Storicamente. Laboratorio di storia

Studi e ricerche

Innovation and Resilience: Post-Disaster Architecture in Fourth-Century Delphi


In the 370s BC, Delphi was struck by a natural disaster. The temple of Apollo itself was severely damaged and needed to be rebuilt almost entirely. A few years later, the Third Sacred War (356-346 BC) interrupted the reconstruction. These two events must have been traumatic, yet the reconstruction of the temple of Apollo started relatively quickly after the catastrophe of the 370s and resumed swiftly after the end of the war. In addition to the temple of Apollo, other monuments were erected in the aftermath of these events. This paper focuses on the temple of Apollo, the Treasury of the Thebans and the so-called Limestone Temple and offers an interpretation of their architectural features in light of the distressing events that preceded them, using the concepts of anchoring innovation and conspicuous consumption. Resilience, it is argued, is not only evident in the vast funding scheme for reconstructing the temple of Apollo, but also in the design and construction of the monuments erected during that period.


The fourth century saw a number of distressing moments and periods for Delphi and its sanctuaries of Apollo Pythion and Athena Pronaia. In or around 373 BC, Delphi was hit by a major natural catastrophe that ruined a number of its monuments, including the temple of Apollo (SD 422) [1] itself. The event must have been traumatic, but the sanctuaries of Delphi evidently recovered: ruined monuments were repaired and new ones were built, or started to be built shortly after the catastrophe. Only a few years later, however, Delphi experienced a distressing event of another sort, as the Third Sacred War began in 356 BC. The Phocians (the people of Phocis, the region surrounding Delphi), having gained control over the sanctuary of Apollo, melted down many of the offerings in order to pay their mercenaries. The war also interrupted the reconstruction of the temple of Apollo. However, the works resumed quickly after the conflict, and other buildings were erected in the second half of the century.

Taking these two traumatic events as turning points in the history of both sanctuaries, this paper focuses on the architectural features of a series of monuments erected during that time, in order to explore in what ways their design responded to the traumas that preceded them [2]. This investigation does not intend to contradict or challenge the rich literature produced for more than a century over Delphi, but rather to suggest another perspective, centred on the appearance of some of the buildings erected in the course of the fourth century. The epigraphic corpus related to the building activity of that time offers invaluable evidence that helps us reconstruct the administrative, financial, technical and practical aspects of monumental construction at the sanctuary of Apollo in the fourth century. The building accounts have been untangled by a series of scholars, notably Bousquet (1988a, 1989), and their potential for the interpretation of the architectural remains has been demonstrated in numerous articles and in the monograph on the fourth-century temple of Apollo by Amandry and Hansen (2010). However rich these inscriptions are, they only partly explain design choices. In order to better understand aspects of design, therefore, other approaches must be used in combination with the textual sources. Using the concepts of anchoring innovation and conspicuous consumption for framing the analysis, this paper aims at showing how the monuments built after distressing events testify, in their visual appearances, to forms of resilience.

Setting the scene: from one upheaval to the other

The natural disaster that struck Delphi is not explicitly recorded in any literary source, but it is usually equated with an earthquake that happened in 373 BC and had devastating consequences on the northern shore of the Peloponnese. That year, the city of Helike disappeared, perhaps engulfed by a massive tidal wave [3], and the earthquake also affected the city of Bura, higher up in the mountains [4]. While the latter survived, the former never recovered.

At around the same time, on the other side of the Corinthian Gulf but some 43 kilometres only from Helike, the sixth-century temple of Apollo, which had been funded in large part by the Alcmeonids, and perhaps other monuments, were damaged. These damages seem to have resulted from a landslide, but what caused this landslide, perhaps an earthquake or heavy rainfalls, remains uncertain. The only thing that is absolutely certain is that the collection of financial contributions for the reconstruction of the temple began at the latest in 366 BC (Bousquet 1988a, 24; CID II, 6). Because of the chronological proximity with the 373 BC disaster, it is often assumed that Delphi was hit by the same earthquake as the one that struck the other side of the Corinthian Gulf. However, the literary sources only mention Helike and Bura, while only one inscription might make a reference to the destruction of the temple (Syll. 3 295; Jacquemin, Mulliez, and Rougemont 2012, n° 36). The catastrophe must have happened before the construction of the Treasury of the Thebans (SD 124), since the latter rested on a repaired portion of a wall that was likely destroyed during the landslide, and since this repaired portion and the northern part of the Treasury’s foundations seem to have been made of reused blocks from the Alcmeonid temple of Apollo (Michaud 1973, 20-1). The Treasury, says Pausanias, was built from the spoils of the Battle of Leuctra, which saw the victory of the Thebans (Paus. 10.11.5). It is therefore generally assumed that it was erected shortly after 371 BC, although “shortly” could be less than one year or several years. A perhaps more compelling, but still debatable piece of evidence is a passage of Xenophon’s Hellenika where the Spartan Prothoos suggests that various cities make contributions εἰς τὸν ναὸν τοῦ Ἀπόλλωνος (Xen., Hell. 6.4.2). The episode dates from 371 BC and has been taken as evidence that there was a plan to rebuild the temple at that time. However, it is not certain that it refers to the temple of Delphi, nor to a reconstruction at all [5].

Whatever provoked the damages at Delphi, its impact must have been high: at the very least, retaining walls were swept away and the south-western corner of the terrace supporting the temple of Apollo collapsed, along with parts of the foundations of the temple. It is unlikely that the whole temple fell apart. In fact, it probably stopped much of the rubble hurtling down from above, shielding a significant part of the lower area of the sanctuary. Indeed, the monuments that stood there do not show any trace of major destruction or repair [6]. Nonetheless, the temple was too badly damaged to be repaired; whatever remained standing was dismantled and reused where possible [7], and only the major parts of the foundations were kept as a starting point for the reconstruction (Amandry and Hansen 2010, 161-3). Numerous buildings were damaged, mostly in the northern and western parts of the sanctuary of Apollo (Bousquet 1988); the nearby sanctuary of Athena Pronaia probably suffered some damage too (see below). Part of the reason why it is so difficult to assess the magnitude of the destruction at Delphi is the fact that the sanctuary of Apollo eventually recovered completely. Ruins were cleaned off, buildings were repaired or rebuilt, and new monuments were erected, some of which very shortly after the accident.

The Amphictiony set up a vast and original funding scheme for the reconstruction of the temple. Voluntary contributions were sought from states and individuals, and each individual of the member states had to pay one obol (CID II, 1-30). This first per capita tax probably did not bring in nearly enough money, and a “second obolos” was therefore introduced after five years, but does not seem to have been much followed and was probably not very effective (Bousquet 1989, 20). Fortunately, the naopoioi (the international board in charge of overseeing the reconstruction of the temple) also benefitted from a line of credit opened for them by the city of Delphi, and to be used exclusively for the reconstruction of the temple (CID II 31-32). The funding of the reconstruction rested, therefore, on a variety of sources, some compulsory, other voluntary [8].

The first payments were received in the spring of 366 BC [9], seven years after the destruction of the temple if the date of 373 BC is retained. This is not necessarily a long time, considering the complex political framework and what needed to be done beforehand: determining the extent and cost of the works, removing the debris, preparing and agreeing on a funding plan and, finally, making it known to the member states and to the rest of the Greek world. Some work, one could think, might have been done before the first payments arrived with the sanctuary’s own cash reserve, which would have included (in addition to the activities already mentioned above) the securing of the sanctuary for its visitors [10] and prospections for new quarry sites (Hansen, Algreen-Ussing, and Frederiksen 2017, 214-5). If one also considers that the temple of Apollo was one of the largest of its kind at the time and that the sanctuary was located in a remote, mountainous location, the initial steps taken for its recovery seem to have been up to the challenge.

However, the reconstruction of the temple was soon interrupted by the outbreak of the Third Sacred War (356-346 BC). While some important works were still going on shortly after the start of the war, in 356, the naopoioi did not meet in 355 and 354. In 353, special “war naopoioi” were called upon, and some activity went on until 351. From then on, the works were put on hold until the end of the war [11]. The Phocians, who had taken over the sanctuary of Apollo, melted down numerous votive offerings to finance their war against the Amphictionic states. Shortly after the end of the war, in 346, the works at the temple of Apollo resumed. The Phocians were condemned to reimburse the money that they had stolen from the sanctuary by melting down the offerings. This influx of cash money was probably seen as a windfall: the wealth stored in the sanctuary, unusable until then, had been turned into cash and could now be injected into the construction of the temple and other projects [12]. The Amphictions, in sum, made the best of a seemingly bad situation. The temple could be finished and was adorned with sumptuous marble decoration, and other new buildings were erected.

Scope and concepts used in the investigation

The literature about fourth-century Delphi and the reconstruction of the temple is extremely abundant and reflects more than a century of discoveries, interpretations, reinterpretations and debates, some of which are still unresolved. It is well beyond the scope of this paper to account for all this scholarship. Rather, this contribution aims at exploring, from an architectural perspective, a period marked by the destruction of the temple of Apollo and the aftermath of the Third Sacred War, that is between ca. 373 BC and the second half of the fourth century. Three buildings in particular, which were each financed by different sources of funding, will be addressed: the Treasury of the Thebans, the so-called Limestone Temple (SD 43) and, of course, the new Temple of Apollo (SD 422) [13].

The Treasury of the Thebans was built shortly after 371 BC, which makes it arguably the first monument to have been completed in the sanctuary after the disaster of the 370s. It was located in the south-western corner of the sanctuary of Apollo, in an odd, yet conspicuous position on top of the southern section of the peribolos wall. The so-called Limestone Temple is located at the western end of the sanctuary of Athena, about 500 metres east of the sanctuary of Apollo. Both the Treasury of the Thebans and the Limestone Temple are almost exclusively built from a greyish limestone coming from the nearby Profitis Ilias quarries (about 5 kilometres west of the sanctuary of Apollo). They also share striking similarities with regard to their technique and formal purity. These affinities have led researchers to date the Limestone Temple to shortly after the Treasury of the Thebans, around 365-360 BC [14]. However, recent investigations in Marmaria, the site of the sanctuary of Athena Pronaia, concluded that the building must have been erected in the second half of the century. Furthermore, according to the same study, the building was not a temple but, rather, a meeting place [15]. The reconstruction of Temple of Apollo started in the 360s. Its euthynteria and krepis were made of the same limestone as the one used for the Treasury of the Thebans and the Limestone Temple. These three buildings, this paper argues, attest to both continuity and change in the architectural projects that immediately followed the disaster of the 370s and the Third Sacred War. From the perspective of building activity, it can be considered a period of resilience as well as innovation.

In the field of archaeology, the term “innovation” is usually distinguished from the word “invention”. An invention is the process by which something completely new is created, while an innovation is the successful adoption of an invention in a variety of contexts (Sluiter 2017). Unsurprisingly, in the context of Classical Greece, inventions are difficult to detect unless clearly labelled as such by ancient sources. The recorded material evidence, indeed, can rarely be considered sufficient alone to determine exactly when and where things were invented, as there will always be a suspicion that there was a predecessor to the earliest recorded attestation. Furthermore, “inventions” and “innovations” are not concepts that exist outside a specific socio-cultural context; they are a matter of perception, and what we might see, in retrospect, as an “invention” could have just been considered, at the time, as a natural step in the evolution of traditional forms and techniques. For these reasons, this paper takes a cautious approach by only referring to novelties as “innovations”, even in the cases of forms or techniques that are seemingly unprecedented in the Greek world.

The rest of this article is divided in three parts. First, I will discuss how constraints can foster or, on the contrary, hamper innovations and how this is illustrated by the architectural works in the sanctuaries of Delphi during the period under consideration. Second, I will discuss the ways in which such innovations were anchored in tradition. The concept of “anchoring innovation” allows one to overcome the tension between tradition and innovation as it acknowledges that they are, so to speak, two sides of the same coin. Not only do innovations only exist within (or in contrast to) a set of traditions; in order to be successfully adopted, they also need to be embedded and anchored in a preexisting, familiar system of values or information. This system can be described as a “system of associated commonplaces”, a concept coined by Black, which was fruitfully applied by Ferrari to Greek sanctuaries as built spaces (Black 1962; Ferrari 2006). Only when anchored in such system of associated commonplaces will the innovations make sense, both for the people that actively implement them and those who experience them as receivers. Finally, I will suggest an analysis of the reviewed buildings using the more familiar conceptual framework of conspicuous consumption and, more specifically, Trigger’s notion of “thermodynamics” according to which monumental buildings are the expression of a conspicuous and deliberate waste of energy (Trigger 1990). I will argue that, in the present case, the lack of conspicuous consumption is as important as conspicuous consumption itself, if not more. Ultimately, I will suggest an interpretation of the architectural features of the reviewed buildings in light of the events that preceded them.

Constraints and innovations

While the diffusion of innovations is a much-discussed concept in archaeology [16], the causal factors that explain their adoption in specific contexts have been little discussed. The buildings reviewed in this paper allow this issue to be addressed by analysing both their appearance and techniques. Of course, the role of the architect cannot be underestimated. Both the Treasury of the Thebans and the so-called Limestone Temple attest to an influence of mathematics (Bommelaer 1979), which manifested itself in a constant effort to attain the purest shapes and proportions. The round Tholos (SD 40), located in the sanctuary of Athena, was the most striking illustration of this effort (Bousquet 1993). The incorporation of ionicizing elements, noticeable in the Limestone Temple, was in line with an old trend recently revived by Attic architects. However, architects were only completely free on paper. They had to deal with the built and natural environment, as well as with material and financial constraints.

Constraints have a lot to do with innovation. The interaction of the two has been mostly studied, so far, in modern business and entrepreneurship studies. In these contexts, constraints are usually seen as obstacles to innovation that need to be abolished, and most studies exhort entrepreneurs to remove as many constraints as possible to favour innovative thinking among their associates and employees. However, a more recent trend in the same field of research also claims that constraints can foster innovation. Acar, Tarakci, and van Knippenberg (2019) reached that conclusion after reviewing no less than 145 empirical studies on the subject. Of course, a lot of their observations apply to a modern capitalistic company setting where innovation is considered to be inherently good and is a goal in itself. Innovation and change in general were not always perceived in such a positive way in Ancient Greece, even if there seems to have been a shift toward a more welcoming view on innovation in Athens in the fifth century BC (D’Angour 2011; Sluiter 2017).

It is outside of the scope of this paper to elaborate a complete theory of constraints and their effects on architectural innovation in the Greek world. Architectural constraints and the responses of the Greek architects have already been well studied (Coulton 1977). The objective here is merely to highlight the fruitful (rather than conflicting) relation between constraints and innovation by stressing some basic elements. Constraints can be external or internal, that is, either dictated by the environment or self-imposed. However, one should not draw too sharp a line between the two: for instance, a given physical environment, presumably an external factor, only becomes a constraint if one decides to build something there in the first place. Therefore, external constraints are always, to various degrees, also self-imposed. These constraints can lead to three archetypal reactions: making concessions (which can consist in lowering the initial standards of the project, or more radically, in abandoning it altogether), adapting, or innovating. In this perspective, innovation is understood as a form of adaptation with, in addition, a disruptive aspect with regard to tradition.

The issue of space, a major constraint in architectural projects, was crucial for all three buildings under review in this paper. More specifically, the lack of space (sometimes combined with rather unstable building grounds) arguably led to concessions in some monuments, and to adaptive innovations in others. This spatial constraint explains in part the audacious position of the Treasury of the Thebans, directly above a ca. 4.5 metres tall peribolos wall. To some extent, when seen from the south, the foundation courses could be considered part of an uninterrupted elevation running from the bottom of the temenos wall to the top of the south wall of the Treasury. While such a position ensured that the Treasury was as visible as possible from every viewpoint and probably caused the awe of the visitors, it was not chosen specifically for the technical challenge that it represented. More mundanely, the sanctuary was already crowded with monuments at the time, and it was probably one of the few spots available for a building of that size. This specific location was probably picked, among the available ones, because it was perceived to be important within the sanctuary. Indeed, the Treasury was located at a crossroad where the paths coming respectively from the west and east entrances met and where the ascending path toward the temple began. Nonetheless, the building was facing west, not east, as used to be believed, with its entrance toward the western entrance of the sanctuary rather than toward the other treasuries (Jacquemin and Laroche 2012, 107-8). Therefore, while being located in a cluster of treasuries, the Treasury of the Thebans was turning its back to them, setting up its own route starting at the western entrance of the sanctuary. Its position also had a political meaning, as it cleverly mirrored that of the Spartan monuments, near the east entrance (Scott 2017). However, the impressive superimposition of the wall and the Treasury would not have been necessary, had the Thebans built a slightly smaller monument. Instead, the treasury they chose to make was the largest (in terms of plan dimensions) of all the treasuries of Delphi [17]. The width of the building was, therefore, a self-imposed constraint. Another constraint was the stability of the terrace, which had proved to be insufficient in the face of the landslide that had swept away a portion of the peribolos wall.

This set of constraints (location, size, stability) was dealt with by adaptive innovation. One of the most striking technical features of the Treasury of the Thebans are the channels cut in the upper surfaces of the two upper courses of the foundations. The channels, about 10 cm wide and 7.5 cm deep, served to accommodate a framing that ran along the foundations, with crossings at the angles. This unusual feature has been much discussed since its discovery. For a long time, the opinio communis was that the framing was made of iron, following an influential article by Dinsmoor on the use of iron reinforcements in ancient Greek constructions (Dinsmoor 1922), but Michaud convincingly demonstrated, in his architectural study of the treasury, that it was more likely made of wood (Michaud 1973, 24-5). He insisted on the absence of any trace of oxidation or lead in the channels and pointed out that iron was quite expensive and required advanced metallurgical techniques. The long beams, it should be added, were about 13 metres long, and it is doubtful that iron beams of such length would have been financially or technically affordable at the time. Wooden beams of that size, on the contrary, were probably all too normal, as is attested by the numerous examples of extremely long wooden pieces used in ancient architecture (Meiggs 1982, passim). The purpose of this double wooden framing has never been in dispute: it served to reinforce the cohesion of the foundations. However, the causes of disruption that the builders had in mind were subject to several interpretations: while some claimed that the main concern was water torrents, others, like Michaud, have stressed that such seasonal torrents are hypothetical and that the main concern must have been the seismic activity, combined with differential settlements (Michaud 1973, 24). In any case, it is the conjunction of heavy constraints that prompted the original solution of the wooden framing [18].

Space was also a constraint in the case of the Limestone Temple, although with different consequences. The building was erected in the terrace of the sanctuary of Athena, at the western end of a series of preexisting buildings and next to the Tholos. The main innovation of the Limestone Temple is the prostyle hexastyle plan (that is, with only six columns at the front), an arrangement rarely seen before in the Doric order. According to Michaud, this innovative plan might have been a response to the limited space on the terrace. A more traditional, peripteral plan (that is, with columns all around the building) would have only been possible with a reduced cella, which would have been in contradiction with the intended effect of the interior design. Indeed, the cella was deliberately wide, with no interior colonnade, and it was largely open to the pronaos, with a broad three-bay entrance that ensured that a large amount of light could enter the room [19]. The prostyle hexastyle plan solution was perhaps inspired by Ionic temples or more directly, perhaps, by the Doric amphiprostyle hexastyle plan (with six columns at the front and at the back) of the Temple of the Athenians at Delos that was built around 425-420 BC. Interestingly, in the latter case, the abandonment of the peripteral plan in favour of an amphiprostyle plan might also have been prompted by the limited available space in the area (Winter 1980, 415). By keeping only the columns in the front, the Limestone Temple went a step further with this innovation.

The hexastyle façade of the Limestone Temple naturally implied a wider porch than with an in antis temple; the krepis steps, which ran all around the temple, were wider in the front and on the sides of the porch, which can be explained by their functional role as stairs. This arrangement gave the temple a T-shaped plan that would become standard in Doric architecture during the Hellenistic period. If Michaud’s conclusions are correct, the constraint of space might have had an important impact over the development of Greek architecture in the Late Classical and Hellenistic Period.

A similar issue of space arguably also played a role in shaping the new Temple of Apollo. After the partial destruction of the Alcmeonid temple, the first major step was to clean up the area by removing all the elevation blocks and rubble, in order to start the construction of a new temple (almost) from scratch. The south-western angle of the foundations appeared to have been badly damaged; they were exposed on the side, fixed and girdled with a strong masonry made of reused blocks from the Alcmeonid temple (Amandry and Hansen 2010, 161-3). The rest of the foundations, which had been spared by the catastrophe, was kept as they were and the new temple was erected on them. These preexisting structures made it more complicated to install the pavement of the peristasis, which perhaps prompted the use of a system rarely encountered in Greek architecture: T-shaped mortises were cut on the resting surfaces of the blocks, allowing them to be pulled and to make new attempts at setting them into place (ibid., 236-9). The constraint by which the old foundations had to be reused, therefore, might have elicited an innovative technique.

Another consequence, however, was that the plan of the new building was almost identical to the plan of the Alcmeonids. It was an elongated, archaic-looking plan, which contrasted with the usual proportions of buildings in the fourth century and, more importantly, with the temple’s own elevation. Indeed, with the exception of the plan, the new temple complied with the canons of the fourth century, such as the raised podium and the more slender column proportions [20]. Scholars have often qualified the plan as “archaizing”. However, one should not lose sight of the practical and economical constraints. The available space was quite limited in the early fourth century. At that time, the sanctuary was already crowded with buildings and the temple itself was surrounded with smaller monuments [21]. Of course, some of them had been ruined by the landslide and the administrators of the sanctuary took advantage of the situation to extensively reorganize the spatial arrangement of the area [22], but the terrace could not be extended much toward either the north or the south, as these areas were densely occupied. In theory, the terrace was probably large enough to increase the width of the temple toward the south. However, it would have considerably reduced the space for building operations. More importantly, the south-western corner of the temple would have been dangerously close to the edge of the terrace, in a particularly unstable area [23]. Concretely, a much larger part of the temple foundations would have rested on artificially built-up soil and the stress incurred on the terrace walls would have been greatly increased. Furthermore, a rearrangement of the foundations (either for decreasing or for increasing the size of the temple) would have implied considerable labour costs, and a widening of the temple would have been a supplementary challenge, as it would have increased the roof span (Coulton 1977, 74-96) [24].

In the cases of all three buildings, constraints pushed the architects and builders to come up with innovative solutions. In the case of the temple of Apollo, they also played (in part) against innovation. One has to keep in mind that the three buildings were very different in scale. The sheer size of the temple of Apollo, and the volume of materials involved in its construction added further levels of complexity and limited the range of possible solutions. However, while constraints did have an impact on architectural projects, deterministic explanations alone can never fully account for architectural features. The “Archaic” proportions of the new Temple of Apollo were not the only possible solution to the problem of space, and other factors than spatial constraints must be taken into account for explaining why this specific solution was retained.

Anchoring innovation and change

Despite the old-fashioned proportions of its plan, the new temple of Apollo, as already mentioned, presented features characteristic of fourth-century Doric temples. These features might have seemed innovative at the time; at the very least, they gave the temple a modern aspect that contrasted with sixth and fifth-century temples and, perhaps more importantly, with its predecessor, the Temple of the Alcmeonids. With regard to these changes, the plan of the temple, which remained virtually unchanged, appeared to the modern scholars as a contradiction and has received the label “archaizing”, without much discussion regarding what this supposed archaism meant at the time. We have pointed out that physical constraints could have played a role in the decision to keep the foundations as they were. However, constraints are not enough to explain why this specific solution was chosen. The Classical temple of Apollo Epikourios at Bassai, for instance, shows a similar continuity of the plan with its Archaic predecessor, in a much less constraining topographical context [25].

The term “archaizing” implies a deliberate attempt to give the new building technical features connoting a sense of “oldness”. However, one can reasonably assume that such a discrete reference to ancient canons of temple architecture only interested the specialists (namely, the architects and builders) and went unnoticed by most visitors. Their perception of the new temple must have depended largely on their mental references. Among all the temples with which the visitors could have mentally compared the fourth-century temple of Apollo, the most natural reference would have been its predecessor, the Alcmeonid temple. The dimensions of the building were probably, in the eyes of those who experienced it, a cue to its identity as the Delphic temple of Apollo (as a mental concept) rather than an obscure reference to archaic norms of architecture. The fact that the dimensions of the plan remained (almost) unchanged signalled to the pilgrims that this was the venerable, authentic temple of Apollo (and not an ersatz), an impression perhaps reinforced by the use of the Doric order. In this way, the innovative features of its elevation were anchored in a familiar set of shapes and volumes that participated in the visual identity of the sanctuary.

More concretely, and perhaps more importantly, the continuity of the plan ensured that the cultic, religious spatial arrangement inside and outside the temple remained unchanged. Ancient cults were embedded in space (Wescoat and Ousterhout 2012; Mylonopoulos 2006) and any significant change in their physical environment potentially jeopardized their very identity (Roux 1984, 162-6; Jacquemin 2017, 33-4). In this sense, the plan of the new temple of Apollo was not “archaizing”; it was, quite literally, “archaic”, as was the cult that it articulated. The importance of the preservation of the interior arrangement in relation to the cult is attested by a passage of the inscribed accounts where the Delphic entrepreneur Deinon built a mudbrick hoarding around the omphalos in order to protect it from his own construction works (CID II 49 A, l. 8-11). Amandry and Hansen also observed that the stylobate of the interior colonnade presented irregularities that they interpreted as a consequence of the special arrangement of the cella required by the oracular activities (Amandry and Hansen 2010, 72, 387). According to Roux, the architects took special care of fixing and protecting the “holy of holies” and, while the temple must have been inaccessible most of the time during the works, monthly consultations were still probably held within the temple (Roux 1979, 203-4). The perpetuation of the archaic plan of the temple of Apollo, therefore, might also be a reflection of the perpetuation of the cult. While the physical constraints reduced the number of possibilities, the weight of religious tradition might have motivated, or at least justified the decision to keep the old plan.

The anchoring of visual innovations in a system of “associated commonplaces” is also detectable in the case of the Treasury of the Thebans. From a visual perspective, the building was a mix of traditional and new. Its general shape could not be more conservative, since treasuries had been one of the main architectural forms in the sanctuary since the seventh century. In fact, it had been about a century since the latest treasury had been built. Rather than just another iteration of the shape, the Treasury of the Thebans was a revival of the form and a conscious allusion to the glorious past of the sanctuary.

By contrast, some features of the Treasury made it stand out from the other nearby monuments. The treasury was unmistakably Doric, but in an extremely sober way. In place of the usual distyle in antis, the façade was a plain wall only pierced by the entrance door and a small window presumably placed on top of the door lintel, a device designed for ventilating the prodomos that also acted as a sort of relieving triangle [26]. The walls were bare but very finely dressed, and show an expert carving of the local limestone from the Profitis Ilias quarry. No sculpture adorned the treasury, except perhaps on the angles of the roof. The ornaments were limited to the framing of the door and the Doric frieze of the entablature. The number of triglyphs and metopes was unusually large, which was made possible by the absence of columns.

The monuments in the close vicinity of the Treasury of the Thebans are not well known. On the other side of the road to the north were a wide niche (SD 230) and a Doric treasury (SD 226) that is usually considered to be Boiotian but could well be Phocian, based on the lettering of the names that were inscribed on the blocks. It seems that the treasury 226 was ruined before the niche was constructed, and therefore that it no longer existed when the Treasury of the Thebans was built (Bommelaer and Laroche 20152, 153-4). A little bit more distant was the Treasury of the Siphnians (SD 122), which was nevertheless close enough to the Treasury of the Thebans that a person standing in the centre of the crossroad could see both of them by just turning their head. The contrast between the two would have been striking. Indeed, the Treasury of the Siphnians was everything that the Treasury of the Thebans was not (and vice versa): Ionic, distyle in antis and, above all, exuberant in its ornamentation. In fact, the Treasury of the Siphnians was probably the most lavishly decorated treasury of the sanctuary, with its two caryatids, its vivid frieze with figures and its elaborate mouldings [27]. Any visitor could notice that, while both monuments belonged to the same category of buildings, their spirits were radically different, and the proximity of the Treasury of the Thebans to the Treasury of the Siphnians made the former stand out all the more (Perrier 2019, 75). At the same time, the fact that the Theban Treasury was located in an area already packed with treasuries reinforced its anchoring in the tradition of this particular shape of monumental offering. However daring and innovative the visual features of the Treasury of the Thebans were, they were anchored in the familiar, small temple-like treasury shape which, over the centuries, had become part of the set of visual “associated commonplaces” that conveyed the identity of the sanctuary.

The Limestone Temple resembles the Treasury of the Thebans in many aspects [28]. So much so, in fact, that they have usually been considered to be more or less contemporary and the work of the same team of builders. The stone, carefully selected in the Profitis Ilias quarry, is the same as the one used in the Treasury of the Thebans, and the carving and the dressing of the blocks attest to an expert craftmanship: the faces are either finely pointed, polished or worked with a toothed chisel, which creates a subtle contrast between the different elements of the building. The temple also shows a similar attention to the stereotomy and an analogous pursuit of mathematical perfection with regard to the proportions. The Limestone Temple, just like the Treasury of the Thebans, gives an overall impression of sobriety. It also possesses some refinements that are not attested in the treasury, like a batter of the walls and a domed base [29].

While offering an overall sober appearance, the Limestone Temple also introduces a number of innovations. These novelties are significant, but not particularly spectacular, as they are balanced out by a series of more traditional aspects [30]. We have already mentioned the innovative prostyle hexastyle plan, as well as the enlargement of the footing of the porch. This protruding was mitigated by the continuous krepis around the building. The krepis reinforced the coherence between the porch and the rest of the temple, which could have otherwise seemed only juxtaposed. Quite remarkably, the porch and the pronaos were not separated by any kind of support. The entrance of the naos itself was divided in three openings by two Ionic semi-columns attached to pillars, another innovative arrangement [31]. A further original feature is the shape of the capitals of the semi-columns, with the balusters shaped as calyx ends. The Ionic half capitals with calyx ends were widespread in the Hellenistic period; according to Michaud (1977, 111-2), the half capitals of the Limestone Temple, halfway between tradition and innovation, were precursors.

Both the prostyle arrangement and the three-bay entrance are adapted from Ionic standards. The introduction of Ionic elements in Doric architecture was not new at the time. In fact, the influence of the Ionic order can perhaps be traced back to the sixth century BC with the trend of more slender columns in Doric temples of the Greek West (Winter 1976). This tendency appeared in the mainland in the early fifth century, and was later picked up in Classical Athens, where other ionicizing traits, such as mouldings, columns and friezes, were implemented in Doric buildings (notably the Hephaisteion, the Parthenon, and the Propylaia; Winter 1978). The Athenians also introduced the hexastyle façade in the late fifth-century amphiprostyle Temple of the Athenians in Delos (Winter 1980, 415-6). The Limestone Temple, therefore, was part of a long-term trend of renewing the Doric style by adding to it some more or less subtle Ionic touches.

Yet some elements were unprecedented or, at least, had rarely been seen in Doric architecture up to that point, such as the hexastyle façade on one front only, the three-bay entrance to the naos and the wider porch. The prostyle hexastyle plan was perhaps inspired by the Athenian experiments like the Erechtheion and the Temple of the Athenians at Delos (Winter 1982, 397), or even the Propylaia of the Athenian Acropolis, which also already featured a wider porch. In the case of the Limestone Temple, this wider porch was required, so to speak, by the hexastyle front; the krepis steps were wider in the front and on the sides of the porch, which can be explained by their functional role as stairs. The triple entrance of the naos may have been inspired, according to Michaud, by the Temple of Athena Nike on the Acropolis, a “fully” Ionic building with a similar arrangement (Michaud 1977, 110).

However, the temple of Nike, being narrower and tetrastyle amphiprostyle, does not present anta walls on either side of the bay, as is the case in the Limestone Temple (Mark 1993, 72-5). With such features, the Limestone Temple prefigures the developments of the Hellenistic period, when some of its characteristics would become standard in Doric architecture: the prostyle plan largely replaced the more archaic peristyle plan and the so-called “T-Plan” that it induced became recurrent.

However, as innovative as it was, the Limestone Temple did not clash with the Doric tradition, quite the contrary. The robustness and austerity typical of the Doric buildings are here very much present, as we have seen: the precise stereotomy and the absence of any decorative flourish are well in line with the Doric classical aspirations. The angle contractions, typical of the Doric order, are still present in the Limestone Temple. Furthermore, while the columns are more slender than those of the older Doric temples, their proportions are very similar to those of the Doric buildings of the same period. The number of metopes, 10 at the front and back and 20 on each long side, is typical of the peristyle Doric temples of the fourth century; thus the Limestone Temple, while introducing the prostyle hexastyle plan in the Doric order, was in accordance with its contemporary peristyle counterparts with regard to its frieze and general proportions (Michaud 1977, 107-11). The Limestone Temple is therefore a striking example of innovation anchored in tradition: the new elements that it introduced, most of which were borrowed from the Ionic order, were embedded in a consensual Doric structure and arrangement. In this way, the radicalism of the new features was attenuated and made acceptable. This is not to say that the architect made a conscious effort of mitigating the new with the traditional, or that such a process was limited to the Limestone Temple, of course. Rather, this case exemplifies that the new cannot exist outside of tradition and that forms can evolve while still being perceived as traditional.

Resilience and moderation

As already discussed, the sanctuary of Apollo proved resilient on a financial level, despite the difficulties, both after the catastrophe of ca. 373 and after the Third Sacred War. After the destruction, a large-scale funding scheme was set up in order to rebuilt the temple of Apollo, and after the end of the Third Sacred War, the “fine” of the Phocians was used as a timely income of cash money that helped finish the building. However, a form of resilience can also be detected, arguably, in the appearance of the monuments themselves, including buildings that were financed by other sources of funding, like the Treasury of the Thebans and the so-called Limestone Temple.

Driessen identified three features that, when appearing in conjunction with one another, signal an “architecture of crisis”: a decrease of energy input in production and maintenance, a change of original function and a change of original plan (Driessen 1995). None of the above-mentioned buildings presents a combination of all three features, which suggests that the crisis was successfully overcome. It could be argued, however, that a decrease of energy input is noticeable, since two of the buildings (the Treasury of the Thebans and the Limestone Temple) are entirely made of a local stone, and do not present any elaborate sculptural element. However, the limestone from the Profitis Ilias quarries, while not as prestigious as marble, was far from being a cheap, default material. Its aesthetic and structural qualities actually made it a stone of choice [32]. As for the absence (or near absence) of sculpture on the two buildings, it is largely compensated, in terms of energy expenditure, by the extreme precision of the stereotomy and dressing of the walls. Therefore money does not seem to have been an issue, and it certainly was not for the Thebans in the wake of their victory.

The absence of sculptural elements on the metopes and on the tympanum and the general sobriety of the building, however, might not say something so much on a material level as on a symbolic one. The intentions of the architects and builders who erected the Treasury of the Thebans and the Limestone Temple are, of course, difficult to fathom. An echo of these intentions can be retrieved through the architectural remains, but we lack textual sources that could help us interpret them. The notion of conspicuous consumption, I argue, and in particular the lack of it, can serve as an interpretational tool for understanding the relative ‘frugality’ of both the Theban Treasury and the Limestone Temple [33]. The term conspicuous consumption, which was introduced by the economist and sociologist Thorstein Veblen at the end of the nineteenth century, refers to the tendency of wealthy people to spend their resources in objects that predominantly serve to publicly display their wealth (Veblen 1899). In an influential article published in 1990, Trigger applied this concept to monumental architecture. According to him, monumental constructions demand an expense of energy that largely exceeds their functional purposes (hence the term “thermodynamics” that Trigger coined to label his approach). Therefore, this expense of energy must reflect symbolic purposes. The common trait of all monumental constructions, he argues, is that they all represent a considerable waste of energy – in the sense that this energy could be used for more essential needs, like sustenance – and that this “petrified” waste attests to the power of those who control this energy (Trigger 1990). Trigger cautiously stressed that monumental buildings are not a “direct reflection of social reality”. Nevertheless, he convincingly showed that they can contain social cues regarding the use of resources.

I suggest that the deliberate alleviation of conspicuous consumption (that is, the conspicuous waste of resources) in a context where it would normally be expected (i.e. temple or treasury architecture) could also be interpreted as a social cue. In a post-disaster context, the sobriety of the material and decoration would have seemed more appropriate than an abundance of wealth. Lavishly decorated monuments were the material expression of a waste of resources, that is, of expenses made for non-essential things. In normal circumstances, such waste was a way of entrenching the power of their funders; in a crisis situation, it might have been perceived as unfair in relation to more urgent expenses pertaining to the survival of people or institutions. More specifically in the case of Delphi, the survival of the sanctuary depended primarily on the recovery of the temple of Apollo. In sum, this was not a time for wasting resources, and the relative austerity of the post-disaster constructions could have echoed concerns about priorities in resource spending in a context of crisis.

While being innovative in some respects, the Treasury of the Thebans and the Limestone Temple are remarkably modest, if not in the detail of their execution, at least in their general appearance. Made of good, but local limestone, they were not particularly spectacular at first sight, and, from an architectural perspective, their capacity to retain the attention lied mostly in the contrast that they created with the nearby monuments. I have shown that this was likely the case with the Treasury of the Thebans; it could not be more true in the case of the Limestone Temple, which stood next to the Tholos, perhaps the boldest architectural attempt of its time with an unapologetically innovative form [34]. At least two distinct phases have been identified in the Tholos, which can be differentiated by two distinct series of sima blocks [35]. The first series shows virtually no trace of erosion, which suggests that it had not been in place for a long time when it was discarded and, therefore, that the building was damaged shortly after its completion. Traditionally, this destruction is dated to 373 BC, and the initial construction of the Tholos to around 380; this dating was recently challenged [36]. Following the destruction, the interior colonnade made of free-standing columns was probably replaced by half-columns attached to the walls of the cella and the first series of sima blocks was replaced by the second one. It is assumed that the roof, initially made of marble tiles, was remade with clay tiles, on the basis of the absence of any remains of tiles that could have corresponded to the second phase of the roof (Bommelaer 2015). If confirmed, this would have constituted a significant setback compared to the initial ambition of producing an all-marble building, and a token of humility. The nearby Limestone Temple would have been another reflection of this humility.

The modest austerity of the Theban Treasury and the Limestone Temple seem to be a trademark of the period during which they were built. Later constructions at Delphi, such as the Treasury of the Cyreneans or the upper parts of the Temple of Apollo itself, returned to a more traditional lavishness in terms of decoration and materials. While presenting some particularities itself, the Treasury of the Cyreneans was a familiar Doric treasury with an in antis plan, which contrasted with the plain façade of the Theban Treasury. It also had a more luxurious appearance, with its elevation made of Paros and Pentelic marble and its elaborate mouldings (Bousquet 1952; Jacquemin and Laroche 2012, 114-22). As for the temple of Apollo, the sima, interior metopes and sculptures of the east and west pediments were all sculpted in marble, which must have given it a much more exuberant outlook than the Limestone Temple (Amandry and Hansen 2010; Croissant 2003).


While focusing on the architecture of a specific place over a relatively short period of time has obvious limitations, it can shed light on the immediate reactions (that is, at the scale of a generation) to traumatic events such as the disaster that struck Delphi in the 370s or the occupation of the sanctuary by the Phocians around the middle of the fourth century. By limiting the timespan of the investigation, it becomes possible to differentiate between short-term responses to the disaster, on the one hand, and practices that are more spread out over time and do not necessarily reflect a specific situation, on the other hand. Some observations have been made that help characterize, from an architectural point of view, the period spanning from ca. 373 to the second half of the fourth century. This is not to say that these architectural features must all be seen as direct consequences of the unfortunate events that marked the century but, rather, that they cannot be understood without taking these events into account.

Considering the complex political framework of the time, the difficulty to gather resources and the logistical problems of transporting materials to the site and assembling them in a space already crowded with monuments, the efficiency of the rebuilding attest to the resilience capacities of the Amphictiony and of the sanctuaries of Delphi. Reconstructions and new constructions implied a reflection on design choices. It is not so surprising that the Theban Treasury and the Limestone Temple, having been built from scratch, show a greater freedom in this respect than the temple of Apollo, which had to take into account a preexisting structure. The architects of the Theban Treasury and the Limestone Temple took advantage of this freedom to introduce some radical innovations. However, they took great care of anchoring these innovations in shapes that were familiar and recognizable by the many. In this way, the new was made acceptable. The process of anchoring innovation is not limited to the period and the place investigated in this paper, but it was perhaps more important than ever in a post-traumatic context, when there was arguably a need for familiar (one might say reassuring) forms. The relatively austere appearance of the Theban Treasury and the Limestone temple, I have argued, could have been the expression of a concern (whether fully conscious or not) with regard to an excessive display of wealth in such context*.

* I am grateful to Floris van den Eijnde and Josine Blok for their useful comments, as well as to the anonymous reviewers for their corrections and suggestions, which greatly improved this text. This study was supported by the Dutch ministry of Education, Culture and Science (OCW) through the Dutch Research Council (NWO), as part of the Anchoring Innovation Gravitation Grant research agenda of OIKOS, the National Research School in Classical Studies, the Netherlands (project number 024.003.012). For more information see



  • BE = Bulletin épigraphique
  • CID = Corpus des inscriptions de Delphes (CID II = Bousquet 1989)


  • Acar, Oguz A., Murat Tarakci, and Daan van Knippenberg. 2019. “Creativity and Innovation Under Constraints: A Cross-Disciplinary Integrative Review.” Journal of Management 45: 96-121.
  • Amandry, Pierre. 1953. Fouilles de Delphes II. Topographie et architecture. La Colonne des Naxiens et le Portique des Athéniens. Paris: De Boccard.
  • ––, and Erik Hansen. 2010. Fouilles de Delphes II. Topographie et architecture 14. Le temple d’Apollon du IVe siècle. Paris: De Boccard.
  • D’Angour, Armand. 2011. The Greeks and the New Novelty in Ancient Greek Imagination and Experience. Cambridge: Cambridge University Press.
  • Black, Max. 1962. Models and Metaphors. Studies in Language and Philosophy. Ithaca: Cornell University Press.
  • Bommelaer, Jean-François. 1979. “Architecture et mathématiques : arithmétique ou géométrie ?” Revue des études Grecques 92: 208-19.
  • –– 2015. “Nouveautés concernant l’architecture de la Tholos de Delphes.” Comptes rendus des séances de l’Académie des Inscriptions et Belles-Lettres 159, 4: 1733-58.
  • ––, and Didier Laroche. 20152. Guide de Delphes. Le site. Athens: École française d’Athènes.
  • Bousquet, Jean. 1952. Fouilles de Delphes II. Topographie et architecture 5. Le Trésor de Cyrène. Paris: De Boccard.
  • –– 1988a. Études sur les comptes de Delphes. Athens: École française d’Athènes.
  • –– 1988b. “La reconstruction du temple d’Apollon à Delphes au IVe s. av. J.-C.” In Comptes et inventaires dans la cité grecque. Actes du colloque international d’épigraphie en l’honneur de Jacques Tréheux (Neuchâtel, 23-26 septembre 1986), edited by Denis Knoepfler, 15-25. Genève: Droz.
  • –– 1989. Corpus des Inscriptions de Delphes. Tome II. Les comptes du quatrième et du troisième siècle. Paris: De Boccard.
  • –– 1993. “La Tholos de Delphes et les mathématiques préeuclidiennes.” Bulletin de Correspondance Hellénique 117, 1: 285-313.
  • Coulton, John James. 1977. Greek Architects at Work. Problems of Structure and Design. London: Paul Elek.
  • Croissant, Francis. 2003. Fouilles de Delphes IV. Monuments figurés, sculpture 7. Les frontons du temple du IVe siècle. Paris: De Boccard.
  • Daux, Georges, and Erik Hansen. 1987. Fouilles de Delphes II. Topographie et architecture 23. Le Trésor de Siphnos. Paris: De Boccard.
  • Dinsmoor, William Bell. 1922. “Structural Iron in Greek Architecture.” American Journal of Archaeology 26, 2: 148-58.
  • Driessen, Jan. 1995. “‘Crisis architecture’? Some observations on architectural adaptations and immediate responses to changing socio-cultural conditions.” Topoi 5: 63-88.
  • Ferrari, Gloria. 2006. “Architectural Space as Metaphor in the Greek Sanctuary.” In The Artful Mind. Coginitive Science and the Riddle of Human Creativity, edited by Mark Turner, 225-40. Oxford: Oxford University Press.
  • Gallet, Yves. 2003. “Le chevet flamboyant du Mont-Saint-Michel et ses modèles dans l’architecture gothique des XIIIème et XIVème siècles.” Les amis du Mont-Saint-Michel. Bulletin annuel 108: 43-55.
  • Hansen, Erik, Gregers Algreen-Ussing, and Rune Frederiksen. 2017. “A Short-cut to Delphi. Indications of a Vehicle Track from a Stone Quarry to the Sanctuary of Apollo at Delphi.” Proceedings of the Danish Institute at Athens 8: 209-63.
  • Hodge, A. Trevor. 1960. The Woodwork of Greek Roofs. Cambridge: Cambridge University Press.
  • Huber, Sandrine, Didier Laroche, Manon Bublot and Anne Jacquemin. 2022. “Delphes : les vestiges sur la terrasse de Marmaria (2017-2021). [Note of archaeological project].” Bulletin archéologique des Écoles françaises à l’étranger, (accessed July 17, 2022).
  • Jacquemin, Anne. 2017. “La fondation de l’oracle de Delphes et les fondations du temple d’Apollon.” In Quand naissent les dieux. Fondation des sanctuaires antiques : motivations, agents, lieux, edited by Sandrine Augusta-Boularot, Sandrine Hubert and William Van Andriga, 33-45. Rome: École française de Rome.
  • ––, and Didier Laroche. 1992. “La terrasse d’Attale Ier revisitée.” Bulletin de Correspondance Hellénique 116, 1: 228-58.
  • ––, and Didier Laroche. 2012. “Notes sur quatre édifices d’époque classique à Delphes.” Bulletin de Correspondance Hellénique 136-137, 1: 83-122.
  • ––, Dominique Mulliez, and Georges Rougemont. 2012. Choix d’inscriptions de Delphes, traduites et commentées. Athens: École française d’Athènes.
  • Jockey, Philippe. 2008. “Delphes, inv. 25584, 25585, 25586. À propos de quelques fragments inédits d’une nouvelle sphinge archaïque colossale.” In La sculpture des Cyclades à l’époque archaïque, edited by Yannos Kouraryos and Francis Prost, 437-66. Athens: École française d’Athènes.
  • Koukouvelas, Ioannis K., David J.W. Piper, Dora Katsonopoulou, Nikolaos Kontopoulos, Sotirios Verroios, Konstantinos Nikolakopoulos, and Vasiliki Zygouri. 2020. “Earthquake-triggered Landslides and Mudflows: Was this the Wave that Engulfed Ancient Helike?” Holocene 30: 1653-68.
  • Laroche, Didier. 1992. “La Tholos de Delphes : forme et destination.” In Delphes. Centraine de la “Grande Fouille” réalisée par l’école française d’Athènes (1892-1903). Actes du Colloque Paul Perdrizet (Strasbourg, 6-9 novembre 1991), edited by Jean-François Bommelaer, 207-23. Leiden: Brill.
  • Mark, Ira S. 1993. The Sanctuary of Athena Nike in Athens. Architectural Stages and Chronology. Princeton: American School of Classical Studies at Athens.
  • Meiggs, Russell. 1982. Trees and Timber in the Ancient Mediterranean World. Oxford: Oxford University Press.
  • Michaud, Jean-Pierre. 1973. Fouilles de Delphes II. Topographie et architecture 15. Le Trésor de Thèbes. Paris: De Boccard.
  • –– 1977. Fouilles de Delphes II. Topographie et architecture. Le sanctuaire d’Athéna. Le temple en calcaire. Paris: De Boccard.
  • Mylonopoulos, Joannis. 2006. “Greek Sanctuaries as Places of Communication through Rituals: An Archaeological Perspective.” In Ritual and Communication in the Graeco-Roman World, edited by Eftychia Stavrianopoulou, 69-110. Liège: Presses Universitaires de Liège.
  • Pakkanen, Jari. 1997. “Entasis in Fourth-Century BC Doric Buildings in the Peloponnese and at Delphi.” The Annual of the British School at Athens 92: 323-44.
  • Partida, Elena C. 2017. “The Disaster and the Experience of 373 BC at Delphi.” In Helike V. Ancient Helike and Aigialeia. Poseidon, God of Earthquakes and Waters. Cult and Sanctuaries. Proceedings of the Fifth International Conference, Aigion 4-6 October 2013, edited by Dora Katsonopoulou, 227-57. Athens: Εταιρεία Φίλων της Αρχαίας Ελίκης.
  • Perrier, Amélie. 2019. “La réorganisation de l’espace au sanctuaire d’Apollon à Delphes au IVe siècle av. J.-C.” In La question de l’espace au IVe siècle avant J.-C. dans les mondes grec et étrusco-italique : continuités, ruptures, reprises, edited by Sophie Montel and Airton Pollini, 71-91. Besançon: Presses Universitaires de Franche-Comté.
  • Rougement, Georges. 2013. “L’oracle de Delphes : quelques mises au point.” Kernos 26: 45-58.
  • Roux, Georges. 1979. L’amphictionie, Delphes et le temple d’Apollon au IVe siècle. Lyon: Maison de l’Orient.
  • –– 1984. “Trésors, temples, tholos”, in Temples et sanctuaires. Séminaire de recherche 1981-1983, edited by Georges Roux, 153-71. Lyon: Maison de l’Orient et de la Méditerranée Jean Pouilloux.
  • Sánchez, Pierre. 2001. L’Amphictionie des Pyles et de Delphes. Recherches sur son rôle historique, des origines au IIe siècle de notre ère. Stuttgart: Franz Steiner.
  • Scott, Michael. 2017. “The Performance of Boiotian Identity at Delphi.” in Boiotia in the Fourth Century B.C., edited by Samuel D. Gartland, 99-120. Philadelphia: University of Pennsylvania Press.
  • Sève, Michel. 2001. “Rapports avec l’archéologie.” Bulletin épigraphique. Revue des études Grecques 114, 2: 480-97.
  • Sluiter, Ineke. 2017. “Anchoring Innovation: A Classical Research Agenda.” European Review 25: 20-38.
  • Thély, Ludovic. 2016. Les Grecs face aux catastrophes naturelles. Athens: École française d’Athènes.
  • Veblen, Thorstein. 1899. The Theory of the Leisure Class: An Economic Study of Institutions. London: MacMillan.
  • Vintzileou, Elizabeth. 2011. “Timber-reinforced Structures in Greece: 2500 BC-1900 AD.” Structures and Buildings 164: 167-80.
  • Viollet-le-Duc, Eugène. 1875. Dictionnaire raisonné de l’architecture française du XIe au XVIe siècle, II. Paris: A. Morel.
  • Wescoat, Bonna D., and Robert G. Ousterhout, eds. 2012. Architecture and the Sacred. Space, Ritual, and Experience from Classical Greece to Byzantium. Cambridge: Cambridge University Press.
  • Winter, Frederick E. 1976. “Tradition and Innovation in Doric Design I: Western Greek Temples.” American Journal of Archaeology 80, 2 (Spring): 139-45.
  • –– 1978. “Tradition and Innovation in Doric Design II: Archaic and Classical Doric East of the Adriatic.” American Journal of Archaeology 82, 2 (Spring): 151-61.
  • –– 1980. “Tradition and Innovation in Doric Design III: The Work of Iktinos.” American Journal of Archaeology 84, 4 (Oct.): 399-416.
  • –– 1982. “Tradition and Innovation in Doric Design IV: The Fourth Century.” American Journal of Archaeology 86, 3 (Jul.): 387-400.
  • –– 2006. Studies in Hellenistic Architecture. Toronto: University of Toronto Press.


1. For the convenience of the reader I give, for each newly mentioned monument, its number in the Guide de Delphes, here with the usual abbreviation SD: Bommelaer and Laroche 20152.

2. Building techniques aiming at countering the effects of earthquakes have been dealt with by Thély 2016 and, for fourth-century Delphi, Partida 2017.

3. The traditional narrative of a tsunami coming from the sea was recently challenged by a team of geologists and archaeologists, who claim that the destruction of Helike was instead caused by mudflows and floods coming from the inland (Koukouvelas et al. 2020).

4. Polyb. 2.41; Paus. 7.25.8-9; Diod. Sic. 15.48-49; Strabo 1.3.18, 8.7.1.

5. See the useful summing up of the question by Rougemont 2013.

6. Bousquet 1988a, 19 n. 4; 1988b, 17-8. Jacquemin and Laroche 2010, 5 have noted that the fragile column of the Naxians, which was standing immediately south of the terrace of the Temple with a sculpted sphinx on its top, had survived the catastrophe, which would confirm Bousquet’s hypothesis that the south part of the sanctuary was spared by the disaster. Partida recently objected that the Naxian offering could have been repaired, on the basis of an archaizing fourth-century inscription mentioning the Naxians and of fragments stored in the museum of Delphi that she suspects belonged to an archaizing replacement of the sphinx: Partida 2017, 237. This interesting hypothesis demands confirmation; Jockey, who identified this new statue, thinks it is archaic, not archaizing: Jockey 2008, 445-7. Furthermore, the inscription could have nothing to do with a repair (Amandry 1953, 7, n. 4 suspects that it was reengraved only after the Third Sacred War, i.e. after 346, like many dedicatory inscriptions) or with the Column of the Naxians at all, as it could also have belonged to another Naxian monument.

7. They were notably used to reinforce the south-western angle of the foundations of the temple, as well as in a retaining wall located east of the Terrace of Attalos, which was built afterwards. Other possibly reused blocks from the Alcmeonid temple have been spotted in various places, including in the foundations of the Treasury of the Thebans (see below).

8. Roux 1979, 137-64. See also the good summary in Jacquemin, Mulliez, and Rougemont 2012, 77-8.

9. See above. The inscription recording to the first contributions is lost, but we know that they date from spring 366 because the Amphictions numbered the pylaiai (six-month periods in between two sessions of the Amphictiony) during which contributions were paid. Thanks to later, well-dated inscriptions that mention the pylaia number to which they pertain, it can be deduced when the contributions started.

10. The oracular consultations did not cease during the reconstruction, and it is likely that the access to the sanctuary was resumed as quickly as possible, perhaps to a limited number of people. The temple itself probably remained inaccessible until the very end of the construction works. In 334/3, a temporary “cover” (στέγαν) was set up “against the Iskhegaion” (CID II 62 A 1, l. 12-15), which perhaps served as a “waiting room” just outside of the temple, at a time when the main structure was completed. This “cover” structure might be identified with an unattributed foundation on the parvis in front of the temple: Bommelaer and Laroche 20152, 223.

11. Bousquet 1988a, 27.

12. Sánchez 2001, 152.

13. These three buildings all benefited from fairly recent architectural monographs; see Michaud 1973 with some details and corrections in Jacquemin and Laroche 2012 (Treasury of the Thebans), Michaud 1977 (Limestone Temple) and Amandry and Hansen 2010 (fourth-century temple of Apollo).

14. Michaud 1977, 115-8, with earlier date suggestions. See also the short discussion about the date in Bommelaer and Laroche 20152, 90.

15. See Huber et al. 2022. In the present paper, the edifice is still referred to as the Limestone Temple.

16. It is often studied through Social Network Analysis; see Amati et al. 2019 for an overview of the basic concepts and recent works.

17. The plan dimensions of the Treasury of the Thebans are 12.33 x 7.22 m; compare with the Treasury of the Athenians (9.65 x 6.57) and the “Doric Treasury” in the sanctuary of Athena (9.74 x 6.60), the two largest Delphic treasuries until then, both dating to the early Classical period. They were significantly larger than the Archaic treasuries and, overall, there is an impression of a continuous evolution toward larger dimensions, with the Treasury of the Thebans as the pinnacle of this evolution. However, the Treasury of the Cyreneans, built around 333-321/1, was quite smaller (5.97 x 9.10).

18. It is true, as it has been stressed many times, that such a device is rather unique in Greek Antiquity, and it is undeniably innovative. However, it is unlikely that it was invented from scratch at the occasion of the construction of the Theban Treasury, and the paucity of timber remains dating from this period should not deter us from trying to trace its origins. A comparable usage of structural wooden framings in ashlar walls is still quite common in vernacular architecture. At Delphi, wooden framings were later used in the stoa of Attalos (Jacquemin and Laroche 1992, 244). For other examples of wooden framings in historical buildings (including in the foundations of Merovingian and Carolingian constructions), see Viollet-le-Duc 1875. For examples of timber reinforcements in Greece at various periods (but not in Classical times), see Vintzileou 2011.

19. Michaud 1977, 107. The cella was about 9.5 metres wide, which is a considerable span given the absence of any interior support. Few temples of Mainland Greece had such a wide, uninterrupted cella (Hodge 1960, 39, table 1).

20. See Winter 1982. The columns also presented an entasis, a feature common to most of the fourth-century Doric buildings in Delphi and in the Peloponnese: Pakkanen 1997, 344.

21. A similar opinion was expressed by Sève 2001, 482, n. 17.

22. Perrier 2019, 80 citing an unpublished report of 1947 by François Chamoux.

23. See Amandry and Hansen 2010, 154-151. Assuming a simple 2:1 length-width ratio (a rough average of the usual ratio in the Doric order at the time), and keeping the north-western angle where it is, the south-western angle would have been situated at about 5 metres of the polygonal wall.

24. The case of the abbey of the Mont-Saint-Michel (Normandy, France) is an interesting parallel to the case of the fourth-century temple of Apollo: in the course of the 15th and in the beginning of the 16th century, the Romanesque chevet of the abbey was destroyed and then rebuilt in the Flamboyant Gothic style. The new chevet was heavily inspired by the older chevet of the cathedral of Évreux (13th-14th century). The choice of the model was likely dictated, in large part, by structural constraints, since the new chevet of the Mont-Saint-Michel was built on an earlier substructure (the crypt), which supported the earlier Romanesque chevet and was preserved when it was destroyed (Gallet 2003). Like in the case of the foundations of the temple of Apollo, the preexisting substructure dictated, at least in part, the shape of the building. I am grateful to the anonymous reviewer for bringing this interesting example to my attention.

25. Winter 1982, 389. This “archaic” feature of the temple of Bassai did not prevent the architect to give it an innovative interior design; see Winter 1980.

26. Michaud thought that the entrance of the treasury was on the east, but Jacquemin and Laroche have shown that it was actually on the west. The window was always, and still is, thought to be on the west side: Jacquemin and Laroche 2012, 106-44.

27. Daux and Hansen 1987 for the architectural study of the treasury.

28. For a detailed study of the architecture of the Limestone Temple, Michaud 1977, to be complemented with Bommelaer’s important comments about the design: Bommelaer 1979.

29. These refinements might suggest that the builders had acquired more experience since the construction of the Theban Treasury and, therefore, that the Limestone Temple was built somewhat later: Michaud 1977, 117-8. However, this argument remains highly conjectural and one has to take into account the hazardous grounds on which the Treasury was built, as well as the difference in scale between the two buildings.

30. “Le charme du temple en calcaire, modeste dans ses dimensions et d’apparence austère, tient essentiellement à l’accord, et même à la fusion, de schémas anciens et de formes neuves”: ibid., 103.

31. According to Winter 2006, 47, it might have been inspired by the treatment of the western front of the Erechtheion.

32. The stone is ideal for sharp, straight edges, but is too brittle for sculpture. However this does not explain the absence of sculpture on the Treasury of the Thebans and on the Limestone Temple, as it could have been made (as was often the case) of another material, and there is no reason to suspect that there was a shortage of supply of other types of stone at the time. One could invoke political reasons preventing the Thebans to access other types of stone, but there is not explicit evidence of a blockade of any sort. In any case, such an explanation cannot apply to the case of the case of the Limestone Temple, or to the Rhodian Pillar (SD 408), which was made of the same stone.

33. It is worth noting that in their latest study of the Limestone Temple, Huber et al. 2022 interpret the building as a meeting place, perhaps for the synedrion of the Amphictions, rather than as a temple, in part based on the absence of decoration.

34. The Tholos of the sanctuary of Athena was not the very first round peripteral building in Delphi: a building of this type had been erected in the Archaic period in the sanctuary of Apollo. The Archaic Tholos was possibly a Sikyonian offering; in any case, almost all of its blocks were reused in the foundations of the Treasury of the Sikyonians (SD 121, around 550-525 BC), along with another small peripteral building, the so-called “Monopteros”. Whatever its filiation with the Classical Tholos may have been, the Archaic Tholos had been removed from the visible landscape more than a century before the former was erected.

35. For a long time, the two series of sima blocks were believed to belong to a single-phased double roof: Laroche 1992.

36. A recent study (Huber et al. 2022) suggests that the Tholos must in fact be dated to the second half of the fourth century. If that was the case, it is possible that it was inspired by the Thymele of Epidauros, built in the second quarter of the fourth century, and not the other way around, as has been thought until now.