Timber Circles at Poverty Point

Near-surface magnetic gradiometry has detected as many as 36 large timber post circles, as part of a larger research effort that also include limited resistance survey and uncovered extensive data on the construction of the distinctive concentric ridges (Hargrave and Clay Reference Hargrave and Clay2016; Hargrave et al. Reference Hargrave, Berle Clay, Dalan and Greenlee2021:198; see Figure 2). Minimal testing of four circles indicates that they are circular holes lacking significant evidence of posts; they are “likely to represent decommission of a structure or the scavenging of wood from a deteriorated structure” (Kennedy and Carter Reference Kennedy, Carter, Redmond and Genheimer2015:338). These circles are elements of what were once geometrically precise circular structures, torn down after use and repeatedly rebuilt (Dalan and Sharp Reference Dalan and Sharp2012; Greenlee Reference Greenlee2009; Reference Greenlee2011; Hargrave et al. Reference Hargrave, Berle Clay, Dalan and Greenlee2021:198–199). Similar decommissioning of wood post structures by post removal was reported earlier for Poverty Point by Haag (Reference Haag and Byrd1990), although the nature of the structures he encountered was not clear because of the limited nature of his test excavations. Despite the perimeter testing of only four post circles and their interiors being known only from magnetic signatures, I interpret them as hypaethral or unroofed spaces or “stages” for ritual performance because of their size and apparent lack of interior features that might be expected in a roofed structure (Pacheco et al. Reference Pacheco, Burks, Wymer, Redmond, Ruby and Burks2021:84–65; Riordan Reference Riordan, Redmond and Genheimer2015, Reference Riordan, Redmond, Ruby and Burks2021); this interpretation is also influenced by current thinking on later timber circles (e.g., Clay Reference Clay2009; Henry and Barrier Reference Henry and Barrier2016:100; Purtill et al. Reference Purtill, Norr and Frodge2014; Zink Reference Zink2009).

Figure 2. Large timber post circles identified by magnetometry at Poverty Point. (Color online)

Occurring in unexpected profusion and in some cases repetitively rebuilt overlapping each other, they are the earliest known timber post circles in North America (Figure 3). They were erected in a historically complex site context that we know includes many smaller posts, isolated timbers, possibly other timber alignments (first noted by Haag [Reference Haag and Byrd1990]), and abundant artifact evidence of habitation. In short, they occurred in a site where there was intensive construction and removal of wooden architectural features, as well as extensive evidence of habitation activity in seemingly isolated cooking features, baked clay balls (Poverty Point objects), and lithic chippage, as well as specialized ground tool forms (Webb Reference Webb1968) not directly associated with the timber circles themselves.

The “ritual” nature of the Poverty Point timber circles is possibly reinforced by data from Jaketown, a far less monumental Poverty Point site in Mississippi (Ford et al. Reference Ford, Phillips and Haag1955; Phillips et al. Reference Phillips, Ford and Griffin1951; Ward et al. Reference Ward, Grooms, Schroll and Kidder2022) where two structures of markedly different form and hypothetical use were identified (Figure 4). The first was a roughly 3.5 m square structure with rounded corners and walls that was constructed of multiple light poles, probably using “flexed pole, bent pole, curved pole” construction techniques (Ford et al. Reference Ford, Phillips and Haag1955:Figure 10; Kennedy and Carter Reference Kennedy, Carter, Redmond and Genheimer2015:343). Because this fragmentary structure lacks interior features like hearths, pits, or other furniture, it is probably a stretch to call it a “domestic” space. Still, it was implicitly a roofed shelter suggesting more informal, vernacular planning and an economy of labor, techniques, and materials than were the case with large post timber circles. Not specifically torn down, although possibly repaired multiple times, its space was not used again for a later structure. As at Poverty Point, this Jaketown example occurs in a stratum with abundant artifactual (fired clay balls and lithics) evidence of habitation.

Figure 3. Detail of large timber post circle cluster magnetic signature partially obscured by grading for the Poverty Point Visitor Center.

The second Jaketown structure was a segment of a timber post circle (Ford et al. Reference Ford, Phillips and Haag1955: Figure 8) consisting of “2 feet deep . . . clearly marked post holes, about 8 inches in diameter and 18 inches apart, . . . a segment of a circle . . . about 5 meters in diameter” (Ford et al. Reference Ford, Phillips and Haag1955: 32). There were no internal features, and although the text suggests that the posts were removed, this is not certain. Identified in “Trench 1,” this post circle originated at the surface of a “brown silt” stratum stratigraphically above the house (Ford et al. Reference Ford, Phillips and Haag1955: Figure 39a:30–34). Poverty Point materials were identified at this surface, and subsequent excavation recovered additional portions of the circle (Ward et al. Reference Ward, Grooms, Schroll and Kidder2022:8) that are firmly associated with an intensive Poverty Point period occupation.

There is a bimodal, non-overlapping distribution of large timber post circle sizes at Poverty Point, with smaller circles ranging between 5 and 40 m in diameter and large circles between 45 and 65 m (Hargrave et al. Reference Hargrave, Berle Clay, Dalan and Greenlee2021; Figure 3). This Jaketown circle falls at the small end of the range of the smaller timber circles at Poverty Point. In the Ohio Valley Middle Woodland where a similar bimodal distribution of timber circles occurs (no cultural historical connection is implied), the smaller of the circles were first interpreted as houses and the larger as open enclosures (Webb and Snow Reference Webb and Snow1945:52–53). Currently both large and small timber circles are now considered open spaces associated with unspecified ritual events (Clay Reference Clay2009; Purtill et al. Reference Purtill, Norr and Frodge2014:75–79). The same may probably be said for Poverty Point, where limited excavation has not identified interpretable features inside the smaller circles that would suggest that they were enclosed. The size difference perhaps reflects the size of the groups that constructed and used a particular circle, suggesting that the circles were used by congregating groups of variable size and with perhaps variable intent. How large and small timber post circles relate to each other is an interesting question that can only be answered by more extensive excavation.

Evidence in the form of a more informal enclosed shelter, as at Jaketown and the Middle Archaic Monte Sano site (Kuttruff et al. Reference Kuttruff, Haag, Brookes and Connaway2005), has not been specifically identified at Poverty Point; however, lithic analysis suggests that there may have been a resident population engaged in “nonritual” activities (Carr and Stewart Reference Carr, Gibson and Carr2004:143) requiring such a construction. Structures to house may be implied by post alignments in Haag's (Reference Haag and Byrd1990:19) limited test, although he specifically rejected this interpretation because the posts lacked associated, identifiable “floors.” In addition, although Greene identified “floors” in Ridge 2 NW, they lacked associated posts (Miller and Greene Reference Miller and Greene2002). Floors are also suggested in Gibson's (Reference Gibson1994:176, Reference Gibson2019) test of the Dunbar Mound, which is associated with a distinctively ritual space. Potentially extensive habitation evidence on or below the ridges (which may have involved structures) was removed by rapid ridge construction or destroyed by erosion/historic cultivation; the anthropogenic materials throughout the ridge fill reflect the churned-up use of midden-rich soils in their construction, not the incremental use of ridge surfaces when they were raised in elevation (e.g., Gibson Reference Gibson2001:105; Kidder et al. Reference Kidder, Kai, Henry, Grooms and Irvin2021:224–225; Spivey et al. Reference Spivey, Kidder, Ortmann, Arco, Gilmore and O'Donoughue2015:150–151). In addition, geophysical survey has suggested that ridge construction was extremely complex, involving smaller “components” that were built and possibly modified, with both activities obliterating earlier or contemporary structural evidence (Hargrave et al. Reference Hargrave, Berle Clay, Dalan and Greenlee2021:203–205).

Although Connolly (Reference Connolly2001:186) failed to identify any structures in one intensively explored area, less substantial structures, not unlike the one recorded for Jaketown, may have been constructed but not been detected by magnetometry because their posts were not removed, as was the case with the timber circles. They might be revealed by extensive area excavation, as occurred in recent Ohio Adena/Hopewell research where less substantial structures were revealed in heavily ritual contexts (with major timber post circles) that were first thought to lack them (Cowan Reference Cowan, Charles and Buikstra2006; Cowan et al. Reference Cowan, Sunderhaus, Genheimer, Connolly and Lepper2004; Lazazzera Reference Lazazzera, Connolly and Lepper2004; Purtill Reference Purtill, Redmond and Genheimer2015). Perhaps they might be found within the Poverty Point “plaza” where they were not destroyed by ridge construction.

There seem to be three significant groupings of Poverty Point circles (Hargrave et al. Reference Hargrave, Berle Clay, Dalan and Greenlee2021:Figure 4, 189–190). The first is located west of the modern highway that bisects the site. It consists of a diffuse scatter of only small circles, many representing single building episodes or rebuilds, indicated by our testing to involve the addition of posts between standing posts (Hargrave et al. Reference Hargrave, Berle Clay, Dalan and Greenlee2021: Figure 4, 199).

Figure 4. Structures at the Jaketown Site: (A) “flexed pole” structure and (B) timber post circle (after Ford et al. Reference Ford, Phillips and Haag1955:Figures 10 and 8).

East of the highway are two tighter clusters of overlapping large circles that represent repeated construction/removal, rather than repair, of much larger structures. In the larger of the two clusters, as many as nine circles were identified by gradiometer survey (Hargrave et al. Reference Hargrave, Berle Clay, Dalan and Greenlee2021:Figure 44); this is an area that was also disturbed by historic construction and there is a possibility that others existed. Circle 26 is 50.13 m in diameter and consists of post holes refilled with magnetic materials—fragmentary Poverty Point objects. The excavated post hole was 1 m in diameter, narrowing to 0.65 m at the base, which was 2 m from its vertical point of excavation. If a post is buried for one-third of its length (Gibson, Alex Reference Gibson1998:106; Ruby Reference Ruby, Redmond, Ruby and Burks2019:96), the suggestion is that a perimeter post could have been as much as 6 m long, of which 4 m were exposed above ground. Such a post would have to be made from massive logs, probably cypress and possibly weighting as much as 1,530 kg (3,390 lb)—these are far removed from the light saplings involved in nonritual construction. A date on miscellaneous charcoal from a hole is 1150 ± 40 BC (2SD cal BC = 1265–1445 BC, Beta-260708; USDOI 2013:227), suggesting that it was associated with later use of the site.

The intense, localized construction of these large timber post circles, which required laborers able to manhandle large cypress posts, suggests that this location had an enduring significance punctuated by acts of circle renewal. In addition, given the size of the posts, it is doubtful that they were removed because they had rotted; as structural members they would have maintained their integrity for some time; it is suggested that a fence post today has the life span of a “decade or less” (Riordan Reference Riordan, Redmond and Genheimer2015:134), and these posts are much larger. Instead, the posts were possibly removed as elements in ritual performance—in effect, a renewal of ritual space—before they decayed. In contrast, the small timber circles west of the road, used differently or by smaller groups of ritual congregants, were repaired, rather than replaced. Importantly, because of the apparent lack of features within them or identifiable floors associated with them, congregants likely moved through the circles, perhaps aligning them with a larger symbolic space, rather them performing ritual events that might have produced more concrete archaeological remains.

A circle is the most economical way to enclose space: for a given area it involves the shortest perimeter. It is not conducive to “packing,” however, and circular enclosures tend to be isolated structures, although they may be contained within larger square buildings for packing purposes (Chacoan culture kivas in the American Southwest). Constructed of many different materials, they have been used for many purposes worldwide, and their possible functions varied. Because uses of them—or seemingly related circular instruments such as the clock, sundial, magnetic compass, or the Qibla compass of Islam (Bonine Reference Bonine, Holbrook, Medupe and Urama2008)—are associated with the development of our early systems of Western thought, we tend to think of them as instruments for geospatial/temporal orientation used to orient the user to the practical or spiritual world around them: Where am I, where am I headed, what time is it, and “so what”? This probably holds true for non-Western uses of circles such as the “earth centering” devices that Hall (Reference Hall1985, Reference Hall1996) delineated for Indigenous North American groups. Thus, in contemporary thinking, circles are assumed to have been used to define time and the cardinal directions. Using a demarcated circle as an “artificial horizon,” direction may be defined in a single day by observing the point at which the rising sun intersects the circle and then where the sun sets. True north (or south) is a point on the periphery of the circle halfway between these two marks. East and west, which are also defined by the equinox, may be determined by erecting a line at right angles to a north–south lie determined in this way.

Observation over a year tracks seasonal changes, from summer to winter solstice and back, crossing the equinox (which also marks true east and west): in this way the circle can act as a mnemonic device to reckon seasonal time and direction. However, timber circles are not efficient devices to measure daily time because the sun passes over the circle, not around the perimeter.

Poverty Point hunter-gatherers did not need an architectural device to practically orient themselves in space and time. Given their economy and their distinctive use of distant, specifically sourced raw materials, we view them as fully acquainted with an extensive and complex surrounding world. Either as pilgrims to Poverty Point from afar or nearby locals, they populated a world with distinct, targeted economic resources and social groups, through which they traveled with purpose, engaging it in the context of the changing seasons. If timber post circles never served as “tactical” structures used to inform daily actions (i.e., which way to go or the time of the year), they certainly served as specialized strategic structures created to physically construct and conceptually interact with rich, geospatially anchored symbolic landscapes. Beyond a mapped universe of people and resources, these landscapes were, as Claassen (Reference Claassen and Classen2016:xiii) might suggest, also “filled with very active spirits and deities that demanded continual human attention.”

In the cycles of large timber post renewal at Poverty Point, we probably see reflected the ongoing spiritual needs of a wide range of human actors, interacting with their shared symbolic world individually or in groups. Their intense, localized construction and renewal created “timely” reused spaces at Poverty Point, which served as continuing stages for recurring ritual performance through which actors moved. That ritual use may also have been specifically involved with their renewal: the circles themselves and the acts of their renewal were sui generis acts of ritual performance, as well as loci for geospatial orientation. The apparent lack of features in the circles would emphasize the “transient” nature of their ritual use.

Finally, a demarcated circle with center and perimeter markers may be used to perform local cadastral tasks. Like a transit, it may be used to lay out alignments with respect to geodesic directions or other factors, such as solstice alignments. Haag (Reference Haag and Byrd1990) suggested some alignments based on his limited excavations in the 1970s. Situated on a grassy open woodland well above the active channel and the Mississippi floodplain, these circles had an unobstructed, 360-degree horizon view (Cummins Reference Cummins2003; Scharf Reference Scharf2011). Gibson (Reference Gibson, Mainfort and Sullivan1998:27) wrote that the site provided the “perfect vista,” in contrast to locations just to the north or south where the view may have been obscured by tall bottomland stands of cypress: it was “open to the east, along a thirty-foot high bluff, providing a perfect vista for observing the rising sun over the treetops.” Timber circles at Poverty Point, specifically those closest to the east bluff edge, were well placed to make sunrise observations and from these make geodesic observations and mark seasonal changes relevant to in-site ritual performance. In them, actors could symbolically center themselves in temporal and geodesic space. Robert Hall (Reference Hall1985, Reference Hall1996) has proposed an omnibus term “world center shrines” to refer to them in all their known variations. Significantly for timber post circle use, the ritual congregant within the circle used the structure to relate one's persona to a world external to the circle itself.

The Mound C Space and Recursive Ritual

The sub-mound C locale, if not the mounded structure itself, is perhaps the best understood space at Poverty Point, despite the extensive testing of Mound B by Ford (Ford and Webb Reference Ford and Webb1956:33–38) and tests at other areas, notably Mound A (Kidder and Sassaman Reference Kidder, Arco, Ortmann, Schilling, Boeke, Bielitz and Adelsberger2009; Kidder et al. Reference Kidder, Ortmann and Arco2008, Reference Kidder, Arco, Ortmann, Schilling, Boeke, Bielitz and Adelsberger2009). However, Mound C has had a checkered career (Ortmann Reference Ortmann2010:659). Treated as a mound by Moore (Reference Moore1913), it was passed over by early workers (Ford and Webb Reference Ford and Webb1956:14–19; Haag Reference Haag and Byrd1990:11) and became identified as a mound through Gibson's work. Gibson (Reference Gibson1994:176, Reference Gibson2001:88–90) revealed inner construction in the form of multiple, sequential activity surfaces with hearths, post molds, and artifact associations quite unlike anything encountered in the ridge excavations. Gibson (Reference Gibson2019:78–81, Reference Gibson2021:44) suggested that the Dunbar Mound, a name he continues to use, was the locus of ritual activities involving concentrations of distinctive imported cherts. There is a sub-mound date of 1546 BC (median age Wk-11253) and a date of 1343 BC (median age Wk-11283) for a feature associated with one of the activity surfaces. A date of 1683 BC (median age Wk-11284) from later mound fill, which is stratigraphically out of sequence, is considered anomalous (Gibson Reference Gibson2019:71; Ortmann Reference Ortmann2007). Gibson's results have been expanded by Ortmann (Reference Ortmann2010:668–672; Ortmann and Schmidt Reference Ortmann and Schmidt2016).

Now termed Mound C by many, the structure has entered mound-building theory in the American Southeast as an example of one corner of a “principal equilateral triangle” of earthworks, now seductively termed in part the “Clark-Sassaman-Heckenberger Triangle” (Clark Reference Clark, Gibson and Carr2004:305–306; Gibson Reference Gibson2019:78–80; Romain and Davis Reference Romain and Davis2011; Sassaman Reference Sassaman2005; Sassaman and Heckenberger Reference Sassaman, Heckenberger, Gibson and Carr2004). This formulation, which stresses that the site was planned as a totality, has uncertain significance in the interpretation of Archaic Period earthworks in general (Milner Reference Milner, Gibson and Carr2004:305–306).

Mound C is one of a series of features forming an eastern north–south axis at Poverty Point, which includes the Motley Mound (16WC7) 2.4 km north of Poverty Point, Mound C, and the easternmost gap through the concentric ridges, outlined by a combination of magnetic and resistance survey (Figure 3). This axis is paralleled by a western north–south axis that includes mounds B, A, and E; some extend this area considerably farther south to include the Lower Jackson Mound (16WC10; Clark Reference Clark, Gibson and Carr2004:174; Gibson Reference Gibson2001:98–99, Reference Gibson2019:32–34, 171; Kidder Reference Kidder, Sassaman and Holly2011:112; Kidder et al. Reference Kidder, Arco, Ortmann, Schilling, Boeke, Bielitz and Adelsberger2009:131). Most assume that these axes are not coincidental, and the western axis, including the Lower Jackson Mound, is now seen as a possible “conceptual continuity” in earthwork siting between Middle Archaic mound building (to which the Lower Jackson Mound dates) and Late Archaic Poverty Point (Clark Reference Clark, Gibson and Carr2004; Gibson Reference Gibson1996:3, Reference Gibson and Marshall2006:315–316; Reference Gibson2007:517; Kidder Reference Kidder, Sassaman and Holly2011; Kidder et al. Reference Kidder, Arco, Ortmann, Schilling, Boeke, Bielitz and Adelsberger2009). Granting them a sense of conceptual finality and ritual importance, Gibson now refers to the axes as the Western (earlier) and Eastern (later) Meridians (Reference Gibson2019:171). For the western axis extended to include Lower Jackson, this name is possibly farfetched because of the 2,000-year temporal gap involved, representing a speculation that cannot be realistically tested (Saunders et al. Reference Saunders, Allen, LaBatt, James and Griffing2001:76): rather it may be an incidental product of the north–south trend of the east-facing bluff on which all are located.

The eastern axis, apart from its possible northern extension to the Motley Mound, aligns only Poverty Point features, bisecting the major clustered rebuilding of large timber circles (Figure 5). It is probable that mound locale and post circles were related in ritual organization at Poverty Point along this axis. Romain and Davis (Reference Romain and Davis2011:47) describe the “northern” orientations” as follows:

It seems likely that this was the case with Poverty Point and those who used it: standing timber circles were likely used in this symbolic effort, and the Dunbar/Mound C locale was significantly involved in related ritual performance.

Figure 5. Poverty Point eastern north–south axis showing relationship of C, timber circle cluster, and south passage through the ridges clarified through resistance overlay. (Color online)

There is an earlier Poverty Point occupation below Mound C. Gibson (Reference Gibson2001:88) described subsequent mound activity as “a succession of horizontal caps lined by hard packed floors. . . . Caps were up to six inches thick, and their topping floors were about as thick as indoor-outdoor carpeting. Floors were strewn with charcoal and fragmentary Poverty Point objects. Post-in-ground buildings had been erected on at least three floors.” The cultural debris suggested a sacred and ceremonial character for the platform activities because it included a wide range of imports, including a concentration of Burlington chert debitage reflecting raw material from the Upper Mississippi Valley.

Ortmann's (Reference Ortmann2010:668) testing has expanded on Gibson's work: “Mound C was built as a series of thin, prepared surfaces ultimately capped with a thick, construction stage. . . . Sixteen [emphasis added] distinct, artificially constructed surfaces were identified in the lowest 60 cm of the artificial fill” (see also Gibson Reference Gibson2019:71; Kidder et al. Reference Kidder, Arco, Ortmann, Schilling, Boeke, Bielitz and Adelsberger2009:112). Distinctively colored sediments were chosen to form these surfaces; Sherwood and Kidder (Reference Sherwood and Kidde2011:74) consider the selection of appropriately colored materials as ritually significant in earthwork construction at Poverty Point. Weathering and grass growing on some surfaces suggest that their use was intermittent and that they were “cleaned” (Ortmann and Schmidt Reference Ortmann and Schmidt2016:28). The combined excavations underscore that multiple, recursive activities were conducted as these surfaces were cleaned, renewed with selected materials, intermittently used in ritual performance, and then terminated when the final mound cap covered them—thereby drastically changing the form and presumably the use and function of the Mound C location (Ortmann and Schmidt Reference Ortmann and Schmidt2016: 34–35). The stratigraphic interpretation of the event sequence at the locale fits well with the interpretation of Poverty Point as a place of pilgrimage, or at least not a place of daily Archaic use.

But the Mound C space was not a mound during its intensive use. According to Ortmann (Reference Ortmann2010:668–689),

In its stratigraphic details, Mound C differs from other mounds (B, E) where sub-mound activities both lack the intense recursive character of those under C and raised activity surfaces in mound fill significantly make them prominently unlike C, until they were terminated by massive “mound” fill (Ortmann and Schmidt Reference Ortmann and Schmidt2016:25–26). The scale of the activity cancelling these loci also sets them apart from Mounded C, which was so much smaller than the others that it escaped attention by a generation of workers (there is a question whether Mound E was ever terminated with a cap). “Mound C” was in fact a reoccurring ritual performance at this space in the artificially raised plaza, and not a prominent, raised, or rising mound stage.

The ritual importance of the Mound C location is likely not the final earthen cap but the sub-mound activities: the 16 successive floors renewals occurred not with the anticipation that they would be obliterated by a mound but that they would be continued. It is potentially misleading to “treat these . . . as closed, stable monuments . . . and more appropriate to view [them] . . . as projects” (Mainfort and Sullivan Reference Mainfort, Sullivan, Mainfort and Sullivan1998:13). An important shift occurred when this space was abandoned, in effect, destroyed: no further use of the locale is shown in the mound fill. In contrast to the multiple large timber circles, which were decommissioned by pulling up posts and emplacing new ones, at Mound C—where successive ritual performance was preceded by thin clay capping of previous acts, perhaps using specially chosen soils and cleaning of the activity surfaces themselves—the place was finally decommissioned altogether after sequential iterations by covering it (but only in part) with a pile of dirt; this covering made the locale obvious but marked it as “not in use” in the context of the developing landscape. The space that this renewed, sub-mound surface represents was effectively removed from further ritual performances, in contrast to the timber circles.

Without denying that the physical “Mound” C meant something to its builders, what that was and how it related to the preceding recursive ritual performances raise other questions. As Kidder and Sassaman (Reference Kidder, Sassaman, Emerson, McElrath and Fortier2009:675) point out, this is an interpretive problem for all Archaic mounds in this portion of the Mississippi Valley. Both those in the Middle Archaic and in Poverty Point end locale activity with no evidence that they were built for a specific purpose going forward in time. On the one hand, the mantled dirt could be an attempt to terminate (end, cancel, contain, bury, hide, conceal, redirect) a symbolically important place: on the other, the mound could be a way to memorialize an earlier sequence of ritual performances or reset the place for new ones that differed in character “as an axis mundi or as a territorial marker” (Saunders Reference Saunders, Gibson and Carr2004:134). Kidder (Reference Kidder, Sassaman and Holly2011:117–118) suggests that “the mounds and earthworks at Poverty Point represent points of social insertion—wherein people consciously used material objects to translate legacies of the past into tools for framing the present and constraining the future”—yet the unknowable “function” of the place before “social insertion” may have been considerably less or quite different than after it. For its builders, the homogeneous mound fill used to cover a locus of repetitive ritual performance may have had much less, perhaps negligible, significance going forward.

Earthworks can also be considered “abandoned projects” (Barrett Reference Barrett1994:13). As Bradley (Reference Bradley1993:2) suggests, “Monuments are made to last, but their meanings are often elusive, and not just for archaeologists.” Clearly, the structure and performance of ritual shifted with the capping of the C locale. Furthermore, to refer to mound C in its totality, including the activity surfaces below it, as “Mound C,” a unitary construct relevant to a larger, comprehensive ritual landscape, is to ignore the shifting importance of space. Yet whereas timber circles were simply renewed with no suggestion that the rituals they formatted were notably canceled at some point in time, at Mound C, the opposite took place. The locale C was itself the site of the ritual performance that engaged actors with imported materials, producing distinctive archaeological remains: ritual performance within the timber circles merely focused the participant outward to geospatial, temporal events—for example, at C—without leaving behind distinctive patterned archaeological remains.

Gibson (Reference Gibson2019:78–81) with others (Romain and Davis Reference Romain and Davis2011) has argued that Mound C was a critical observation point in the Poverty Point landscape involved in solstice sightings. Unless it is determined that the activities that took place at the Mound C locale would be mounded over in the future, they were in no way involved with mound construction. Because mounds seem to “terminate” activity at sites like Poverty Point and earlier Archaic mound sites (Kidder and Sassaman Reference Kidder, Sassaman, Emerson, McElrath and Fortier2009:675), it would seem a stretch to argue that they were in fact critical observation points in celestial observation schemes prior to their construction.

The southern end of the east axis is the south aisle passage through the concentric ridges. Here, the “conceptual” north–south axis appears to have also been used to develop the concentric ridges with their record of coordinated, intense, perhaps largely secular use, because it clearly passes through the ridges. If there is a sequence to ridge construction, this axis—cast as a defined passageway—was perpetuated through the construction of all ridges, possibly underscoring their rapid construction as Gibson (Reference Gibson2001:96) has suggested and, certainly, the continuing importance of the north–south site orientation. As with the Mound C locale, it is possible that the passage was sited and conceptually reiterated by using the timber circles on this eastern axis as sighting devices. But whereas Mound C may have been built to inform those to come or to simply obliterate a locus of abandoned ritual importance, this aisle was built as an ongoing element of a larger developing ritual landscape, and the timber circles seemingly continued as orienting devices for congregants perhaps approaching from the south. In this case the ridges with this gap through them are elements of continuing ritual performance and not, as with Mound C, a termination or afterthought: the ridges and the occupation they represent remained engaged with developing ritual performance at the site along this eastern “meridian,” as well as the timber circles at the point where the southern passage enters the site plaza.

Ritual Spaces in a Larger Site Context

How we construct 400 years of Poverty Point history is a work in progress, although a framework narrative is emerging (Gibson Reference Gibson2019:67–75; Kidder Reference Kidder, Sassaman and Holly2011; Kidder and Sassaman Reference Kidder, Sassaman, Emerson, McElrath and Fortier2009; Kidder et al. Reference Kidder, Arco, Ortmann, Schilling, Boeke, Bielitz and Adelsberger2009; Ortmann Reference Ortmann2010:672–675; Spivey et al. Reference Spivey, Kidder, Ortmann, Arco, Gilmore and O'Donoughue2015). It will continue to be shaped by the scale of the site, the novelty of its components and their stratigraphic implications, the nature of ¹⁴C dating, and, finally and equally importantly, the interests of its interpreters. Combined, these contingencies will continually shape the historiography we use to interpret the logic of individual events and of the whole ritual enterprise represented by the site.

The Mound C locale and the clustered large timber circles date to a rapidly evolving period of the site (about 1450–1250 BC), which saw infilling to level the center of the site, thereby raising its surface along with the activities on it, and the construction of the six concentric ridges bordering it (Gibson Reference Gibson2019; Haag Reference Haag and Byrd1990; Kidder et al. Reference Kidder, Arco, Ortmann, Schilling, Boeke, Bielitz and Adelsberger2009:110; Ortmann Reference Ortmann2010; Ortmann and Kidder Reference Ortmann and Kidder2013). How the mound and the timber circles relate to the final site structure is not clear, but it is clearly an historical question whose answer requires teasing apart the braided history of events that created it. Stratigraphic analysis (Kidder et al. Reference Kidder, Ortmann and Thurman Allen2009, Reference Kidder, Kai, Henry, Grooms and Irvin2021; Ortmann Reference Ortmann2010; Ortmann and Kidder Reference Ortmann and Kidder2013) is demonstrating that certain earthworks, perhaps all, were constructed rapidly, as measured in months, not years. In addition, the largest of them, Mound A, is perhaps the last construction at the site (though a small Mound F to one side seems to be later in time); it was preceded by the six concentric ridges that were rapidly built well after the site was first used, leading Kidder (Reference Kidder, Sassaman and Holly2011) to ironically suggest that this site presents us with “history in reverse.” What were viewed as the structural results of the coordinated development of a ritually composed landscape are now best approached through the understanding of events behind each of them that emphasize their temporality, the pace of their construction, their use or non-use, and their abandonment. When these emerging details are added to, for example, Ortmann's dissection of activities in the mound C locale into 16 recursive episodes, we are increasingly faced with the fact that discrete events occurred with differing implications. What these were and how they were causally linked by human action and developing ritual traditions through time affect how we interpret the overall logic of the site.

However, the imprecisions of statistical ¹⁴C data blur both the existential reality of events and their impact going forward, reducing them to generalizations, and explanations tend to heap generalization upon generalization, “smear[ing] into the past and the future” (Thomas Reference Thomas2006:81, cited by Gilmore and O'Donoughue Reference Gilmore and O'Donoughue2015:15). For Poverty Point the most obvious generalization that I think we can set aside is that it is somehow the planned sum of its parts constructed over 400 years. Even the less categorical assertion that “elements of deep ancestry are evident in the mathematics and engineering of the complex” (Sassaman Reference Sassaman, Sassaman and Holly2011:204) perhaps constrains the contingency of events as they unfolded. The emerging historiography of the site suggests it developed piecemeal, and there are suggestions, notably in Mound C, that functions of the parts of the site space were not static.

Engaging in speculations to be challenged by new data and dating, I argue that the intense recursive activity at the Mound C locale and the clustered construction of large timber circles out of massive logs were features of what we now term Poverty Point's “Plaza.” But although “plaza” tends to refer to a centrally located arena in a larger settlement (and so it has been generally viewed in the case of Poverty Point), it is not so certain that these activities occurred in such a focused space. It is probable that, at least at first, they took place before construction of the concentric ridges and, last of all, Mound A, which really defined a plaza fronting them. Habitation occurred within the general area, and its remains are reflected in tests (Haag Reference Haag and Byrd1990). Although there is no evidence at present of domestic features like enclosed houses, it is quite probable that they existed and were similar to the structure identified at Jaketown. They were possibly temporary, but when combined with the evidence of the timber circles, would support the suggestion that the site was the object of pilgrimages with an obviously ritual intent, not a continuous “everyday” occupation. Habitation also occurred outside the “plaza” prior to ridge construction (Connolly Reference Connolly2001) without evidence of substantial domestic structures (several small timber circles we have identified west of the modern road would appear to have been built prior to episodes of ridge construction [Hargrave et al. Reference Hargrave, Berle Clay, Dalan and Greenlee2021:198–199]). It is not clear whether the use of the plaza interior was combined with contemporary use of the surroundings such that the latter may merely have been an extension of the former. In other words, we do not know whether the area to be covered by the concentric ridges was only used after a pattern of ritual performance had been developed within the plaza: occupation may have been contemporaneous.

There is a geophysical suggestion that at some point the plaza was structurally defined, possibly prior to ridge construction and perhaps by a palisade. At present it is known only from a curving, negative magnetic signal that has not been tested and only from general magnetic survey; this signal suggests a curving structure beginning in the vicinity of Mound C and passing through a disturbed area where it was obscured by complex events, including an historic occupation. A speculative reconstruction of this feature, assuming it is a palisade, loops it to the north to contain the Mound C space and to the south traversing Haag's test unit (Figure 6). Haag identified what he considered to be three stages of construction involving posts. The earliest stage consisted of massive post holes 60–75 cm in diameter and 3 m deep, from which all traces of posts were removed and replaced with nonmagnetic loess filling (Haag Reference Haag and Byrd1990:18). It is tempting to suggest that this feature could be a palisade or delimiter of some sort.

Figure 6. Possible palisade signature inside the concentric earthworks extending from the vicinity of Mound C and crossing the highway, a passage through the earthworks, and the vicinity of Haag's excavations.

An important problem in interpretation of the use of the cluster of large timber circles is their asymmetrical location in what was later to become the “central” plaza. Located on its southern edge, they become not central devices for ritual orientation in the Poverty Point mounded landscape as it would finally develop with the additions of ridges and Mound A, but rather southern entrances to a larger space that potentially was not central to the ritual landscape at that time. There they emphasized the north–south “meridian” to use Gibson's appropriate term, perhaps for congregants entering from the south. They may also have persisted in use with the definition of a plaza through ridge construction, defining the southern entrance through the mounded ridges as they were constructed. Still, because they are off-center in the final mounded landscape, they could not have been used to lay out other entrances through these ridges that have figured importantly in reconstructions of the site as a calendrical observatory (Brecher and Haag Reference Brecher and Haag1980, Reference Brecher and Haag1983; Purrington Reference Purrington1983; Romain and Davis Reference Romain and Davis2011). Because of this, they almost certainly imply that there were additional similar structures in the plaza that were destroyed by recent construction or eroded away by the precipitous bank of Bayou Macon, as suggested by Haag (Reference Haag and Stoltman1993).

Discussion and Coda

There is a contrast between the nature of ritual at the sub-mound C locale and the clustered timber post circles built of massive logs along the eastern north–south axis of Poverty Point. The circles were built with substantial materials, used for an unknown period, and renewed. However, they may have functioned in on-site ritual: it would appear that congregants passed through them, without using their circular space for recurring ritual performances that would have produced distinctive, magnetically susceptible archaeological signatures—in other words, “floors”—as was the case at the Mound C locale. Ironically, they were also built ultimately to be torn down, suggesting that acts of circle renewal were purposeful ritual acts. By analogy to the great temple of Ise Dai Jingu in Japan and to other Shinto shrines in general, where deconstruction and rebuilding were elements in spiritual renewal (in the case of Grand Shrine, about every 20 years; Adams Reference Adams1998; Mikl Reference Mikl2014), these periodic construction events at Poverty Point may have been similarly multiple expressions of renewal. Still, and importantly as architectural devices, they were not designed to define loci of concentrated ritual performance; rather they were used to focus congregants outward to a much larger, developing ritual landscape centered on the Poverty Point locale. That landscape probably included recurring ritual events at the Mound C locale.

The sequential events at the Mound C locale produced distinctive archaeological remains in a way that ritual performance within timber circles did not. These events reflected acts of construction and renewal but were followed by intense use, unlike the interiors of the timber circles. For Mound C locale history, use involved repeated renewal and cleaning of activity surfaces and then their conclusive ending by covering their remains with mounded fill. The cancellation of the Mound C locale suggests that ritual performance there became no longer relevant to the evolving landscape.

Whether this is also true of the clustered large timber circles on the east axis is unknown. Because they are not central to the developing C-shaped site plaza at Poverty Point—by all accounts, a very late development of the ritual landscape—it is just possible that they reflect one temporal stage in ritual alignment. The eastern “Meridian” clearly persisted in importance, leading to construction of the south passage through the concentric ridges; yet the ritual importance of the acts of timber circle construction/renewal on that axis may or may not have continued. Finally, unlike Mound C, the space of the timber circles was not “obliterated” by mounding over: the large timber posts, if they remained, were simply removed.

Other earthworks at the site also suggest sequences of events that are not clearly understood. In the case of Mound B, perhaps the earliest construction at Poverty Point (Kidder et al. Reference Kidder, Arco, Ortmann, Schilling, Boeke, Bielitz and Adelsberger2009:110; Ortmann Reference Ortmann2010:673), there was a sequence of activities involving successive platform-like structures clearly terminated by a final mound cap; this prompts the question how the locale, possibly focused on the field adjacent to the south and west of Mound A, figured in the larger Poverty Point ritual landscape before and after that terminating construction. Recent geophysical survey suggests that there may have been a circle enclosure just south of that activity area, later cancelled by the mound construction itself (Lipo et al. Reference Lipo, Sanger, de Smet and Patchen2018).

The iconic crescentic earth ridges appear to have been built rapidly (Kidder et al. Reference Kidder, Kai, Henry, Grooms and Irvin2021:225) and piecemeal (Hargrave et al. Reference Hargrave, Berle Clay, Dalan and Greenlee2021:Figure 12, 205), using midden from the extensive pre-ridge use of the site. However they were used, they dramatically changed human use outside what became defined by their construction as a central “plaza” in the site. They suggest important changes in the plan of Poverty Point during its development as a ritual landscape. As Lepper (Reference Lepper, Martin Byers and Wymer2010:124) has put it, “Places have a history and cultural landscapes are continually evolving.” By chipping away at the structure of events, recognizing that they define beginnings, uses, and endings (using Riordan's [Reference Riordan, Redmond and Genheimer2015] apt term), we get the impression that the Poverty Point site was the locus of physical landscape modification, which reflected changes in the projected symbolic ritual landscape, rather than the projection of a master plan. Recalling Kidder's perceptive comment that elements of the site ironically reflect “history in reverse,” some understanding of these events leads us backward in time to an understanding of why it all occurred. Likewise, they raise the equally important question: What did the earthworks mean going forward to their builders? According to Mainfort and Sullivan “Final form cannot necessarily be assumed in . . . initial construction because form was accumulative and resolved only through a sequence of events” (Mainfort and Sullivan Reference Mainfort, Sullivan, Mainfort and Sullivan1998:13).

For example, as the final earthwork—a massive singularity and the product of intense communal effort over a very short period—Mound A would appear to close out both the development of and the use of the Poverty Point ritual landscape. Was it built simply as a memorial to what had been, but if so for whom? The use of Poverty Point apparently ceased soon after. Could it have been rather a “coming together of diverse people” to create a novel and “elaborate and emphatic” act of “cultural representation” (Sassaman Reference Sassaman, Sassaman and Holly2011:205)? As such, and as a dramatic ending to the site, could it not also represent a failure, metaphorically somewhere between “a bridge too far” and the biblical Tower of Babel, and perhaps a dramatic cancellation of what had been an unsustainable communal construction effort or a failure to attain and project a consensual function? The site was never used again (in effect avoided?) as such a prominent place of ritual, and Woodland period occupation essentially shifted just south to the vicinity of the Jackson Place (16WC6) site.

By contrast, other Poverty Point sites, notably Jaketown, continued in use throughout the Woodland period after a flood episode ended its Poverty Point occupation (Ward et al. Reference Ward, Grooms, Schroll and Kidder2022:10–11). Period activity at Jaketown, indeed all Poverty Point Period sites, did not mirror what went on at the Louisiana-type site (Ward et al. Reference Ward, Grooms, Schroll and Kidder2022). This prompts the question of the intent of the notable elements of the larger site: Just what was it that its users sought to project ahead in time with their earthwork constructions? Dissection by event-oriented archaeology does not presently produce a masterful reconstruction of site function, but it does provide a running critique of it. There is a continuing need for these larger exercises, and Gibson's (Reference Gibson2019:161–179) most recent attempt points the way. Tempering these with the nuance of events and their implications as we begin to recognize them in temporal perspective can only enhance their heuristic ability. To follow Barrett's (Reference Barrett1994:13) thinking, I propose that the historiography of Poverty Point as we develop it be built around the proposition that the site “is the physical remnant of a number of abandoned projects . . . undertaken at different locations within the landscape, and at these locales actions and exchanges between people created the material conditions” they hoped to attain. At Poverty Point, as we progressively attempt to understand with additional work what generations of actors attained or sought to attain, we approach their interpretation with caution and flexibility.