Inscriptions as dedications

The normality, in the Greek world, of the use of religious sanctuaries as places for the publication of public documents is well recognised (Lalonde Reference Lalonde1971, 52–67; Detienne Reference Detienne and Detienne1988, 42; Hölkeskamp Reference Hölkeskamp1992, 100; Thomas Reference Thomas, Foxhall and Lewis1996, 28–9). For Athens, the central importance of the Acropolis as a place of publication for the development of Athenian public epigraphy has been clearly established (Liddel Reference Liddel2003),Footnote ⁴⁰ and its prominence has come to be widely accepted as reflecting the importance of the Acropolis as a religious space, which thereby conferred ‘a religious aura’ on the enactments erected there (Osborne, R. Reference Osborne, Goldhill and Osborne1999, 347; Liddel Reference Liddel2003, 81). Other features of the form and fabric of the document stelai at Athens (and elsewhere) have also been identified as drawing these monuments into a relationship with the gods – including the pediments adorning many of the stelai which echo the architecture of surrounding temples (Lambert Reference Lambert2018, 22), the care given to the layout of the lettering, to make it pleasing to the gods (Meyer, E. Reference Meyer2013, 461) and even the very fabric of the stelai themselves, imperishable marble (Bresson Reference Bresson, Bresson, Cocula and Pébarthe2005). Two recent contributions, by Elizabeth Meyer and Stephen Lambert, have explored at some length the importance of the Acropolis as a sacred context for the interpretation of these monuments. The key question they raise concerns the status of these monuments – in particular whether, given the fact that they were deliberately set up in the sanctuaries, they should be understood as dedications to the gods. For Meyer, public document stelai set up in sanctuaries should, like any other property placed in a sacred space, be recognised as gifts for the gods, but Lambert highlights certain difficulties (Meyer, E. Reference Meyer2013, 459–60; Lambert Reference Lambert2018, 22–3).Footnote ⁴¹

In part, this is an issue of terminology. Dedications tend to be identified, in accompanying inscriptions, by the use of the aorist verb anetheken, he/she ‘set up’, often specifying a recipient deity in the dative, or are typically described as an anathema, a ‘thing set up’, or an agalma, an ‘adornment’ (van Straten Reference Straten and Versnel1981; Parker Reference Parker2004; Jim Reference Jim2012; Patera Reference Patera2012). Although there is at least one honorific decree, from Elis, inscribed on a bronze disc, which describes itself as an ‘agalma belonging to Zeus’,Footnote ⁴² at Athens, documentary inscriptions tend to refer to themselves in different terms. The publication clauses of public decrees instead refer to ‘writing up’ (anagrapsai) the text on a stone stele, which is to be ‘stood’ (stesai) or, sometimes, ‘set down’ (katatheinai) in a particular place (usually on the Acropolis), but not for a particular god (Liddel Reference Liddel2003; Lambert Reference Lambert2018, 22–3).

The real crux of this debate, however, is the sense in which document stelai could have been considered gifts for the gods. How far could they have been conceived of as gifts benefitting the gods in any of the ways which seem to have been recognised by contemporaries, for example those set out by the eponymous interlocutor in Plato's Euthyphro – as giving honours (timai), the god's dues (gera) or thanks (charis)?Footnote ⁴³ Other kinds of dedications, following victory or the fulfilment of a prayer, obviously fulfil more than one of these functions (van Straten Reference Straten and Versnel1981, 65–75). They give thanks for a positive result brought about by a god and honour the deity in question by bearing witness to their power. How the stelai inscribed with public documents could be thought of as benefitting the god in this way is, on the face of it, harder to pin down.Footnote ⁴⁴ That is, unless we take the evidence of the document reliefs seriously.

The document/relief monuments seem to provide answers to both of the central questions in this debate, making the ‘strong conceptual overlap’, which Lambert acknowledges between document stelai and dedications in general, nearly perfect (Lambert Reference Lambert2018, 23). In form, these monuments identify themselves with dedicatory reliefs, as they would have appeared in situ, much more closely than is now apparent: visually, they combine the relief which was ‘set up’ (anatheken) with the pillar which was the means by which it was ‘set up’ off the ground (van Straten Reference Straten and Schachter1992; Lawton Reference Lawton2017, 19, fig. 2). The reliefs themselves, moreover, by depicting deities performing an action which was equivalent with and therefore linked in some way to the action of the decree, suggest a way in which these monuments could have been thought of as suitable gifts for the gods. As I have argued, the implication in these cases is that the gods were in some way involved in the decisions and actions which the inscriptions documented. On this basis, the document stelai (both those set up with a relief and potentially those without) could have been understood as gifts to the gods, like other dedications, honouring them by highlighting their power to influence the Athenian state and, perhaps, expressing thanks.

ΘΕΟΙ

The practice of inscribing, above the text of a public inscription, the Greek word for god, in either the singular (theos) or the plural (theoi), was a widespread epigraphic practice in the Greek world, attested, by my count, at more than 60 communities (see Larfeld Reference Larfeld1902, 436–7, for an extensive list). The four letters of the heading are often inscribed in larger or smaller lettering than the document they accompany, and, at Athens, are further distinguished from it by often being incorporated in the architectural mouldings of the inscription or by being spaced evenly across its width (Θ  Ε  Ο  Ι). At Athens, this heading (always in the plural) is first attested in public epigraphy in the 440s bc, in relation to the accounts for Pheidias’ statue of Athena Parthenos, not long before the first document reliefs.Footnote ⁴⁵ In this first case it appeared alongside the words ‘Athena’ and ‘Tyche’ (Fortune). Thereafter, on its own, it was apparently a regular, if apparently haphazard feature of public Athenian epigraphic practice, preserved in relation to 136 documents produced under civic authority and 34 documents produced by civic subdivisions (tribes, demes, phratries, orgeones) until the end of the fourth century.Footnote ⁴⁶ Its irregular use is also attested, idiosyncratically, in a small handful of mostly private texts of different kinds.Footnote ⁴⁷

A variety of terms have been applied by scholars to this heading, usually in passing, without much by way of discussion:Footnote ⁴⁸ Weiheformel (‘dedication formula’: Larfeld Reference Larfeld1902, 436–7), ‘words of good omen’ (Robinson Reference Robinson1958, 74) and, most frequently, ‘invocation’ (Hill and Merrit Reference Hill and Meritt1944, 8; Guarducci Reference Guarducci1969, 43; Chaniotis Reference Chaniotis1996, 84–5; Rhodes with Lewis Reference Rhodes and Lewis1997, 4). Despite this variation in terminology, it has generally been assumed to reflect some kind of ritualised appeal to the gods, and one made in relation to the text recorded and what it represents (e.g. decree), rather than the monument on which it was inscribed as a physical object.Footnote ⁴⁹ The reason for this is presumably the fact that this heading is almost never found in relation to other monuments explicitly set up as dedications to the gods.Footnote ⁵⁰ However, opinions have differed concerning the precise timing and object of the ritualised appeal. Geoffrey Woodhead, in particular, has taken it, in an Athenian context, as indicating that ‘the proper religious exercises had been performed or invocations made’ before the assembly which made the decree,Footnote ⁵¹ and, in fact, such prayers, along with a purificatory sacrifice of a piglet, are well attested at Athens (Rhodes Reference Rhodes1972, 36–7; Parker Reference Parker2005, 99–101; Sommerstein in Sommerstein and Bayliss Reference Sommerstein and Bayliss2012, 47–56), with a number of references surviving in the Attic Orators (Aeschines 1.23; Demosthenes 19.70, 23.97, 24.20), as well as a lengthy parody in Aristophanes’ Thesmophoriazusae (l. 295–372). Other scholars, conversely, have tended to take it as a reference to a subsequent act, commending the contents of the text to divine protection and/or praying that its consequences will be conducive to the best interests of the community – and here a model has been identified in decrees which (perhaps exceptionally) make specific provision for subsequent sacrifices and prayers.Footnote ⁵²

One difficulty posed to the interpretation of this heading is the haphazardness with which it seems to have been used in all of the local epigraphic traditions in which it is attested. At Delos, for example, where the inscription of the theoi heading was a regular feature of epigraphic practice, the 65 examples identified by Robert Pounder constitute only 18 per cent of the stelai which survive with their top intact on his analysis.Footnote ⁵³ At Eretria, where the number of surviving texts is lower, for the period before 300 bc the proportion is higher – it is present in nine of 17 (53 per cent) examples where the top is intact – but for some reason it falls out of use entirely on decrees produced after 300 bc.Footnote ⁵⁴ It is difficult to derive strictly comparable figures for Athens for the Classical period as a whole. On the face of it, the number of examples which we know of for Athens in the Classical period – 174, a very small fraction indeed of the total number of public documents inscribed – suggests that the use of this heading was even more sporadic at Athens. However, the highly fragmentary state in which most decree stelai are preserved makes this kind of simplistic comparison problematic. The recently published, well-illustrated volume of Athenian decrees, IG II³ 1, makes it possible, for particular periods, to compare the number of uses of this heading and the number of tops of decree stelai preserved intact without it. Thus for the only fourth-century period currently covered, 352/351–322/321 bc, there are 41 attestations of the heading against only three apparently clear examples of its absence.Footnote ⁵⁵ For the next published period, 300/299–230/229 bc, there are 28 attestations as opposed to nine apparent examples of absence.Footnote ⁵⁶ Nonetheless, although the use of this heading at Athens was perhaps not as intermittent as the raw figures would imply, the variation which we do see, at Athens and elsewhere, suggests that, whatever the precise significance of the heading, inscribing it cannot have been felt to be a ritual necessity.

The best means of assessing the meaning of this heading is to consider the ways in which it could have been read. Details of grammar and vocabulary, in particular, are important in terms of the theological readings of the inscribed headings which they allow and privilege.Footnote ⁵⁷ The heading – always written as theoi at Athens – was capable of being read as either nominative or vocative, which are indistinguishable in the plural.Footnote ⁵⁸ The nominative reading takes strong support from the fact that the version of this heading used by Greek cities of the mainland, the singular theos, is only capable of being read in the nominative. There is also some support for a contemporary reading of this in the vocative, given the use of theoi in the plural as an interjection in speech in Classical Athenian literary texts (see below).

The use of the dative or genitive cases would have expressed specific roles for the gods, whom one would expect to be either specifically named or obvious in context – in the first instance, as the recipient/s of an object which was being dedicated and, in the second, as the owner/s of the object in consequence of its dedication (Parker Reference Parker2004, 274–81). Instead, a reading in the nominative case would place the gods in the role of active agents, performing some kind of action (indeed, a verb, although omitted, is sometimes implied where the heading is used in the singular in central Greece, when it is juxtaposed with an object in the accusative, clearly implying a syntactic relation).Footnote ⁵⁹ The significance of a vocative reading is somewhat harder to pin down. In particular, there are some problems with reading theoi in the vocative as the direct quotation of the invocation from a formal prayer, given that such prayers generally use both names and cult epithets in an effort to obtain the aid of particular gods (Burkert Reference Burkert1985, 74–5; e.g. Aristophanes, Thesmophoriazusae 312–30). Where it occurs in literary texts, usually following the interjection ‘o’ (lacking here), and often in connection with other deities (o Zeus and gods!), theoi in the vocative often seems to function as an exclamation, generally at a moment of revelation or strong emotion.Footnote ⁶⁰ On a few occasions, however, a sense of this as a deliberate invocation of divine power is more clearly apparent, especially in the cases where it seems to be connected with the expression of a wish using a verb in the optative, which sometimes functions as a kind of informal prayer.Footnote ⁶¹

The word theos itself was the primary way in which divine agency, in both the singular and the plural, was described from the human perspective, in historical and literary sources.Footnote ⁶² Its use as an anonymous term for a divine actor seems to have reflected a conventional understanding that divine intervention in human affairs was real and that its occurrence could be recognised, but that the identity of the divine actor in question was usually beyond human knowledge. It was also, however, used in other contexts, like divination, in instances when the identity of the deity should not have been in doubt. Thus, although the oracular sanctuaries at Dodona and Delphi were inseparably connected with particular gods, Zeus and Apollo respectively, who might be addressed directly by consultants in posing their questions, replies were often presented as being made by or sought from not Zeus or Apollo, but ho theos, the god.Footnote ⁶³ In fact, this usage is embedded in the language used to describe divination itself, from the term for divine utterances and signs (thesphaton), to the term for seer (theopropos) and the state of divine inspiration proper to them (theazein; on these, see Burkert Reference Burkert1985, 112). In Walter Burkert's analysis, what connects the uses of the word theos, including the exclamation theoi, is the idea of the manifestation of divine power in action.Footnote ⁶⁴

The reference of the divine action, according to both readings, is not explicitly pinned down, and because of this it is possible to reconstruct a range of different ways in which this agency could have been read. The two modern mainstream interpretations of this heading, in fact, lead in different directions. In the first instance, in which this reference to the gods is related to the performance of prayers before the act commemorated by the inscription was performed (a decision reached by the assembly, and now documented), theoi would make for a stronger statement and imply a finite verb (‘gods have done …’) or, in the case of the vocative reading, the recognition of a manifestation of divine power (‘o gods’). In the second instance, in which the heading is read as referring to a subsequent ritual invocation performed in aid of the success of the action which has been decided upon, that future assistance would imply an optative (‘may the gods …’;Footnote ⁶⁵ in the case of the vocative reading ‘gods, may …’).

The theoi heading, in presenting the gods as active agents, converges with the document reliefs, which depict gods in precisely this way. Whether or not the theoi heading was one of the inspirations for the figured document reliefs – and, given their relative chronology, this is possible – both seem to reflect a wider belief in the involvement of the gods in political action at Athens. The document reliefs, however, arguably resolve the potential ambiguity of the headings, at least partially, and suggest that the stronger reading was favoured at Athens – that it served as a statement or recognition of divine power. The depiction of the gods performing a concrete action equivalent to that performed by the document below seems to assert a role for them in relation to the action of the decree, whatever the precise mechanism of that involvement, rather than to express an uncertain wish for their future interest in it.

Modalities of production

The study of public epigraphic culture rapidly raises the question of why inscribed monuments, and particular features of them, were produced. In this case, why were some decrees and public documents (and apparently not others) inscribed on imperishable media and deposited in sanctuaries – and, why, at Athens, above some of these inscriptions (and not others) were reliefs sculpted and/or the heading theoi?

Despite a recent challenge from Michael Osborne, it is clear that not all public enactments were inscribed in the Greek world, even at Athens, where the epigraphic record is richest.Footnote ⁶⁶ Inscription, instead, was selective, and from the quite patchy collection of material which has survived, it is clear that selection was largely dictated by particular functions which inscription was understood to perform that were relevant to the purpose of particular kinds of enactments, rather than for general democratic transparency and accountability, or, indeed, as a means of highlighting texts of particular religious concern (Meyer, E. Reference Meyer2013). This emerges quite clearly from consideration of the relative numbers of different kinds of decree which survive, which have been categorised and quantified for the fifth century to 404/403 bc, by James Sickinger (Reference Sickinger1999, 242 n. 45), and for the short fourth, 403/402–322 bc, by Mogens Herman Hansen (Reference Hansen1987, 110). The largest category of surviving Athenian decrees for the fifth and fourth centuries bc, honorific decrees, seems to reflect the fact that inscription of such a decree was an honour in itself – constituting 38 per cent (68 of 177) of all identifiable decrees for the fifth century and 74 per cent (288 of 388) of all identifiable decrees for the fourth century.Footnote ⁶⁷ Some of the next largest category, decrees relating to interstate relations – 31 per cent (54 of 177) of identifiable fifth-century decrees and 17 per cent (67 of 388) of identifiable fourth-century decrees – are also explicitly honorific, granting praise and privileges to foreign communities; most of the remainder are treaties, deposited according to what was already a well-established practice by which states sought to endow them with additional authority and binding power by placing them under the protection – and scrutiny – of the gods (e.g. Meiggs – Lewis, SGHI 10; Rhodes – Osborne, GHI 39 l. 17–22 with Thomas Reference Thomas1989, 46), reinforcing the role the gods were already called on to play in the oaths that were sworn (Steiner Reference Steiner1994, 64–6; Bayliss in Sommerstein and Bayliss Reference Sommerstein and Bayliss2012, 151–75). Specifically religious enactments are only the third-largest category quantified by Sickinger and Hansen – 25 per cent (46 of 177) in the fifth century and five per cent (21 of 388) in the fourth. Other laws and decrees account for five per cent (nine of 177) of the fifth century dataset and three per cent of the fourth (12 of 388). In the case of public decrees of other kinds, and particularly laws (which were only rarely inscribed, especially in the fourth century), there is usually a clear logic apparent for the inscription and erection of that particular text (often in a specific place beyond the Acropolis), which suggests that there was not a strong underlying general drive for publication (Richardson Reference Richardson, Flensted-Jensen, Nielsen and Rubinstein2000; Lambert Reference Lambert2018, 19–46). As we have seen, in the case of reliefs, it is similarly hard to identify any particular religious preoccupation in their general use.

There is, moreover, little reason to believe that document reliefs in particular were the result of a specific public mandate, as none of the publication clauses which have been preserved in texts makes any reference to an accompanying relief, and there is no clear correlation even between the sum allocated for publication and the presence of a document relief (Lawton Reference Lawton1995, 22–8; Clinton Reference Clinton1996). Because of this silence, the detail of the process involved is disputed – whether document reliefs were paid for by supplements to the formal grant by those particularly associated with a monument (e.g. the honorand, his family, the proposer of the decree or, indeed, the secretary, whose service it commemorated)Footnote ⁶⁸ or were obtained by the commissioning officials within the designated budget through a competitive tendering procedure (Lawton Reference Lawton1995, 22–8; Deene Reference Deene2016). However they were paid for (and they need not all have been paid for in the same way), the reliefs were probably the result of some form of personal initiative, for which the primary motivation was presumably the desire to enhance the visibility, prestige and perhaps the beauty of particular monuments.

In the case of the theoi heading, as in the case of the two other kinds of epigraphic practice, there is no evidence that the decision to add this heading was determined or even particularly strongly influenced by the religious significance of the content of the document. Instead the numbers of different kinds of document for which headers have been preserved, correspond quite closely to the proportion of that kind of text inscribed in the first place (compare above).Footnote ⁶⁹ Excluding 26 decrees which cannot be categorised with certaintyFootnote ⁷⁰ and 24 accounts (a category not included in Sickinger's and Hansen's datasets),Footnote ⁷¹ as well as a single civic lease (Agora 19 L6) and the two dedications by prytaneis (IG II³ 4 29, 79), we are left, by my count, with 83 attestations in relation to public documents. For the fifth and fourth centuries, honorific decrees account for 62 per cent (52 of 83) of those with the heading,Footnote ⁷² a further 19 per cent (16 of 83) relate to interstate relations,Footnote ⁷³ 13 per cent (11 of 83) relate to specifically religious affairsFootnote ⁷⁴ and the remaining five per cent (four of 83) are laws of other kinds.Footnote ⁷⁵ Unlike document reliefs, however, it is not clear that this heading served a function other than as a reference to divine agency, which would explain its sporadic use. In this case the answer probably relates to the fact that, within the wider context of making these monuments, the reference to divine agency which it made was already made clear to contemporaries in other ways – through a document relief, the physical form of the inscription, the deposition of the monument in a sanctuary or, indeed, rituals and references embedded in the process of making the public enactment which the document recorded.Footnote ⁷⁶ In consequence, the decision to inscribe this heading (the cost of which would have been effectively nil) – whether made by the secretary commissioning the decree or perhaps even the mason inscribing it – may not have been particularly marked.

The desire to make specific reference to the gods does not seem to have been the primary driver behind decisions relating to the inscription of particular decrees, or their adornment with a document relief or the theoi heading. Nonetheless, the fact that references to divine agency were made – in some cases in pursuit of other ends – was not accidental, and nor was the way in which they were made idiosyncratic. These practices clearly reflect a strong sense of what was appropriate (especially in terms of the iconography used), which suggests that they reflected a wider understanding of the relationship between public authority and the gods.

As above, so below: divine determination

The Athenians honoured Proxenides, Athena honoured Proxenides: one of the readings which this juxtaposition of equivalent actions arguably invites is to explain the relationship between the two in terms of causation. From the modern viewer's perspective the Athenian decree preceded and is necessary for explaining the existence of the relief. The physical layout of the monument, however, suggests a different causal connection between the two coinciding acts. The visual implication, on this reading, is that the divine action, depicted above, preceded and determined the human action, presented below. In the case of the Proxenides monument, according to this deterministic reading, the Athenians honour Proxenides because or as a result of the fact that Athena honours him and intervened in some way to bring about this result on the mortal plane. By the same token, in the case of the divine dexiosis monuments, the friendly relations between Athens and other cities would be understood as being rooted in an act of friendship between Athena and the tutelary deity of the other city in question.

On the face of it, this sort of deterministic reading of these monuments sits uncomfortably with the emphasis in the public decrees they bear, on the agency of the popular assembly in making these specific decisions. Decrees at Athens and elsewhere began with the confident statement ‘it was decided by the council and the assembly …’ and often note who was formally responsible for proposing a particular motion. Literary depictions of the gods, such as the Homeric poems, Athenian tragedies and, indeed, Herodotus’ Histories, however, suggest a framework for understanding how these two different levels of determination could co-exist, a phenomenon which modern scholars know as ‘dual motivation’ or ‘double determination’.Footnote ⁷⁷ According to this model, a complex range of decisions and causal factors which were understood as operating on both human and divine levels coincided to produce a particular result, for which moral responsibility could be represented as being borne by actors at both levels.

Divine determination within these genres occurs through a variety of mechanisms. In the Homeric poems, in particular, gods are recognised as being able to intervene in both the physical and the mental worlds of men, physically restraining them, instructing them and implanting emotional states and intentions, to achieve particular outcomes which they personally desire within overarching deterministic frameworks of ‘fate’ and/or ‘the plan of Zeus’. For example, Agamemnon represents his actions in dishonouring Achilles as the result of ate, a stubborn blindness inflicted upon him by Zeus, and Achilles implicitly accepts this, but this does not negate the moral responsibility of Agamemnon and thus the necessity of compensation (Iliad 19.85–144). In other cases, mortals are endowed with menos by the gods, a kind of inner, mental force which enables mortals to do extraordinary things on and off the battlefield (Dodds Reference Dodds1951, 8–10). The Homeric epics also offer direct models for understanding the intervention of gods in the collective political institutions of men – by causing an assembly to be called, either by inspiration (Hera: Iliad 1.55) or suggestion (Athena in disguise: Odyssey 1.272–3) and, in one case, by determining the outcome of an assembly by influencing the psychological state of the crowd (‘Pallas Athena took away their judgement’: Iliad 18.311). The clearest expression of the idea of double determination in relation to an assembly, however, comes in Aeschylus’ Suppliants, in which an assembly of the Argives to decide the fate of the Danaids is reported by Danaos. The description in general emphasises the agency of the Argive people in agreeing the proposal put to them by their king, Pelasgos, but right at the end it is also described, in closely echoing language, as an event determined by Zeus:

Evidence of similar modes of thought regarding divine intervention can be identified in the surviving Athenian oratory, contemporary with the document reliefs (Dover Reference Dover1974, 136–41; Mikalson Reference Mikalson1983, 18–26). Speakers in the courts regularly talk about the direct intervention of gods in the mental and emotional processes of individuals, especially with a view to bringing about the destruction of offenders. In one particularly convoluted example, the orator Andocides refutes the claim of his accusers that the goddesses, seeking his destruction, dazzled him and thus caused him to place a suppliant-branch on the altar in the Eleusinion (an act which would have been in contravention of Athenian law and an indication of his own guilt). Andocides’ counterargument reveals the inherent flexibility of this discourse of divine intervention, in which he argues that, had the goddesses done this, they would certainly have caused him to admit the offence, whether or not he had committed it, when the herald asked who was responsible (Andocides 1.113–14). Elsewhere, divine intervention is credited, more positively, with inspiration (Isocrates 5.149–50). In other contexts orators also talk more generally in terms of an anonymous god (theos) or gods (theoi) intervening to bring about particular outcomes in a similar way to the human speakers of epic (Mikalson Reference Mikalson1983, 66–8; Parker Reference Parker2011, 65). In these cases, divine intervention is represented as a reality in the lives of individuals, but the particular agent or means is, in most cases, beyond human knowledge.

In one speech transmitted in the Demosthenic corpus, the speaker envisages direct divine involvement in the legislative processes of the Athenians as a matter of course:

This remarkably strong emphasis on the divine origin of law is exceptional in Attic oratory and is perhaps another indication, alongside the many recently marshalled by Edward Harris, that this speech is itself a Hellenistic imitation (Harris Reference Harris2018, 193–7; but see Martin Reference Martin2009, 183–202). The understanding of law attested here is, however, well paralleled in philosophical sources and narratives of law-making and law-makers (Harris Reference Harris2006, 51–3; Mikalson Reference Mikalson2010, 224, 227–8; Willey Reference Willey, Eidinow, Kindt and Osborne2016). What the passage usefully highlights is the different levels of agent which could be represented as responsible for creating a law and the way in which each layer of authorship lent it authority: beginning with the gods (of course) before proceeding to ‘wise’ human proposers and the collective will of the polis which formally instituted it.Footnote ⁸⁰ The document reliefs, on this interpretation, could be read as reflecting a belief that Athena was the prime mover responsible for the decision of the assembly.

Looking up to the gods: imitation

Another way of viewing these monuments is to begin with the inscribed decree on the lower part of the stele, which generally takes up the majority of the stone and is present even when the decision has not been taken to add a sculptural relief. The decree, as we have seen, emphasises the initiative and agency of the assembly itself – so what is the relationship between the two, if this element is foregrounded? In theological terms, the causal arrow cannot simply be inverted – it is difficult to see how the equivalent action of the goddess in honouring Proxenides could be determined by and subsequent to the action of the human assembly. The action of the deity ought to retain a primacy over, just as it is depicted above, the equivalent human action. However, instead of determining the human decision, in this reading agency remains with the assembly and its decree is to be read as imitating the model – both general and specific – provided by divine action, as depicted on the relief.

This reading works best for honorific document relief monuments. The key point is that honorific decrees themselves are always represented as being reactive – as responding to benefactions performed by individuals and other communities. For those of the second half of the fourth century bc the decrees regularly include a clause which gives the following as the official reason for their promulgation and, sometimes specifically, publication: ‘so that all may know that the demos of the Athenians gives great thanks for those who perform benefactions’.Footnote ⁸¹ The point is that there was a strongly felt moral obligation, on the part of recipients of benefaction, to reciprocate with appropriate thanks, and the decrees are inscribed as a way of communicating the fact that the Athenians comply with this obligation in this specific case, but they also, with their stereotypical and often non-specific descriptions of the honorand, communicate a paradigm of responding to benefactions in general.Footnote ⁸²

The divine imagery would arguably function in a similar way in this reading. In the case of the Proxenides monument, the logic would be as follows: Athena can be represented as honouring Proxenides, because that is what reciprocity demands as a result of his services for the polis of which she is the tutelary deity. Indeed, within this framework, services for Athenians can be interpreted as motivated by a desire to honour their goddess, Athena, and thus as services for her to respond to. Honorific decrees at Delos and Delphi in particular frequently link services performed for the polis by benefactors to their reverence for the god or shrine with which the polis was closely associated (e.g. FD III 2 72; IG XI 4 558), and there are also a number of cases where reverence for particular deities is clearly signalled on the part of proxenoi who are attested making dedications to the gods of the city in question (Mack Reference Mack2015, 10 n. 23). The gods in this case would be read as paragons or exemplars of moral behaviour – an approach which has less in common with the complex, vengeful characters of epic and tragedy than with the understanding of gods as perfect moral agents more common in popular and philosophical discourses (Dover Reference Dover1974, 78–80; Mikalson Reference Mikalson1983, 58–62; Reference Mikalson1991, 208–41; Parker Reference Parker, Gill, Postlethwaite and Seaford1998).

The Proxenides relief, with its non-specific portrait of Proxenides, ties in with the stereotypical language in which the decree describes his actions – and similarly communicates a paradigm of honouring those who, like the figure depicted raising his hand up before Athena in a gesture of adoration, benefit Athens and honour the goddess through their actions. On this reading the action of the Athenian assembly in honouring Proxenides is presented not as a direct consequence of Athena's intervention so much as a deliberate decision to imitate the act which she, as a paradigm of moral behaviour, could be expected to have performed. Even on this reading, however, the presence of this general and specific divine paradigm still has the effect of investing this monument and the decision of the Athenian assembly with divine authority.

Praying to the gods: ritual invocation

There is also a third reading, however, which would place a greater emphasis on the agency of the Athenians in initiating the decree for a benefactor than the first reading, but also a stronger emphasis on the active intervention of Athena in decision-making at Athens than the second. This would be to read the relief as depicting divine action as a result of a specific appeal made by the Athenian political community – of prayers before the assembly. The starting point for such a reading might be the heading, theoi, if Geoffrey Woodhead and others are right to believe that it made a reference to such prayers (or, indeed, if it was a reference to divine action resulting from such prayers). This heading, where it occurs on the document relief stelai, serves as the clearest point of contact between the texts and images on these monuments.

The assembly prayers, which we know from an extensive parody in Aristophanes’ Thesmophoriazusae, set out a series of different ways in which the gods were entreated to intervene in the Athenian assembly, influence the result and police the conduct of its participants.Footnote ⁸³ From this material it is clear that, after a purificatory sacrifice of a pig, attested elsewhere,Footnote ⁸⁴ the gods were entreated to be present at the assembly:

They were also, however, specifically entreated to intervene in two ways in the process of the decree – to ensure that the assembly itself was conducted properly, and with results that were for the benefit of the city, and that the speaker who gave the best advice be victorious:

The second of these provisions is particularly interesting. In this prayer, the assembly, like the battlefield or athletic games, is envisaged as a competitive arena of complex human interaction in which there is room for direct divine intervention to ensure the victory (nike) of one combatant, for example, by bestowing particular force (of persuasion rather than arms) on one participant.Footnote ⁸⁷

The first of these provisions, however, is also significant, and it is echoed in the request which the chorus expresses at the end of its lengthy invocation of the gods: ‘may we well-born women of Athens hold our assembly teleōs’ (Thesmophoriazusae 329–30). Here the adverb (translated as ‘faultless’ by Jeffrey Henderson in the Loeb edition) connects the idea of ritual correctness, emphasised by the context of the prayer, with procedural correctness, political authority and finality – and suggests that the assembly could be conceived of as a ritual as well as an institution for reaching collective decisions.Footnote ⁸⁸ Why should it matter that the assembly was conducted well, if the god was capable of ensuring that the correct decision was made regardless? The point is presumably that assemblies could be believed to operate under the guidance of the gods as long as the humans involved played their part honestly and diligently, and as though acting independently. In this context it makes sense that the surviving speeches should make infrequent reference to the gods, as certain knowledge of the gods, their actions and desires, could not plausibly be claimed, especially in speeches intended to persuade (Martin Reference Martin2009; Reference Martin, Eidinow, Kindt and Osborne2016). The divine framework did not replace, but operated externally, alongside the human institution, and, in fact, relied upon the latter's proper working, because there was always the possibility that it could be corrupted by the improper behaviour of human participants. This is illustrated by the other element of the prayer: the curses uttered against malefactors (the most heavily elaborated element in Aristophanes’ parody) amongst whom were listed ‘those who deceive’ as well as ‘those who plot against’ the city.Footnote ⁸⁹ As Alan Sommerstein has shown, this served as a practical alternative to individual oaths which could not have been administered to the many participants of the assembly, and, unlike the other elements of the prayer, invoked the intervention of the gods after the assembly proper, as a threat to ensure that it was held ‘faultlessly’ by participants (Sommerstein in Sommerstein and Bayliss Reference Sommerstein and Bayliss2012, 47–56).

There were also some direct communications from the gods that were looked for and acted upon in relation to public assemblies – in the entrails of sacrificial animals and certain meteorological phenomena (Parker Reference Parker2005, 100–1). In this case, it seems that these forms of divination concerned the question of whether it was propitious to hold a meeting – that is the absence of unpropitious signs, such as rain – rather than in relation to particular proposals, but this could nonetheless be taken as a form of positive guidance by the gods.

On this reading, the Proxenides relief depicts the ideal (and expected) result of this prayer, always supposing that it was completed in a ritually appropriate way and that the subsequent assembly had not been corrupted by malign participants – the normal explanation for decisions taken by the demos which turned out to be bad ([Xenophon], Constitution of the Athenians 2.17; Parker Reference Parker and Pelling1997). Specifically, it depicts Athena present, as requested, and the decree honouring Proxenides as the result of her benign influence, a manifestation of her will. This reading of the monument as depicting the Athenian decree as a kind of epiphany may be highlighted where the theoi heading occurs alongside the document relief, which perhaps invites the viewer, reading it aloud, to exclaim theoi, addressing the gods depicted above.

Variation and theme

The importance of the possibility of filling the gap between documents and reliefs in different ways is illustrated when we look beyond the honorific decrees to other kinds of public document bearing these reliefs. In these cases, some of the different readings of the relationship presented by the document relief stelai seem to work better than others. So, for example, for the document relief monuments involving interstate relations and dexiosis between gods, the determination and invocation readings make better sense than the imitation reading. The idea that relationships between communities might be determined by relationships between their respective patron deities, or that they would be invoked to provide guidance in such a context, fits easily within the wider theological frameworks which I have discussed. It is a little harder to see how treaties or even interstate honours could be thought of as imitating the gods as moral exemplars of behaviour, given that such documents inevitably also suggest interstate conflict and warfare. By contrast, when the same scene is deployed (briefly) in relation to inventories of sacred property, the suggestion that the treasurers of Athena and the other gods might be presented as seeking to imitate an inherent divine order – in the case of the first preserved example, the friendship of Athena and Demeter – makes good sense, especially in the context of the then recent reconciliation between Athens and Eleusis which had resulted in the transfer of the Eleusinian goddesses into the care of these officials.Footnote ⁹⁰ However, while ritualised appeals to the gods for guidance were surely made by the treasurers, permitting the third reading in this case as well, it is harder to read their administrative activities as likely to have been seen as determined directly by the gods.

More generally, the possibility of these parallel readings which was enabled by this juxtaposition of human and divine actors arguably allowed these monuments to function effectively in an intellectual context in which, while divine agency was for the most part an accepted fact of human life, uncertainty was necessarily built into discourse about it at every level. Platt's comparison of the equivocating rhetoric of dedicatory reliefs with multi-stable images – images which simultaneously depict, for example, either an old woman or a duck – is equally apt for the document reliefs (Platt Reference Platt2011, 36–41). In both cases, interpretation of the relationship presented involves making a choice between distinct alternatives but also recognising that multiple readings are possible (Mitchell, W.J.T. Reference Mitchell1994, 45–57).

Nonetheless, although the ways of reading these monuments outlined here have different starting points and explain the connection between relief and document in distinct ways, they do converge to assert what seems to us a very marked claim about the role of the gods, but which to contemporaries seemed, on the basis of this imagery, to have been a public orthodoxy: whether they suggest that the decree was caused by direct divine intervention, was made in imitation of divine action or was the result of an effective prayer for the guidance of the gods, the effect of each of these readings is to assert that the human decisions of Athenian democracy were underwritten in some way by divine authority.

Whether this understanding was, for any citizens, a deeply held personal belief, we cannot say. Indeed, it seems most likely that viewers would have been prompted to engage in this process of ‘reading’ the relationship between the assembly's decision and the gods in individual cases where they were inclined to disagree with the wisdom of a particular decree, because of opposition at the time and/or the benefit of hindsight. In such cases, the different ways of reading the relationship between relief and document would have provided space for downplaying the strength of divine intervention in a particular case or even, through reference to human corruption, making a complete exception of it. To account for the production of this material, however, it is necessary to posit this kind of understanding of the relationship between public authority and the gods, and specific ways of making sense of it, an understanding which was generally taken as true for the purposes of running the polis. It is, in any case, closely paralleled by – or, better, an expression of – a wider, well-documented article of Athenian public faith, that Athens was particularly dear to the gods (which, from the start, entailed the essential corollary, that it was harmed by corrupt political leaders).Footnote ⁹¹

Standing en masse on the Acropolis, with their eye-catching polychrome sculptures amidst the forest of other documentary stelai and votive dedications, the document relief stelai would have offered viewers a continually repeating performance of the relationship between human authority and divine authority – expressing lapidary certainty in the existence of a link, but allowing for a necessary uncertainty, given the limits of human knowledge, about its precise nature in particular cases.