The so-called Second Letter of Clement has long been regarded as an enigmatic text with an elusive message. This article seeks to elicit previously unforeseen coherence from the document by demonstrating how its content conforms with patron–client relationships at the discourse level. Drawing on tools from cognitive linguistics and socio-scientific reading models, a patron–client frame is constructed through which the document’s content can be organised. The frame consists of a patron ‘slot’, a client ‘slot’ and corresponding characteristics of asymmetry in status, an enduring relationship, and a reciprocal exchange of goods and services. When applied, God and Christ consistently occupy the patron slot, while the addressees fill the client slot along with the accompanying characteristics. As a result of the analysis, a more precise communicative goal emerges as most coherent: to evoke an ongoing reciprocal behavioural response from the audience to the gift of salvation.
Contribution
The article demonstrates that Second Clement achieves coherence when understood in the terms of patronage, and as a result, a reciprocal behavioural response to salvation is presented as the most coherent purpose of the document.
Second Clementdiscourse analysiscoherencecognitive linguisticsApostolic Fathersframe semanticssocio-scientificearly ChristianityFunding information This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.Introduction
At first consideration, it may seem trivial to address such a specific matter as the coherence of one of the more neglected documents in the Apostolic Fathers. However, for nearly as long as comments have been recorded on the so-called Second Letter of Clement, readers have been quick to note its opaque literary nature. Photius, the ninth-century Patriarch of Constantinople, may have been the first to comment on its enigmatic character:
The second letter, containing advice and exhortation to a better life, at the beginning proclaims Christ as God, although certain foreign expressions, from which even the first letter (First Clement) is not altogether free, are introduced as if from Holy Writ. Certain passages are strangely interpreted. The sentiments are somewhat poor and at times inconsistent. (Freese 1920:213; cf. Eusebius, Hist. eccl. 3.38.4; Jerome, Vir. Ill. 15)
Similar remarks emerge in the early modern commentary from J.B. Lightfoot (1890:206–208), who concurs that the author is ‘confused in thought and slipshod in expression’, proceeding to add that, as a literary work, the document is ‘almost worthless’ (cf. Crowe 2015:251). More recent commentators provide little challenge to the perplexed status quo, often dismissing or quickly conceding the document’s lack of coherent thought.1 Perhaps telling is that, in Tuckett’s (2012) magisterial single-volume commentary, he conspicuously omits any discussion of the document’s structure in the commentary’s otherwise extensive introductory treatments.2 Such widespread neglect of the structure, it is surmised here, has inadvertently hindered progress in understanding the document’s message in more precise terms. While the document is often rightly recognised as having general ‘ethical’, ‘hortatory’ or ‘paraenetic’ intentions, a unifying communicative goal has nonetheless remained elusive (e.g. Donfried 1974:24; Jefford 2012:125; Pratscher 2010:74; Starr 2018:251–266; Tuckett 2012:65). To date, the document’s structure, or the way the information is packaged and organised, has largely remained an unexplored path of analysis towards a more precise understanding of the document. If one were to survey commentaries, one would find several, often distinct, proposed structures, but one would also discover that those structures serve almost strict organisational purposes for the main commentary.3
The intent of this article is not to explore all aspects of Second Clement’s structure; such an extensive analysis is reserved for a larger forthcoming project. Instead, discussion will be limited to its ‘coherence’, modestly aiming to demonstrate that its challenging nature is often overstated, and when a particular set of lenses is employed, a more unified conception of the discourse emerges. Specifically, the thesis of the article is that conceptualising the discourse on terms of patronage brings a newfound coherence to the message that, in turn, elucidates the communicative goal of the document. At the outset, it is also worth noting that ‘coherence’, as used here, does not refer to a general sense-making of the discourse. Rather, the term is used in a more technical and linguistic sense to refer to the discourse’s conceptual comprehensibility, as perceived from the addressees’ viewpoint.4 Although appeals to the text or surface structure are inevitable, the term specifically denotes the organisation of the information in the addressee’s mind. Accordingly, the criterion for a coherent discourse is whether ‘for a certain hearing or reading, [the addressees are] able to fit its different elements into a single overall mental representation’ (Dooley & Levinsohn 2001:23). In this respect, the conceptual structure proposed to elicit the most coherence from the discourse is that of the Patron–client frame.5 Patron–client dynamics were first observed in Second Clement by Kelhoffer (2013:433–456) in what may now be considered a seminal article that not only offered novel insight into the relationships therein but also generated newfound consideration of their implications. Varner (2020:41, 53), for instance, in the most recent commentary on Second Clement follows Kelhoffer in these observations and offers a rare claim to a ‘central message’: ‘that believers have an ongoing obligation to render “payback” or “repayment” to Christ or God’. Through utilising the Patron–client frame, these already observed dynamics can be extended to the discourse level, as a result bringing a previously unforeseen level of coherence to the message of reciprocity.
The rest of the article unfolds in two stages. Firstly, the article’s methodology is discussed. In this initial section, the Patron–client frame is established, grounding the construct in semantic frame theory and existing socio-scientific reading models, along with defining its parameters. Secondly, the article proceeds to demonstrate the construct’s heuristic capabilities for the discourse by addressing two questions: Firstly, how is the frame initially activated? Secondly, how are the slots filled by the discourse content?
The patron–client frame
The construction of the patron–client frame operates within the broader enterprise of cognitive linguistics, and primarily, under the framework of frame semantics as originally developed by Fillmore (1977, 1982). The fundamental premise of frame semantics is that words only evoke meaning at the conceptual level through their relationship to other words in a discourse context. In contrast to a dictionary view of linguistic meaning, frame semantics understands words as cognitive tools that provide access to an encyclopaedic knowledge base derived from embodied and thus culturally embedded experiences (Evans 2019:376–379; Evans & Green 2006:207–211; Geeraerts 2021:24; Tenbrink 2020:60). Such a theory accounts for different so-called ‘senses’ not strictly as different lexicon entries but as activations of various aspects of stored encyclopaedic knowledge. In this way, it can be explained, for example, why the English word ‘trunk’ can both refer to the storage compartment of a vehicle and the appendage of an elephant – different frames are prompted based on the relationship of the word ‘trunk’ to surrounding words. ‘She used her trunk to drink water’ and ‘Put those items in her trunk’ activate different conceptual structures that allow the utterance to be understood as intended. In this way, too, semantic frames are critical for successful communication, as they serve to fill in presupposed knowledge gaps between the communicator and the addressee (Ensink & Sauer 2003:5–6). As, until recently, patron–client dynamics have been concealed in Second Clement, those knowledge gaps have proved more like chasms rather than crevices.
One additional step must be taken for the notion of semantic frames to be utilised at the discourse level. While frame semantics was developed to explain linguistic phenomena at lower levels, such as clauses and sentences, an extension to the discourse level is required for the notion to be used here – a move that requires a jump in scale that is neither problematic for human cognition nor novel within the scope of the theory’s application (Chafe 2015:395–402; Ensink & Sauer 2003:6; Longacre & Hwang 2012:24–26). Operational schemata at the discourse level have been utilised under the auspices of various terms, including, for instance, ‘organising frame’, ‘script’, ‘scene’, ‘schema’ and even Vernon Robbins’ ‘rhetorolects’.6 All, however, in keeping with semantic frame theory, rest on the shared understanding that linguistic meaning only actualises in relation to other conceptual structures. Each proposes a kind of independent knowledge structure that serves as a heuristic model or navigating schema into which the content can be mapped. Following, ‘determining what level of frame specificity yields the greatest exegetical returns’, as Von Thaden (2016:301) avers, is the basic task of the analyst.
It is important to underscore, too, that these knowledge structures derive from repeated and patterned embodied experiences, so much so that one develops subconscious expectations for their iterations. For example, when one purchases from a store repeatedly throughout their life, she or he develops a transactional set of expectations. Naturally, one may develop anticipations for a ‘buyer’, a ‘seller’, a ‘good (s)’ and corresponding actions of payment, exchange, so on and so forth (Evans 2019:398; Fillmore 1982:122–123). Concurrently, it is far from implausible to propose that repeated exposure to patron–client dynamics, which one scholar quips as ‘part and parcel’ of the ancient world, could result in similar entrenched and conventionalised patterns of expectations (Van Eck 2016:22). In commenting on the ubiquity of reciprocally based relationships in the ancient world, DeSilva (2022) unintentionally but no less effectively illustrates the way repeated experiences result in entrenched knowledge structures. Worth quoting at length, DeSilva (2022) imagines:
It was imprinted through socially observed and socially practiced behavior by means of which first-century people learned about how relationships work, how the world works, and what values are part of the very foundation of life together. A child notices how his parents interact with people who have helped the family in some way. He accompanies his father out into the public places and notices how others treat his father with honor, and perhaps asks why. As the child grows, he notices dedicatory inscriptions giving public honors to people who have performed some service or underwritten the cost of some public building everywhere in his city and hears proclamations made that declare new honors for emperor or governor or civic patron. When he receives a favor, his family and mentors stimulate him to think about how to respond–and about the importance of responding and so on. This formative process of social education shapes that child’s thinking throughout his or her adult life; he or she will bring this knowledge into any new situation where the social dynamics appear similar. (p. 110)
As entrenched conceptual knowledge structures, frames can be conceived as having ‘slots’ that can then be filled by discourse content. In the case of the Patron–client frame, two so-called ‘slots’ are quite obviously required. There is necessarily the ‘patron’ and the ‘client’ slot, as illustrated in the diagram (Figure 1).
Patron–client frame diagram.
Yet, these ‘slots’ still require accompanying characteristics to effectively identify and map corresponding content, and in this regard, existing socio-scientific reading models offer parameters for understanding such relationships. As many have observed and documented across the span of epigraphical evidence, in contrast to relationships between ‘friends’, an imbalance of power typifies patron–client relationships (Crook 2020:45).7 The higher-status patron provides benefits and favours to the lower-status client, or conversely, the client seeks benefits and favours from the patron on whom they ultimately depend for their often-vital support (Malina & Rohrbaugh 2003:388–389; Moxnes 1991:242). Even among critics of such reading models, it is well recognised that these relationships were maintained through generalised reciprocity (e.g. MacGillivray 2009:37–81, 46). In contrast to a balanced reciprocity between equals or a negative and exploitative reciprocity that strictly benefits the superior giver, appropriate patron–client relationships are characterised by mutual benevolence in the interest of the other.8 Such reciprocity can be considered prototypical, extending well beyond the formalised Roman patronage system and rendering the frame a salient option to be employed towards communicative ends. Saller (2002), on whom Kelhoffer (2013:434–436) also relies in his original analysis of Second Clement, identifies three ‘vital elements’ of personal patronage that can subsequently serve as general characteristics around the slots:
asymmetry in the status of the patron and client
‘the reciprocal exchange of goods and services’
a personal relationship of some duration.
To summarise, it is suggested that patterned experiences of patron–client relationships in the ancient world yielded entrenched conceptual structures that can be individually referred to as the Patron–client frame. The frame, operational at the discourse level in Second Clement, consists of two ‘slots’ or roles: that of the superior patron and the dependent client, accompanied most prototypically by a generalised reciprocity.
How is the frame activated?
Indeed, the most critical question regarding the viability of the frame is the first: how is the frame initially activated? That is, from the viewpoint of the addressees, how do linguistic cues in the text prompt or ‘call up’ the conceptual structure? Without evidence of such activation from the outset, the frame cannot govern the full scope of the discourse. In this regard, a careful reading with attention to patronage dynamics elucidates Clement’s accentuation of the aforementioned elements listed above.9
From the very first sentence, Clement construes a power imbalance and establishes an asymmetry between Jesus and the audience through his presentation of Jesus as a determining judge. ‘Brothers and sisters’, he says, ‘in this way, it is necessary for us to think about Jesus Christ as [we do] of God, as “Judge of the living and the dead”’ (2 Clem 1.1).10 As Kelhoffer (2013:439) highlights, the characterisation of Jesus as a judge and, a few sentences later, ‘as a father’ (1.4) is consistent with a ‘paternal stance’ [patronus] common to patronage. One can observe, too, that a reciprocal element to the relationship quickly emerges in Clement’s warning that when the audience thinks little of him, ‘little also’ they hope to receive (2 Clem 1.2; cf. 2 Clem 8.6, 9.5, 11.6, 15.4). The reciprocal exchange between Christ and the audience becomes even more explicit after a brief exposition of their salvation. Christ has provided the gift of salvation, and now Clement asks, ‘Therefore, what repayment will we give him? Or what fruit is worthy of that which he has given us? And what holy works do we owe him?’ (2 Clem 1.3). Ostensibly, the addressees have incurred repayment [ἀντιμισθία] for the salvation they have received (2 Clem 1.5, 9.7, 11.6, 15.2).
With two characteristics initially detected – namely, an asymmetry in status and overt expectations of reciprocity – the third can be considered: an enduring relationship. Initially latent, this element becomes increasingly explicit in later chapters; nevertheless, its pre-emptive presence can be observed early on. As Clement reinforces their idolatrous and dire condition before salvation, he ascribes Christ’s mercy and compassion as motivations for his salvific activity, which implicates duration in the relationship (2 Clem 1.6–1.7). So too, he proceeds to close the first chapter with the remark that Jesus ‘called us, who did not exist, and he willed out what was not–us’ (2 Clem 1.8; cf. Rm 4:17). Such a construal of vital reliance and dependence on Jesus not only reinforces the imbalance in the relationship but also implies that ongoing activity will be necessary to sustain its continuation. As Clement’s message develops, the enduring relationship becomes increasingly apparent. For instance, at one point, Clement instructs, ‘Therefore, let us give him eternal praise, not from [our] mouth only but also from [our] heart, so that he may welcome us as sons’ (2 Clem 9.10). The ongoing repayment for salvation eventually culminates in a second gift from the giver: deliverance from future judgement (see 2 Clem 17.1–7). This evasion from a final judgement, in this way, squarely rests on the duration of the relationship.
Beyond the presence of the three elements, there is also a concentration of terms semantically associated with patronage throughout the opening chapter. Elliott (2003:152), based on his own and other socio-scientific studies, has tentatively developed what he refers to as a ‘semantic field of terms’ reflective of such dynamics (cf. Crook 2020:55–56). Generally, the category includes lexical symbols that are frequently identified in contexts of patronage, or under this study’s theoretical approach, utilised within semantic frames of the same. Evoking concepts of benefaction, giving and receiving, a concentration of content words within what is here re-termed as a semantic domain of patronage is found. These include two forms of λαμβάνω (1.2, 1.5), three forms of δίδωμι (1.3, 1.5), one form of χαρίζομαι (1.4) and one form of ὀφείλω (1.3). Additionally, although not originally included in Elliot’s admitted ‘illustrative but certainly not exhaustive’ list, the surface forms for ‘reward’ [μισθός] (1.5) and ‘repayment’ [ἀντιμισθία] (1.3, 1.5) also warrant consideration based on their consistent use to express the reciprocal dimension of the relationship (Elliott 2003:152; cf. Kelhoffer 2013:441–446). If ‘reward’ and ‘repayment’ are included, then 10 total surface forms occur in the opening chapter.11 Moreover, not only does the opening chapter contain more than any section thereafter, but of the nine distinct lexical symbols that recur throughout the discourse, five first appear in this fledging section.12
Altogether, sufficient evidence can be discerned in the opening chapter to activate the Patron–client frame. Noteworthy, too, is that the document lacks the typical introductory epistolary features.13 Thus, it is no overstatement to emphasise that these characteristics of patron–client relationships evince from the very outset of the discourse, especially those reciprocal elements present in the anticipated response to salvation. Additionally, and critically, lexical symbols within the semantic domain of patronage provide linguistic cues for prompting conceptions of patronage.
How are the slots filled?
With the frame activated, the second question remains: ‘How are the slots filled?’ Put differently, how does the content map into the heuristic, considering particularly the most salient entities or characters? Initially, the answers seem readily apparent, as Christ has already been discussed as the patron and the addressees as the clients. However, further investigation is warranted to demonstrate the extent to which the content fits with the frame and elicits coherence.
As the superior and resourced giver to whom repayment is explicitly owed, Christ fills the patron slot; however, one could rightly question which role God occupies in the frame. To this question, Clement’s opening presentation of Jesus as a determining judge continues with relevance as the author insists that this too is how God is to be understood (2 Clem 1.1). As a result, an overlapping entity is established in the addressees’ mental representation, and through metaphorical extension into the ‘patron’ slot, they are, to a significant degree, to be regarded as one and the same (see also Kelhoffer 2013:438). As this holds, it should be unsurprising that throughout the discourse, many overlapping attributes and functions are shared between God and Christ. For example, in distinct passages, the audience is commanded to do both Christ’s will (5.1) and God’s will (2 Clem 14.1; see also 6.1). So too, both Christ and God are said to have ‘called’ the audience (1.2, 1.8, 2.7, 9.5, 10.1, 16.1), and it is not only Christ who is explicitly owed repayment but also God (2 Clem 9.7, 15.2).14 It may finally be noted that in at least two instances, both Christ and God are separately implied to have authority over an impending death (2 Clem 5.4, 9.7; see also 7.6). Collectively, then, both Christ and God also exhibit the three vital elements of patronage. Namely, they consistently hold a higher status than the audience, expect reciprocity (i.e. repayment) and anticipate an enduring relationship with the addressees. From the standpoint of coherence, these overlapping characterisations are best explained when organised into the same ‘slot’, and additionally, such overlap explains why a ‘broker’ slot is not necessary, since both Christ and God seem to exhibit features consistent with the patron role (cf. Neyrey 2005:475–476).
When considering the ‘client’ role, by contrast, the content adheres more straightforwardly. The author, for one, assumes the audience consists of those who owe repayment. They are the ones who, in numerous instances, are construed as being dependent on Jesus and God for their original salvation, and deliverance from future judgement rests on their appropriate response (e.g. 2 Clem 1.3, 1.7–8, 2.7, 4.5, 6.7, 9.10, 11.7, 15.5, 17.4). Clement, too, binds himself to the audience’s situation through the consistent use of first-person plural morphological markings and pronouns. It can be observed, for example, that throughout the plethora of exhortations, rather than utilising the more potent imperative forms (e.g. you do this), barring a few exceptions, he consistently utilises first-person plural constructions (e.g. 2 Clem 4.1, 4.3, 5.1, 7.1–3, 8.1, 10.1, 11.1, 12.1, 13.1, 15.3, 15.5, 16.1, 17.1–3).15 He, too, sees himself as one who should respond appropriately to salvation. Such solidarity is confirmed in his closing remarks, where he proclaims himself to be ‘utterly sinful’ [πανθαμαρτωλὸσ], yet eager to flee temptation and pursue righteousness in fear of the coming judgement (2 Clem 18.1; see also 15.1).16 Thus, the audience and author obtain a degree of shared overlapping characterisation comparable to that of God and Christ in the ‘patron’ slot.17 As instantiated in Second Clement, an updated depiction of the patron–client frame with the slots filled (Figure 2) illustrates the result:
Patron–client frame diagram with slots filled in Second Clement.
While the patron slot accounts for Jesus and God, and the client slot for the audience and author, several other entities appear throughout the discourse that could compromise the frame’s explanatory utility. For example, one could question where entities such as the ‘outsiders’ (13.1), ‘unbelievers’ (17.5), ‘presbyters’ (17.3, 5) and the presumable ‘false teachers’ (10.5) fit into this scheme.18 On the one hand, none of the entities requires a slot as they only appear briefly in the discourse and, as a result, are only temporarily activated in the addressees’ mental representation. As Todd (2016) summarises in his approach to discourse coherence:
[F]or a concept to be a key concept contributing to the discourse topics (or in this case, roles), it would either need to be expressed frequently over a stretch of discourse or made salient. (p. 126)
Put differently, since these entities do not transcend stretches of any significant length, they only need to be accounted for at localised levels.19 Nevertheless, once activated, the entities could temporarily fit into either of the two slots of the higher patron–client frame. In the case of the ‘outsiders’ and ‘false teachers’, they inevitably reside outside the patron–client dynamics and, as a result, outside the frame as well. Conversely, since the ‘presbyters’, and the so-called ‘unbelievers’ [ἄπιστοι], exhibit close association with the addressees, they appear to occupy the ‘client’ slot; however, briefly that may be the case (2 Clem 17.3, 5).20 Altogether, it remains that the frame can still account for all the most salient entities spanning the whole discourse. This crucial point advantages the current scheme over more traditional anti-gnostic interpretations, which notably have not gone unchallenged by more recent commentators such as Tuckett (2012:47–57) and Kelhoffer (2017:266–295). In these readings, the key problem occasioning the discourse is fundamentally doctrinal (e.g. Donfried 1974:179). Therefore, if a parallel frame were to be developed to fit such understandings, a discourse-level ‘slot’ for some kind of ‘opponents’ or ‘erroneous doctrine’ would necessarily be required to achieve coherence. However, this is improbable, since, as noted above, these slots could only be temporarily and inconsistently filled by the disparate entities throughout. Nonetheless, interpreters who suppose an anti-gnostic agenda often circumvent, perhaps unconsciously, the need for coherence by claiming the polemic is ‘implicit’ and accentuating presumed gnostic features (e.g. 2 Clem 3.1, 12.1–5) (see e.g. Pratscher 2007:50–55). In some respects, these readings can be said to displace the addressees for the interpreter in the equation of comprehension. In other words, they may cohere for the interpreter, but the linguistic evidence that a similar mental representation would actualise in the minds of the addressees remains exceptionally thin.
Conclusion
Efforts began by refusing to succumb to the oft-assumed incoherence of the so-called Second Letter of Clement. The modest attempt of this article has been to fill part of the void left by its structural neglect: its conceptual unity or coherence. When perspectivised from the viewpoint of the addressees, a previously unforeseen level of coherence can be realised when understood on the terms of patron–client relationships. By addressing two critical questions, it was demonstrated how a constructed patron–client frame can provide a navigational schema for the discourse content. Firstly, regarding how the frame is activated, elements of patron–client relationships, including reciprocity, were identified from the very opening sentence. Additionally, a dense concentration of lexical symbols within the semantic domain of patronage served to activate or ‘call up’ the conceptual structure. Secondly, regarding the question of how the content fits into the heuristic model, God and Christ fill the ‘patron’ slot while the audience and author fill the ‘client’ slot (Figure 2). Thirdly, it was suggested that other entities, such as the ‘unbelievers’ and ‘presbyters’, pose no legitimate threat to the discourse frame as they only make brief and fleeting appearances.
Beyond offering another example of the usefulness of both cognitive linguistic and socio-rhetorical approaches to ancient literature, the proposed coherence provides a framework within which a more precise understanding of the document’s message and purpose can be discovered. As the patron–client frame governs the discourse at its highest level, the earlier suggestion of a central message of repayment for salvation is especially accommodated (Kelhoffer 2020:516; Varner 2020:41, 53). Most coherently, Clement should be seen as attempting to evoke an ongoing behavioural response to the gift of salvation that culminates in a future deliverance from judgement. A stream of scholarship, characterised by works such as Zeba Crook’s aptly titled Reconceptualising Conversion (2004), has long explored the way ancients understood gods in terms of patronage.21 Viewing Second Clement through the lens outlined above extends such conceptions into another instance of early Christianity, paving the way for future application of the patron–client frame and similar ones in other early Christian writings.
Acknowledgements
This article is based on research originally conducted as part of Devin Arinder’s doctoral thesis titled ‘Discourse as Reciprocity: A Cognitively Informed Text-Linguistic Analysis of Second Clement’, set to be submitted to the Department of New Testament and Related Literature, University of Pretoria in 2026. The thesis is currently unpublished and not publicly available. The thesis was supervised by Ernest van Eck. The thesis was reworked, revised and adapted into a journal article for publication. The author confirms that the content has not been previously published or disseminated and complies with ethical standards for original publication.
A prior version of this article was presented in the Apostolic Fathers and Related Early Christian Literature unit at 2025 SBL International Meeting in Uppsala, Sweden. The presentation titled ‘Second Clement and the Patron-Client Frame: Coherence in the Document’s Structure’ has since been revised and expanded for journal article form. Permission has been granted by conference organisers for publication. Much gratitude to those in the unit who provided invaluable feedback, including especially James Kelhoffer, Martin Landgren and Katja Kuyanpää, who particularly reviewed an earlier draft of the article.
Competing interests
The authors declare that they have no financial or personal relationships that may have inappropriately influenced them in writing this article.
CRediT authorship contribution
Devin Arinder: Conceptualisation, Investigation, Methodology, and Writing – original draft. Ernest van Eck: Supervision. All authors reviewed the article, contributed to the discussion of results, approved the final version for submission and publication and take responsibility for the integrity of its findings.
Ethical considerations
The research for the doctoral thesis encompassing the methodology and subject matter of the article was approved by the Research Ethics Committee in the Faculty of Theology and Religion at the University of Pretoria (No. T008/24).
Data availability
The authors confirm that the data supporting this study and its findings are available within the article and its listed references.
Disclaimer
The views and opinions expressed in this article are those of the authors and are the product of professional research. They do not necessarily reflect the official policy or position of any affiliated institution, funder, agency or that of the publisher. The authors are responsible for this article’s results, findings and content.
ReferencesBarclay, M.G., 2015, Paul and the gift, Eerdmans, Grand Rapids, MI.Chafe, W., 2015, ‘Constraining and guiding the flow of discourse’, in D.Tannen, H.E.Hamilton & D.Schiffrin (eds.), The handbook of discourse analysis, 2nd edn., pp. 391–405, Wiley Blackwell, Malden, MA.Cienki, A., 2010, ‘Frames, idealized cognitive models, and domains’, in D.Geeraerts & H.Cuyckens (eds.), The Oxford handbook of cognitive linguistics, pp. 170–187, Oxford University Press, Oxford.Crook, Z.A., 2004, Reconceptualising conversion, De Gruyter, Berlin.Crook, Z.A., 2020, ‘Patronage’, in Z.A.Crook (ed.), The ancient Mediterranean social world: A sourcebook, pp. 45–58, Eerdmans, Grand Rapids, MI.Crowe, B.D., 2015, ‘Like father, like son: Unravelling the proto-Trinitarian approach of 2 Clement’, Westminster Theological Journal77(2), 251–268.DeSilva, D.A., 2022, Honor, patronage, kinship, and purity: Unlocking New Testament culture, 2nd edn., InterVarsity Press Academic, Downers Grove, IL.Donfried, K.P., 1974, The setting of Second Clement in Early Christianity, Brill, Leiden.Dooley, R.A. & Levinsohn, S.H., 2001, Analyzing discourse: A manual of basic concepts, SIL International, Dallas, TX.Downs, D., 2009, ‘Is God Paul’s patron? The economy of patronage in Pauline theology’, in B.W.Longenecker & K.D.Liebengood (eds.), Engaging economics: New Testament scenarios and early Christian reception, pp. 129–156, Eerdmans, Grand Rapids, MI.Drobner, H.R., 2007, The fathers of the church: A comprehensive introduction, transl. S.Schatzmann, Baker Academic, Grand Rapids, MI.Elliott, J.H., 2003, ‘Patronage and clientage’, in R.L.Rohrbaugh (ed.), The social sciences and New Testament interpretation, pp. 144–168, Hendrickson Publishers, Peabody, MA.Ensink, T. & Sauer, C., 2003, ‘Social-functional and cognitive approaches to discourse interpretation: The role of frame and perspective’, in T.Ensink & C.Sauer (eds.), Framing and perspectivising in discourse, pp. 1–22, John Benjamins, Amsterdam.Evans, V. & Green, M., 2006, Cognitive linguistics: An introduction, Edinburgh University Press, Edinburgh.Evans, V., 2019, Cognitive linguistics: A complete guide, 2nd edn., Edinburgh University Press, Edinburgh.Fillmore, C.J., 1977, ‘Scenes-and-frames semantics’, in A.Zampolli (ed.), Linguistic structures processing, pp. 55–82, North Holland Publishing, Amsterdam.Fillmore, C.J., 1982, Frame semantics, viewed n.d., from http://brenocon.com/Fillmore%201982_2up.pdfFreese, J.H., 1920, The library of Photius, Macmillan, London.Garnsey, P. & Woolf, G., 1989, ‘Patronage of the rural poor in the Roman world’, in A.Wallace-Hadrill (ed.), Patronage in ancient society, pp. 153–170, Routledge, London.Geeraerts, D., 2021, ‘Cognitive semantics’, in X.Wend & J.R.Taylor (eds.), The Routledge handbook of cognitive linguistics, pp. 19–29, Routledge, New York, NY.Grant, R.M. & Graham, H.H., 1965, First andSecond Clement, vol. 2, R.M.Grant (ed.), Nelson, New York, NY.Gregory, A.F. & Tuckett, C.M., 2007, ‘2 Clement and the writings that later formed the New Testament’, in A.F.Gregory & C.M.Tuckett (eds.), The reception of the New Testament in the Apostolic Fathers, pp. 251–292, Oxford University Press, Oxford.Harrison, J.R., 2017, Paul’s language of grace in its Graeco-Roman context, Wipf & Stock, Eugene, OR.Jefford, C.N., 2012, Reading the Apostolic fathers: A student’s introduction, 2nd edn., Baker Academic, Grand Rapids, MI.Kelhoffer, J.A., 2013, ‘Reciprocity as salvation: Christ as salvific patron and the corresponding “payback” expected of Christ’s earthly clients according to the Second Letter of Clement’, New Testament Studies59(3), 433–456. https://doi.org/10.1017/S0028688512000380Kelhoffer, J.A., 2015, ‘If Second Clement really were a “sermon,” how would we know, and why would we care?: Prolegomena to analyses of the writing’s genre and community’, in M.Grundeken & J.Verheyden (eds.), Early Christian communities between ideal and reality, pp. 83–108, Mohr Siebeck, Tübingen.Kelhoffer, J.A., 2017, ‘Second Clement and Gnosticism: The status quaestionis’, Early Christianity8(1), 124–149. https://doi.org/10.1628/186870317X14876711440240.Kelhoffer, J.A., 2018, ‘How did Second Clement originally end?: A study of 2 Clement 19–20 and the imitation of both First and Second Clement’, Early Christianity9(4), 432–483. https://doi.org/10.1628/ec-2018-0031Kelhoffer, J.A., 2020, ‘Second Clement’, in C.Keith, H.K.Bond, C.Jacobi & J.Schröter (eds.), The reception of Jesus in the first three centuries, pp. 513–530, T&T Clark, London.Levinsohn, S.H., 2023b, Self-Instruction materials on non-narrative discourse analysis, SIL International, Dallas.Lightfoot, J.B., 1890, The Apostolic fathers, Part 1: S. Clement of Rome, vol. 2, Macmillan, New York, NY.Lindemann, A., 1992, Die Clemensbriefe, Mohr Siebeck, Tübingen.Longacre, R.E. & Hwang, S.J.J., 2012, Holistic discourse analysis, 2nd edn., SIL International, Dallas, TX.MacGillivray, E.D., 2009, ‘Re-evaluating patronage and reciprocity in antiquity and New Testament studies’, Journal of Greco-Roman Christianity and Judaism6, 37–81.Malina, B.J. & Rohrbaugh, R.L., 2003, Social-science commentary on the Synoptic Gospels, 2nd edn., Fortress Press, Minneapolis, MN.Moxnes, H., 1991, ‘Patron–client relations and the new community in Luke–Acts’, in J.H.Neyrey (ed.), The social world of Luke–Acts: Models for interpretation, pp. 241–268, Hendrickson Publishers, Peabody, MA.Neyrey, J.H., 2005, ‘God, benefactor and patron: The major cultural model for interpreting the deity in Greco-Roman antiquity’, Journal for the Study of the New Testament27(4), 465–492. https://doi.org/10.1177/0142064X05055749Parvis, P., 2006, ‘2 Clement and the meaning of the Christian homily’, Expository Times117(7), 265–270. https://doi.org/10.1177/0014524606064362Peters, J., 2021, ‘1 and 2 Clement’, in M.F.Bird & S.D.Harrower (eds.), The Cambridge companion to the Apostolic fathers, pp. 186–207, Cambridge University Press, Cambridge.Pratscher, W., 2007, Der Zweite Clemensbrief, Vandenhoeck & Ruprecht, Göttingen.Pratscher, W., 2010, The Apostolic fathers: An introduction, Baylor University Press, Waco, TX.Robbins, V.K., 2009, The invention of Christian discourse, vol. 1, Deo Publishing, Leiden.Sahlins, M., 1972, Stone age economics, Aldine-Atherton, Chicago, IL.Saller, R.P., 2002, Personal patronage under the early empire, Cambridge University Press, Cambridge.Sanders, T. & Spooren, W., 2010, ‘Discourse and text structure’, in D.Geeraerts & H.Cuyckens (eds.), The Oxford handbook of cognitive linguistics, pp. 916–941, Oxford University Press, Oxford.Starr, J., 2018, ‘Protreptic and paraenesis: The Second Epistle of Clement’, in O.Alieva, A.Kotzé & S.Van der Meeren (eds.), When wisdom calls: Philosophical protreptic in antiquity, pp. 251–266, Brepols Publishers, Turnhout.Stegemann, C., 1974, ‘Herkunft und Entstehung des sogenannten zweiten Klemensbriefs’, PhD thesis, University of Bonn, Bonn.Stewart, E.C., 2010, ‘Social stratification and patronage in ancient Mediterranean societies’, in D.Neufeld & R.E.DeMaris (eds.), Understanding the social world of the New Testament, pp. 156–166, Routledge, London.Tenbrink, T., 2020, Cognitive discourse analysis: An introduction, Cambridge University Press, Cambridge.Todd, R.W., 2016, Discourse topics, John Benjamins Publishing Company, Amsterdam.Tuckett, C.M., 2012, 2 Clement: Introduction, text, and commentary, Oxford University Press, Oxford.Tugwell, S., 1989, The Apostolic fathers, Geoffrey Chapman, London.Van Eck, E., 2016, The parables of Jesus the Galilean: Stories of a social prophet, Cascade Books, Eugene, OR.Varner, W., 2020, Second Clement: An introductory commentary, Cascade Books, Eugene, OR.Von Thaden, R.H., 2016, ‘A cognitive turn: Conceptual blending within a sociorhetorical framework’, in V.K.Robbins, R.H.Von Thaden & B.B.Bruehler (eds.), Foundations for sociorhetorical exploration: A rhetoric of religious antiquity reader, pp. 285–328, SBL Press, Atlanta, GA.Wang, Y. & Guo, M., 2014, ‘A short analysis of discourse coherence’, Journal of Language Teaching and Research5(2), 460–465. https://doi.org/10.4304/jltr.5.2.460-465Wengst, K., 1984, Didache (Apostellehre), Barnabasbrief, Zweiter Klemensbrief, Schriften an Diognet, Kösel Verlag, München.Williams, R., 1997, ‘Charismatic patronage and brokerage: Episcopal leadership in the letters of Ignatius of Antioch’, PhD thesis, University of Ottawa, Ottawa.
How to cite this article: Arinder, D. & Van Eck, E., 2026, ‘An apostolic enigma? Dispelling the incoherence of Second Clement through the patron–client frame’, HTS Teologiese Studies/Theological Studies 82(1), a11027. https://doi.org/10.4102/hts.v82i1.11027
See, for example, Wengst (1984:208) who says, ‘Der 2 Klemensbrief hat keinen strengen Aufbau und keine klare Gedankenfolge’. Similarly, Pratscher (2010:73) begins his section on the document’s structure with the remark that ‘The structure of 2 Clem is difficult to determine’ and observes that it lacks an ‘overarching logical thought construction’. Grant and Graham (1965:110) agree that ‘there is no clear logical development’ as the ‘preacher proceeds from one thing to another simply by association of ideas, sometimes by verbal expression’. More recently, Parvis (2006:70) concedes it is ‘neither well written nor well organized’. Starr (2018:258 n. 26) says it ‘is not a tightly structured text’. Lastly, in the most recent commentary, Varner (2020:25) concludes ‘The effort to discern an overall structure to this “discourse,” can be a challenge due to its sometimes hectic and scattered character’.
Tuckett (2012:27–33) does include an introductory section on ‘Literary Unity’; however, this section focuses on the question of originality of chapters 19–20.
Pratscher (2007:22–23) catalogues a near exhaustive list of no less than 13 unique structures.
For a short, helpful introduction to discourse coherence, see Wang and Guo (2014:460–465). The approach to discourse coherence described here is sometimes termed ‘discourse-as-process’. See also Sanders and Spooren (2010:916–941).
As conventional in linguistic literature, the denotation of the frame is rendered in all capitalised letters.
Ensink and Sauer (2003:2–6) show how frames, particularly ‘knowledge frames’ have been used across various fields, including Anthropology, Sociology, Linguistics, and most pertinently, Discourse Analysis. See also Cienki (2010:173) who notes the apparent interchangeable usage of ‘frame’, ‘scene’, ‘script’, and ‘schema’. Additionally, Robbins (2009:98–120) similarly develops his six ‘rhetorolects’ for genre analysis based on frame semantics, prototype theory, and idealized cognitive models.
Beyond the well-known works of the Roman social elite, including Seneca’s De beneficiis and Cicero’s De officiis, evidence has increasingly accumulated in scholarship that such relational dynamics were commonplace to the lower-level social strata. See, for example, Garnsey and Woolf (1989:153–70) and Harrison (2017:64–96). Accompanying Second Clement in the Apostolic Fathers corpus, Williams (1997) has also observed patron–client dynamics in Ignatius’ writings.
Most commonly, three types of reciprocity are delineated amongst relationships. First, there is generalised reciprocity, or the altruistic giving in the primary interest of the other without necessarily expecting a return. Second, there is balanced reciprocity characterised by the mutual exchange of goods either between equals (friends) or non-equals (patron and clients). Third, there is negative reciprocity, or exploitative exchange, primarily benefitting the giver. See Stewart (2010:157); Neyrey (2005:469); see also Sahlins (1972:193–196).
This article refers to the author as ‘Clement’ while recognising that virtually all aspects of the document’s external circumstances remain highly contested and uncertain. For recent overviews, see Varner (2020:3–28) and Peters (2021:202–203).
All translations are original unless otherwise noted.
The 10 surface forms include λαβειν (1.2), ἐλάβομεν (1.5), δώσομεν (1.3, 1.5), ἒδωκεν (1.3), ἐχαρίσατο (1.4), ὀφείλομεν (1.3), μισθὸν (1.5), ἀντιμισθίαν (1.3), and ἀντιμισθίας (1.5).
Those six include forms of λαμβάνω (1.2, 1.5), δίδωμι (1.3, 5), ὀφείλω (1.3), μισθὀν (1.5), and ἀντιμισθία (1.3, 1.5). Though the original cluster of 10 unique symbols is never eclipsed throughout the discourse, the second highest cluster appears in 8.4–9.6 with seven symbols within the semantic domain of patronage.
For example, the document lacks any initial greetings or explicit indication of who the addressees are or their location. Often, the document is put forth as the earliest Christian ‘homily’; however, this designation has more recently been challenged. See Kelhoffer (2015:83–108); cf. Drobner (2007:57).
As several have observed, ‘Lord’ [κύριος] is also used throughout the document both to refer to Jesus and in a few places, introduce citations from the Old Testament, implying some overlap in conception (see e.g. 2 Clem 13.2, Is 52:5; 2 Clem 17.4, Is 66:18; cf. 2 Clem 4.2, 4.5, 5.2, 6.1, 6.2, 8.5, 9.11, 12.2). See Gregory and Tuckett (2007:251–292). At least two commentators observe the ambiguous identification between ‘Jesus’ and ‘God’. Varner (2020:136) suggests that ‘the use of “the Lord” indicates that the author did not reserve a title for the Father as distinct from the Son’. Similarly, commenting on the use of ‘Lord’ in 17.3, Tuckett (2012:281–282) says, ‘Whether the “Lord” here is God or Jesus is not clear; for the author there was doubtless little difference in status, and certainly no difference in real substance, between the commands of Jesus (from gospel traditions) and the commands of God (from scripture)’. In this way, both God and Jesus filling the ‘patron’ slot affirms what has already been discerned by commentators.
In this way too, Clement’s commands are decisively mitigated. See, for instance, Levinsohn’s (2023:82–83) salience or potency ranking scheme for Greek command forms.
Along with several others, this study considers the final two chapters (19–20) to be unoriginal and thus excluded from the analysis (e.g. Donfried 1974; Kelhoffer 2018:432–83; Stegemann 1974; Tuckett 2012:27–33; cf. Varner 2020:25–28).
For further evidence of Clement’s intentional enjoining to the audience, consider his remark that, ‘Now I think that I have not given trivial counsel about self-control, which if anyone does they will not [only] change their mindset, but also will save themselves and me, the one who counsels them’ (2 Clem 15.1, emphasis added).
Other entities could include the supporting citations involving ‘Peter’ (5.3–4) and similarly, ‘Noah, Job, and Daniel’ (6.8).
Thus, for example, the entities can be accounted for by lower level or embedded frames while the patron–client frame continues to govern at the discourse level. On frame embedment, see again Von Thaden (2016:301) who refer to ‘local frames’ that ‘function below the level of the cultural organizing frame and allow exegetes a more precise tool to explicate the meaning of certain segments of discourse’.
Both Pratscher (2007:213) and Lindemann (1992:252) suggest that the prepositional phrase ἐν ἠμιν [among us] identifies the ones ‘who have acted ungodly and distorted the commands of Jesus Christ’ (17.6) as part of the addressees. As no other grammatical distinction can be found, the participles ἀσεβήσαντας [who have acted ungodly] and παραλογισαμένους [who have distorted] likely characterise the ‘unbelievers’ from the previous verse (2 Clem 17.5). Nevertheless, whether the ‘unbelievers’ represent a subset of the audience remains contested among scholars. See, for example, Tuckett (2012:284–286); see also Tugwell (1989:141).
See also, DeSilva (2022:125–175), Neyrey (2005:465–492), and Harrison (2017). John Barclay’s Paul and the Gift (2015) stands out as a landmark work in this respect. Downs (2009:129–156) notably argues that Paul does not view God as a patron.