting a hope exercised by the one who subjected creation or a hope inhering in creation despite its subjection. After surveying and evaluating the standard proposals, I argue for an alternative manner of punctuating vv. 19–21 that makes it possible to preserve the most common meaning of ἐπί with the dative, while also taking creation itself as the agent that acts ἐφ’ ἑλπίδι. This proposal obviates a number of difficulties with conventional readings, and highlights the parallels between Paul's statements about the hope of creation in vv. 19–21 and the hope of believers in vv. 24–5.
2 It may be noted at this point that the form ἐφ’ ἑλπίδι (found in P46 א B* D* F G) results from an unusual shift in aspiration whereby the breathing mark on ελπιδι becomes rough and the pi of the preposition aspirates to phi. All other occurrences of the phrase in the LXX and NT have the more usual form ἐπ’ ἐλπίδι, which appears in Rom 8.20 in e.g. P27 A B2 C D2.
4 On the meaning of ἀποκαραδοκία (of which this verse and Phil 1.20 provide the earliest extant examples) as ‘earnest, confident expectation,’ see Denton, D. R., ‘’Aποκαραδοκία,’ ZNW 73 (1982) 138–40Google Scholar. It is illegitimate to appeal to Paul's application of this language of longing to creation as a grounds for insisting that κτίσις here must refer purely to humanity (as does Schlatter, Adolf, Gottes Gerechtigkeit: Ein Kommentar zum Römerbrief (Stuttgart: Calwer, 1959 3) 274)Google Scholar, since the personification of the natural world is a persistent feature of Jewish and early Christian literature (e.g. 1 Chr 16.31; Ps 96.11; 97.1; Jer 4.28; 12.4; 50.46; Joel 1.10; Hab 2.11; Luke 19.40; Rev 12.12; 4 Ezra 10.9; for numerous references in Jewish apocalyptic texts see Hahne, Corruption and Redemption, 97–152).
5 For arguments in favour of the view that κτίσις here includes humanity, see Gibbs, J. G., Creation and Redemption: A Study in Pauline Theology (NovTSup 16; Leiden: Brill, 1971) 40Google Scholar; Käsemann, E., Commentary on Romans (Grand Rapids: Eerdmans, 1980) 232–3Google Scholar; Eastman, ‘Whose Apocalypse’, 273–6; Gaventa, B. R., Our Mother Saint Paul (Louisville: Westminster John Knox, 2007) 53–5Google Scholar; Hultgren, A. J., Paul's Letter to the Romans: A Commentary (Grand Rapids: Eerdmans, 2011) 321Google Scholar. This view is often supported by appeal to Paul's use of the phrase ‘all creation’ (πᾶσα ἡ κτίσις) in 8.22, which is taken as an all-encompassing reference to the totality of God's created works, including humanity. There is, however, evidence in the LXX and NT that constructions in which πᾶς modifies κτίσις ‘can be less than comprehensive, when a particular class of creature is in focus in the context’ (Hahne, Corruption and Redemption, 179). Thus, in Tobit 8.15 πᾶσαι αἱ κτίσεις σου may well mean ‘all your (non-human) creatures’ (coming as it does between references to ‘all your holy ones’ and ‘all your angels’), while in Mark 16.15 πάσῃ τῇ κτίσει seems to mean ‘every human creature’.
6 Thus also Cranfield, C. E. B., A Critical and Exegetical Commentary on the Epistle to the Romans (2 vols.; ICC; Edinburgh: T&T Clark, 1975–9) i.411–12Google Scholar; Wilckens, U., Der Brief an die Römer (3 vols.; EKKNT 6; Neukirchen-Vluyn: Neukirchener, 1978–82) ii.152–3Google Scholar; Dunn, Romans, i.469; Fitzmyer, J. A., Romans (AB 33; New York: Doubleday, 1993) 506Google Scholar; Moo, D. J., The Epistle to the Romans (NICNT; Grand Rapids: Eerdmans, 1996) 514Google Scholar; Adams, E., Constructing the World: A Study in Paul's Cosmological Language (Edinburgh: T&T Clark, 2000) 175–8Google Scholar; Hahne, Corruption and Redemption, 180; Jewett, R., Romans: A Commentary (Hermeneia; Minneapolis: Fortress, 2007) 511Google Scholar.
7 As for the notion that unbelieving humanity constitutes part of the referent of κτίσις in this passage, it is difficult to conceive of unbelievers ‘eagerly awaiting the revelation of the children of God’ in light of the sharp contrast that Paul draws between the children of God and those who live κατὰ σάρκα, whose thinking is controlled by the flesh, who stand in enmity with God, who are unable to submit to the divine law and powerless to please God, who do not have the Spirit of Christ and do not belong to Christ (8.5–9).
8 Even those who argue for an all-inclusive referent for κτίσις often find it difficult to avoid making an implicit distinction between (believing) humanity and the rest of creation. Gibbs, for example, asserts that ‘all of creation is being designated’, but then, confusingly, argues that ‘if “the creation” may be taken to mean the entire creation, then, Paul says . . . that there is a solidarity between man and creation, so that creation is affected by man's action’ (Creation and Redemption, 40; emphasis added), and elsewhere claims that Rom 8 speaks of the hope ‘which is characteristic of both creation and Christians’ (35; emphasis added). With somewhat greater nuance, Eastman says, ‘the ἀπαρχὴ τοῦ πνεύματος impels Christians themselves also to groan with all creation and thus here minimizes their separation from the rest of the cosmos’ (‘Whose Apocalypse’, 274; emphasis added). Such formulations reflect the fact that Paul's language virtually compels the interpreter to speak of believers and (the rest of) creation as distinct subjects.
9 Thus, one may affirm, with Gaventa, that Paul is speaking of ‘we’ as ‘a featured section of the orchestra [of creation], not a different orchestra’ (Our Mother, 54), while still insisting that Paul's very act of ‘featuring’ one section serves to demarcate it to some degree from the ‘non-featured’ sections.
10 Eastman (‘Whose Apocalypse’, 265–6) rightly points out that the membership of the ‘children of God’ who will take part in this apocalyptic revelation need not be identical with that of the ‘children of God’ as currently constituted in 8.14–17, since Paul appears to envision the possibility that the number of God's children will in the meantime be increased, particularly through the inclusion of ‘all Israel’ (Rom 11.26).
11 Thus e.g. Cranfield, Romans, i.412–13; Dunn, Romans, i.470; Moo, Romans, 515; Jackson, T. R., New Creation in Paul's Letters: A Study of the Historical and Social Setting of a Pauline Concept (WUNT 2.272; Tübingen: Mohr Siebeck, 2010) 162Google Scholar.
12 Cf. 1 Thess 3.13; 4.16–17; 1 Cor 15.23; Rom 8.11; Hahne, Corruption and Redemption, 183–4; Hultgren, Romans, 322.
13 Similarly, Gaventa writes that ‘the glory of God's children … is a function of God, not something that comes from them or in any way inheres in them’ (Our Mother, 59).
14 Thus e.g. Gibbs, Creation and Redemption, 43–4; Cranfield, Romans, i.413; Dunn, Romans, i.470; Gieniusz, Romans 8:18–30, 154–60; Adams, Constructing the World, 178; Hahne, Corruption and Redemption, 188; Jewett, Romans, 513; Jackson, New Creation, 157–8; Hultgren, Romans, 323.
15 Cf. also God's authoritative ‘handing over’ of humanity in Rom 1.24, 26, 28 and ‘imprisonment’ of all under disobedience in Rom 11.32.
16 Thus Cranfield, Romans, i.413; Käsemann, Romans, 235. L. J. Braaten objects that ‘Paul viewed the subjection of creation not as a onetime primeval event, but rather as a repeated occurrence’ and that ‘creation is not redeemed from … a primeval curse on nature by God; rather, she is redeemed from the ongoing effects of human sin’ (‘All Creation Groans: Romans 8:22 in Light of the Biblical Sources’, HBT 28 (2006) 136; emphasis originalGoogle Scholar). In light of the emphasis that Paul places on the sweeping cosmic effects of the actions of the ‘one man’ (Rom 5.12), however, it seems better to see creation's subjection as an on-going (rather than repeated) state that nevertheless has a definitive point of origin in Adam's transgression and the resultant divine decree. Moreover, I would argue that the curse placed on the ground in Gen 3.17 is itself one of the ‘effects of human sin’.
17 I thus follow Cranfield (Romans, i.413) in taking ματαιότης to refer to the frustration of creation's ability to fulfil the purposes for which it was designed as a result of human sin. One may compare the use of the verb ματαιόομαι in Rom 1.21, where humanity's thinking is ‘rendered futile’ in the sense that the process by which knowledge of God catalyses honour of God is disrupted. Dunn notes with reference to the two passages that ‘creation has been caught up in the futility of human self-deception’ (The Theology of Paul the Apostle (Grand Rapids: Eerdmans, 2006) 100)Google Scholar. On other possible nuances of ματαιότης, see especially Cranfield, Romans, i.413–14; Gieniusz, Romans 8:18–30, 150–4; Hahne, Corruption and Redemption, 189–91.
18 E.g. Sir 40.8–9; Jub. 3.28; 4.26; Philo, QG 1.32; Josephus, Ant. 1.50; 4 Ezra 7.11–12; Gen. Rab. 12.6. For an extensive survey of the themes of the corruption and redemption of creation in Jewish apocalyptic literature, see Hahne, Corruption and Redemption, 33–168.
19 Byrne, B. labels these words ‘something of a parenthesis, wedged between the main verb … and the following reference to “hope”’ (Romans (SP 6; Collegeville, MN: Liturgical Press, 1996) 260)Google Scholar.
20 Thus Cranfield, Romans, i.414; similarly Lagrange, M.-J.: ‘de façon qu'il n'y ait pas de sa faute’ (Saint Paul: Épitre aux Romains (Paris: Gabalda, 1931 4) 208)Google Scholar; Balz, H. R.: ‘ohne eigenes Zutun’ (Heilsvertrauen und Welterfahrung: Strukturen der paulinischen Eschatologie nach Römer 8,18–39 (BEvT 59; Munich: Kaiser, 1971) 41)Google Scholar; Schlier, H.: ‘von Schuld nicht die Rede sein kann’ (Der Römerbrief (HThKNT 6; Freiburg: Herder, 1977) 261)Google Scholar; Käsemann: ‘creation did not incur guilt for itself as mankind did’ (Romans, 235).
21 Godet, F., Commentary on St. Paul's Epistle to the Romans (2 vols.; Edinburgh: T&T Clark, 1880–1)Google Scholarii.91. Others who take τὸν ὑποτάξαντα to be Adam include Chrysostom, Hom. Rom. 14.5; Lampe, G. W. H., ‘The New Testament Doctrine of Ktisis’, SJT 17 (1964) 458CrossRefGoogle Scholar; Lyonnet, S., ‘Redemptio “Cosmica” Secundum Rom 8.19–23’, VD 44 (1966) 228Google Scholar; Balz, Heilsvertrauen, 41; Schlier, Römerbrief, 261; Byrne, Romans, 258–61. Surprisingly, some have insisted that τὸν ὑποτάξαντα is Adam even while apparently conceding that ὑπετάγη should be taken as a divine passive (e.g. Balz, Heilsvertrauen, 41; Schlier, Römerbrief, 261; Byrne, Romans, 260). Such an interpretative move, however, is exceedingly strained, as it is surely more natural to see Paul's use of the same verb twice in such close succession as pointing towards a single event of subjection, and thus a single subjecting agent.
22 LSJ s.v. ἑκών 1; BDAG s.v. ἑκών; L&N 25.65.
23 Cf. also the cognates ἑκούσιος in Phlm 14 and ἑκουσίως in Heb 10.26; 1 Pet 5.2.
24 Jewett, Romans, 514. Similarly, Gaventa renders οὐκ ἑκοῦσα ‘not freely’ (Our Mother, 54).
25 Of course, the notion that the non-human creation was subjected ‘involuntarily’ is still capable of evoking a secondary contrast with humanity, particularly in light of Rom 1.24, 26, 28, where Paul thrice states that God ‘handed over’ (παρέδωκεν) humanity – who knew God and yet did not honour him (1.21) – in response to its wilful idolatry and rebellion.
26 This distinction is nicely captured in the statement of E. Gräßer: ‘Adam zwar der Grund für das ὑποτάσσεσθαι, nicht aber selbst der ὑποτάξας ist’ (‘Das Seufzen der Kreatur (Röm 8,19–22): Auf der Suche nach einer “biblischen Tierschutzethik”’, in Schöpfung und Neuschöpfung (ed. Baldermann, Ingo et al. ; Neukirchen-Vluyn: Neukirchener, 1990) 107Google Scholar).
27 As a preliminary point, it is necessary to decide whether the conjunction that follows at the beginning of 8.21 should be read as ὅτι (P46 A B C D2 K L P Ψ 33 and numerous minuscules) or διότι (א D* F G). The decision is complicated by the presence of the -δι ending on ἐλπίδι, for if ὅτι was original, a δι- prefix could have been added through dittography, while if διότι was original it could have dropped out through haplography (cf. Metzger, B. M., A Textual Commentary on the Greek New Testament (Stuttgart: Deutsche Bibelgesellschaft, 1994 2) 456Google Scholar; Jewett, Romans, 504). Alternatively, a scribe who interpreted an original ὅτι in a causal sense could have changed it to διότι in order to make the causal nuance more explicit. Ultimately, however, the weight of the external evidence is strongly in favour of ὅτι. While some commentators who accept that ὅτι is the stronger reading nevertheless favour taking the conjunction in a causal sense (e.g. Dunn, Romans, i.471; Hultgren, Romans, 320), it seems preferable to see Paul using ὅτι to introduce the content of the hope, as he does in Phil 1.20 (and perhaps also 2 Cor 1.10, although here the presence of the conjunction is textually uncertain). Thus also Moo, Romans, 516 and n. 45; Byrne, Romans, 261; L&N 90.21.
28 Cranfield, Romans, i.414; Moo, Romans, 516 n. 43; Byrne, Romans, 260–1.
29 Dunn, Romans, i.470 (viz); Fitzmyer, Romans, 508; Jewett, Romans, 514; Hultgren, Romans, 319, 323; RSV.
30 Gieniusz, Romans 8:18–30, 160. Granted, those who see God's own hope as in view are somewhat more likely to understand ἐφ’ ἑλπίδι as modifying the participle, while those who attribute hope to creation are more likely to link ἐφ’ ἑλπίδι with the finite verb, but there is not a strictly necessary correlation between the grammatical decision and the construal of the agent of hoping.
31 I defer until later in the essay consideration of Acts 2.26, where the interpretation of ἐπ’ ἐλπίδι is complicated by the fact that we are dealing with a quotation from the LXX.
32 A preliminary TLG search for ἐπί (τῇ) ἐλπίδι/ἐπί (ταῖς) ἐλπίσιν (including variant spellings) returned only thirty-six instances predating Paul, of which only nineteen occur in non-Jewish texts. In addition to the texts cited below, cf. e.g. Theognis, Eleg. 1.333; Xenophon, Mem. 2.1.18; Lysias, Orat. 6.23.1.
33 Morris, L., The Epistle to the Romans (Grand Rapids: Eerdmans, 1988) 321–2Google Scholar (emphasis added). F. J. Leenhardt's claim that ‘man himself [in contrast to God] could not connect any kind of hope to the subjection of creation’ also appears to reflect this view (The Epistle to the Romans: A Commentary (London: Lutterworth, 1961Google Scholar) 220; emphasis added). For similar statements concerning the relative appropriateness of God, Adam, and/or Satan subjecting creation ‘in hope’, cf. also Gibbs, Creation and Redemption, 44; Adams, Constructing the World, 178; Hahne, Corruption and Redemption, 188; Jewett, Romans, 514 n. 68. It is not always clear in such passages, however, whether the commentator means to imply ‘only God could, acting in hope, subject creation’, or ‘only God could subject creation in(to) a state of hope’.
34 Hultgren, Romans, 323.
35 Hultgren, Romans, 323.
36 E.g. Ps 30.7 (MT 31.6); 36.3 (MT 37.3); 77.7 (MT 78.7); Isa 26.4; Jer 17.7; 2 Macc 2.18; Acts 24.15; 1 Tim 4.10; 5.5; 6.17.
37 This function of ἐπί + the dative is noted by BDF §235.4; BDAG s.v. ἐπί 16; L&N 89.60.
38 Murray, J., The Epistle to the Romans (2 vols.; NICNT; Grand Rapids: Eerdmans, 1959–65)Google Scholari.303 (emphasis added).
39 Gieniusz, Romans 8:18–30, 161 and n. 542.
40 Meyer, H. A. W., Critical and Exegetical Handbook to the Epistle to the Romans (New York: Funk and Wagnalls, 1889) 324Google Scholar.
41 E. Hill argues along these lines that ἐφ’ ἑλπίδι serves as a contrast to οὐκ ἑκοῦσα (taking διὰ τὸν ὑποτάξαντα as a parenthesis explaining the reason for the hope), and thus translates ‘For the creation was subjected to vanity, not willingly (indeed) but (nonetheless) in hope because of him who subjected it’ (‘The Construction of Three Passages from St. Paul’, CBQ 23 (1961) 297Google Scholar; emphasis added). For a critique of this proposal, which requires ἀλλά to signal a less-than-obvious contrast between ‘willingly’ and ‘in hope’, see Hahne, Corruption and Redemption, 189.
42 Barrett, C. K., The Epistle to the Romans (BNTC; Peabody, MA: Hendrickson, 1991 2) 156Google Scholar.
43 Jerusalem Bible.
44 Byrne, Romans, 258.
45 Hahne, Corruption and Redemption, 192.
46 Jewett, Romans, 514.
47 Cranfield, Romans, i.414.
48 Käsemann, Römer, 219.
49 Nebe, G., ‘Hoffnung’ bei Paulus: Elpis und ihre Synonyme im Zusammenhang der Eschatologie (Göttingen: Vandenhoeck & Ruprecht, 1983) 87Google Scholar.
50 Fitzmyer, Romans, 508 (emphasis added).
51 Moo, Romans, 516 and n. 45 (emphasis added).
52 Jewett, Romans, 514 n. 68 (emphasis added).
53 Jewett, Romans, 514 (emphasis added).
54 Ἐπ’ ἐλπίδι is characteristically used in the LXX to translate Hebrew phrases containing forms of the root בטח (cf., in addition to Ps 15.9, LXX Judg 18.7, 10, 27; Ps 4.9; Prov 1.33; Hos 2.20; Zeph 2.15). In context, these Hebrew phrases carry the meaning ‘in a state of security’ or ‘at ease’, and thus none of these passages appear to have in view the sort of expectant, forward-looking hope of which Paul writes in Romans (which, in Hebrew, is typically conveyed by קוה). Cf. the discussions of R. Bultmann, “ἐλπὶς, κτλ”, TDNT ii.521–3; A. Jepsen, ‘בטח,’ TDOT ii.88–94. The only other occurrences of ἐπ’ ἐλπίδι in the LXX are at Isa 28.10, 13, where ἐλπίδα ἐπ’ ἐλπίδι and ἐλπὶς ἐπ’ ἐλπίδι are used to translate the Hebrew קו לקו, which is rendered ‘line upon line’ by the NRSV, apparently as a result of the assumption that קו, ‘line’, was a substantive derived from קוה, ‘to hope’.
55 Ἐλπίς/ἐλπίζω is similarly linked with ἀπεκδέχομαι at Gal 5.5; Rom 8.25; and with ἀποκαραδοκία at Phil 1.20.
56 On this point, see Stirewalt, M. L. Jr, Paul the Letter Writer (Grand Rapids: Eerdmans, 2003) 20–3Google Scholar. For examples of Pauline parentheses, see Rom 1.13; 7.1; 1 Cor 1.16; 9.20–1; Gal 2.2, 6, 8; 1 Thess 4.1.
57 For possible linkages between Paul's use of στενάζω and images of lamentation and mourning in the Hebrew prophets, see Braaten, ‘All Creation Groans’, 131–59.
58 Gempf, C., ‘The Imagery of Birth Pangs in the New Testament’, TynBul 45 (1994) 123–4Google Scholar; Gaventa, Our Mother, 53, 56–9.
59 Thus, the content of creation's hope in the midst of its labour ‘is not “this pain will produce a future good”, but rather “the present agony will not always be with us”’ (Gempf, ‘Birth Pangs’, 124).
60 While Gaventa (Our Mother, 58) questions the validity of translating the singular σώματος as ‘bodies’, N. Turner notes that the use of the distributive singular of σῶμα with plural possessive pronouns for the plural ‘bodies’ is a regular NT idiom (A Grammar of New Testament Greek, vol. iii: Syntax; ed. Moulton, James Hope; Edinburgh: T&T Clark, 1963) 23–4Google Scholar). Cf. Matt 6.25; Luke 12.22; 1 Cor 6.19, 20; 2 Cor 4.10; Jas 3.3.
61 I thus take τοῦ πνεῦματος to be an epexegetical genitive.
62 With Cranfield, Romans, i.419; Fitzmyer, Romans, 510.
63 Elsewhere Paul typically refers to salvation as something lying in the future (e.g. Rom 5.9–10; 10.9; 13.11; 1 Cor 3.15; 5.5).
64 Dunn, Romans, i.475 (emphasis added). Cranfield similarly argues that τῇ ἐλπίδι provides a ‘necessary qualification’ indicating that ‘the final effect of God's action, namely, our enjoying salvation, still lies in the future’ (Romans, i.419–20). Cf. also Fitzmyer, Romans, 515; Hultgren, Romans, 325; Turner, Syntax, 241. Moo prefers to interpret it as an associative dative, thereby seeing hope ‘as the ever present companion of this salvation’ (Romans, 522 n. 72).