Introduction

The summit of Structure 1 at the ancient Maya center of Actuncan, Belize, rises above the modern forest canopy, a stone reminder of a once great city standing watch like a sentinel over the landscape (Fig. 1). The summit’s 28-m height looms above the Mopan River, which snakes just to the east (Fig. 2). Now covered in patches tropical forest and cow pasture, this river valley once was home to ancient Maya farmers, who, from its construction in the first millennium B.C., could look up and see this monument—first as a brightly colored Preclassic temple pyramid, then as a Classic period ruin slowly crumbling due to neglect. Despite its advancing decrepitude, the remaining residents of Actuncan’s center continued to walk past its ruined monuments. Structure 1’s tree-covered mass, which formed a striking silhouette on the horizon, remained part of daily life and anchored the collective remembering of Actuncan’s past across the passing centuries.

Fig. 1
figure 1

View of Actuncan from the alluvial terraces on the opposite bank of the Mopan River. Structure 1, the tallest at Actuncan, is indicated. Note that Actuncan would have been visible from throughout the valley, even if the forest regrew

Full size image
Fig. 2
figure 2

Map showing Actuncan’s location on the Mopan River within the Belize River drainage (redrawn from LeCount, 2004; Fig. 1)

Full size image

The importance of the past to the ancient Maya has long been recognized through explicit statements of descent made by rulers and ordinary households’ practice of repeatedly burying important ancestors in the same locations under house floors (McAnany, 1995; Proskouriakoff, 1960). In the sixteenth century, Spanish priests recorded the oral traditions of Maya lineages, who legitimized their control over people and territory through records of descent and memories of inherited lands (e.g., Roys, 1939). For the ancient Maya, the collective memory of descent served as a primary source of legitimacy for control over people or land (McAnany, 1995). In the archaeological record, the importance of memory is also evident in the palimpsest use of architectural space by the ancient Maya. Ruins were built over or repurposed by later people as a way to create genealogical ties to the past (Halperin, 2014; Kristan-Graham & Amrhein, 2015; McAnany, 1998; Navarro-Farr & Rich, 2014; Simova, Mixter, & LeCount, 2015; Stanton & Magnoni, 2008). Additionally, old sites and buildings were frequently the destinations of pilgrimage, as evidenced by altars and the remains of ritual practices encountered atop collapsed structures (M. K. Brown, 2011; Hansen, Howell, & Guenter, 2008; Robin et al., 2012a). Drawing on these and other examples, Canuto and Andrews (2008, pp. 269–271) have commented that Maya historical awareness existed at variety of levels of abstraction depending in a large part on temporal distance and occupational history (see also Gosden & Lock, 1998). Generational oral histories, social continuity, and hieroglyphic inscriptions encouraged accurate remembering. Over longer time periods, markers of the past blended into the Maya wilderness and took on meaning as metaphorical natural features and supernatural homes to the ancestors (D. F. Brown, 2008; Halperin, 2014; Taube, 2003, pp. 467–468).

Following Connerton’s (1989) influential exposition on How Societies Remember, much archaeological work on collective memory has focused on the inscribed and embodied qualities of memory production. Although Connerton (1989, pp. 73–79) was more directly interested in text and modern media, archaeologists have rightly expanded the inscribed category to include the placement of tangible markers on the landscape—the materialization of past actions by people of all levels of society (Bradley & Williams, 1998; Rowlands, 1993; Van Dyke & Alcock, 2003; Yoffee, 2007). This theoretical anchoring has led to two approaches to collective memory. One approach focuses on how societies intentionally reworked their pasts through performance and modifications to the lived landscape to legitimize authority or influence the future of a society, state, or collective (Alcock, 2002; Bradley, 1987; Schortman & Urban, 2011; Schwartz, 2013; Sinopoli, 2003). A second approach focuses on the memory created through every day practices and how this kind of quotidian memory creation leads to the production of identity, land tenure, and meaningful places (Ardren, 2015; Bradley, 1998a; Hendon, 2010; Hodder & Cessford, 2004; Joyce, 2003; LeCount, 2010b; Lillios, 2008; McAnany, 1995). These two perspectives are important because they underpin claims to authority by groups and individuals at the state and household scales (Hobsbawn & Ranger, 1983; McAnany, 1995, 1998). However, neither the state nor the household view of memory adequately considers how claims about the past gain salience. The execution of a brilliant performance by a ruler does not guarantee that this event will be remembered as evidence of their greatness and magnanimity. Rather, memories are perpetually digested and interpreted within social, political, and historical contexts. This mental processing provides difficulties for archaeologists who can easily see the materialization of official narratives but have difficulty seeing the interpretation of formal events and official propaganda by everyday people (Overholtzer, 2013; Trouillot, 1995). Yet, minority and marginalized understandings of the past are critical to understanding the stability (or instability) of formal institutions and trajectories of reorganization during political transitions (Pauketat & Alt, 2003; Schwartz, 2006; Yoffee, 2015).

To bridge the divide between performative and practice-based approaches to collective memory, I draw on the concept of collective remembering to develop a relational approach to the archaeological study of collective memory that considers the impact of both official messaging and informal community-level interactions on the process of memory production and perpetuation (Bartlett, 1932; Middleton & Edwards, 1990a; Wertsch, 2002, 2009). Public monuments, ceremonies, and events are not accepted as whole cloth but rather are interpreted by subordinate groups and individuals within a cultural and historical context. Importantly, collective remembering results from repeated interactions and conversations that happen through the practice of daily life (Halbwachs, 1980 [1950], pp. 22–49; Middleton & Edwards, 1990b). Rather than seeing the production of official memory and the everyday creation of memory as discrete processes, collective memory is synthesized relationally through a web of interactions structured by periodic events, individual’s daily schedules, and the physical setting that structures the rhythm of people’s daily lives. Thus, collective remembering becomes a process that results from daily practice and is constrained by the spatial dimensions of interaction (de Certeau, 1984; Mills & Walker, 2008b, pp. 7–8). Furthermore, routinized local interactions and the materiality of city and village facilitate the development of mnemonic communities—defined by Zerubavel (1996, 1997, p. 90) as families, organizations, nations, or other social groups that collectively possess a set of memories, even if the people and events remembered were not personally experienced by all members of the group (see also Bellah, Madsen, Sullivan, Swidler, & Tipton, 1985, pp. 153–155). The collective remembering within these communities reflects a local interpretation of the past and may reify alternatives to official narratives as a kind of “hidden transcript” in resistance to official narratives propagated by hegemons (Scott, 1990). Rather than viewing collective memory as monolithic or a tool of domination, approaching memory from the perspective of interactions draws attention to the potential for locally diverse understandings of the past and their implications for power relations.

In this article, I draw on the idea of collective remembering to explain the sociopolitical transformations that took place at Actuncan, Belize, during the Terminal Classic period (A.D. 780 to 1000). At this time, communities across the Maya lowlands were enacting profound changes to their governing institutions as the royal courts of the Classic period failed and the legitimizing principles of divine kings were rejected (Iannone, Houk, & Schwake, 2016). The timing and nature of these transformations varied across the Maya lowlands (Aimers, 2007; Demarest, Rice, & Rice, 2004), and individual communities used their increasing autonomy to develop solutions adapted to local needs (Masson, 1997; Schwarz, 2013). During this time period, Actuncan transformed from a secondary administrative center to a local capital. A new civic center was constructed within the site’s urban core, and a new council-based political organization replaced the region’s divine king and his elite bureaucracy (Mixter, 2016). These transformations reflect choices made by the Actuncan community to not emigrate; to continue to occupy the same hilltop settlement; and to build a new political regime with council, not royal, leadership.

I argue that Actuncan’s Terminal Classic sociopolitical transformations were inspired and legitimized by the community’s collective remembering. Due to its deep history, Actuncan was no ordinary Late Classic village. Rather, the site was one of one of the region’s earliest capitals. The site’s towering monumental core was constructed during the Late Preclassic (400 to 150 B.C.) and Terminal Preclassic periods (150 B.C. to A.D. 250) when it served as the local seat of authority. In the Early Classic period (A.D. 250 to 600), Actuncan’s leaders lost their power, and authority passed to rulers seated at other centers in the region. Initially, the site’s public architecture fell into disuse, though many of the community’s households remained occupied. In the Late Classic period (A.D. 600 to 780), Actuncan was repurposed as a secondary administrative center within the regional polity centered at nearby Xunantunich.

Throughout this dynamic political history, Actuncan’s monumental center formed a backdrop to daily life for residents of the local households and anchored a community identity fostered by a collective past. As a center of political authority, Actuncan formed a nexus of interactions—both between people and between people and the material landscape—that facilitated the collective remembering of the past. Importantly, these memories were not retained by repeated, elite - centered performances, but rather through the quotidian interactions of daily practice that form an under-considered component of collective remembering. As a result, the common experience of Actuncan’s residents anchored the community’s identity project and, I argue, resisted attempts by Late Classic political leaders to unite the region under a unified identity (Fields, 2004; LeCount, 2010a, 2010b).

When Actuncan’s community built their council-based political organization during the Terminal Classic period, they had to contend with their long and complex history. Indeed, the Terminal Classic community made specific decisions about how to utilize the preexisting urban landscape that necessarily referenced their interpretation of the local past. To understand how Actuncan’s mnemonic community remembered its past, I trace the experiences of community members through three eras in the site’s political history—its Preclassic apogee, its Classic period time as a subordinate center, and its Terminal Classic revitalization. I argue that acts of memorialization from the Terminal Classic period reflect a reemergence of alternative narratives curated during the period of subordination through collective remembering. Terminal Classic ceremonies and community events publicly recalled this local understanding of the past to rally the local community in support of Actuncan’s political revival.

This article is structured in three sections. In the first section, I establish collective remembering as a framework for understanding how to identify collective memory in the archaeological record. In the second section, I propose a methodology for identifying collective remembering in archaeological contexts. Third and finally, I present Actuncan as a case study to show how a collective remembering approach can be used to gain a new perspective on the social changes that take place during periods of political transformation.

Collective Remembering: a Relational Approach to Collective Memory

Ancient Maya rulers were masters of producing elaborate spectacle. The murals from Bonampak, Mexico, portray the pageantry of Maya kings in elaborate full color (M. E. Miller & Brittenham, 2013). Magnificent processional parades are accompanied by musicians, entertainers, and warriors in full regalia (see also Morton, 2012; Reese-Taylor, 2002). Captives stretch sprawled across the stairs of a great pyramid, suffering from the removal of their fingernails or distal phalanges, while Bonampak’s elite stand triumphant (Johnston, 2001; M. E. Miller & Brittenham, 2013). This kind of performance was intended to reify charismatic royal authority by providing a venue for kings to display their wealth, martial power, and cosmological associations (Demarest, 1992; Freidel, Schele, & Parker, 1993; Inomata, 2006a). Spectacles were motivated by specific events—whether a royal accession, a marital triumph, or an important calendrical date—and were intended to establish particular memories in the minds of local polity members. The first-hand experiences of these kinds of commemorative ceremonies are discussed by Connerton (1989), who suggests that their power resides in incorporation—the memory formed through bodily participation either as a performer or spectator. For the Maya, this kind of public event both constructed the polity by gathering the local populous in one place and reified sociopolitical inequality through the physical separation of performing elites and spectating commoners (Inomata, 2006b, pp. 807–809).

Just producing a great spectacle does not guarantee that the event was processed and remembered as intended. After an event takes place, remembering and interpretation begin, providing space for disagreements over the facts and differences in opinion about what transpired. In his foundational study Remembering, psychologist Frederic Bartlett (1932) argued two critical points. First, memory is not a set of information possessed by collectives for individuals to draw on—as previously argued by Maurice Halbwachs (1992 [1925])—but rather a constructive process in which individual memories are influenced by social associations and cultural context. Thus, memory is produced, in part, through individual cognition but also through social interaction. Second, memory does not stand as an accurate representation of the past but is instead an “effort after meaning” (Bartlett, 1932, p. 44). By this, Bartlett meant that individuals reconstruct memories in the moment of recall and in the effort actively reshape them in the context of the present (Schacter, 1996). Numerous psychological studies have pointed to the inaccuracy and memory distortions that result from the constructive nature of recall (Roediger & McDermott, 2000; Schacter, 1995). To emphasize that recalling the past is a collective and constructive process, Bartlett preferred the term “remembering” to “memory,” the latter of which evokes the static connotation of Halbwachs’ initial formulation of memory.

As a psychologist, Bartlett’s writing on remembering ultimately focused on the impacts of sociality on individual cognition (Douglas, 1986, p. 81). From this perspective, individuality and the constructive nature of remembering insert instability into the process of memory production. In contrast, social interaction, public commemoration, and cultural norms provide structure to the collective production of memory. Psychological studies of “collaborative remembering” show how memory is produced through social interactions—the memories of individuals placed together tend to converge on a common narrative (Weldon, 2001; Weldon & Bellinger, 1997). Put another way, individual memory largely derives from internalizing social discourse over time (Wertsch, 2002, p. 37). If commemorative ceremonies, such as those put on by Maya rulers and elites, provide events intended to be fixed in memory, their internalization is ultimately conditioned by the discourse that develops around that event, its associated actors, and the historical moment.

Collective remembering refers to the processes by which a common understanding of the past is resolved from individual versions of events (Wertsch, 2002, 2009). Rather than a static view of collective memory, collective remembering acknowledges both the fallibility of individual memory and the role of social interactions in the process of remembering. The collective memory does not necessarily represent the facts of the past—a reality long acknowledged in archaeology—and is also mutable, unstable, and often not uniform across a population. Rather, as an aggregation of individual memories, collective memory is influenced by the limits and biases of individual observations, interests, and agendas. Although individual memories are never fully reconciled, a collective memory emerges from public discourse, wherein common understandings of the past are reified and the memories of some individuals adjust to accepted and popularized versions of the past.

Collective Remembering and Landscapes

James Wertsch (2002, pp. 37–40) identifies social interaction as a critical locus of collective remembering, where people both remember together and internalize collective memories beyond their autobiographical experience. I specifically focus on three kinds of interactions that are critical to collective remembering: public events, including the kinds of public spectacles typically discussed in archaeology (Coben & Inomata, 2006; Meskell, 2003, pp. 47–49; Reese-Taylor & Koontz, 2001; Roddick & Hastorf, 2010, pp. 167–172); informal interactions based around the conversations that take place in the course of everyday life (Hendon, 2010; Middleton & Edwards, 1990b); and the role of monuments and important points on the landscape as mnemonics that focus and perpetuate memory (Alcock, 2002; Basso, 1996; Bradley, 1998b; Fitzhugh, 2017; Gillespie, 2008; Prent, 2003; Van Dyke, 2003), even if collective remembering may cause their original intended meanings and associations to drift. These categories of interaction are not intended to be understood as exclusive to each other. Rather, all are part of the practice of everyday life, wherein individuals—and their memories—are constituted by their interactions and relationships (de Certeau, 1984). If public events form the seed of what is to be remembered, it is conversations—both at the event and in the days that follow—that cement how events are remembered.

Monuments and the constructed landscape play different roles in framing collective remembering. Public space serves as the setting for the public events and informal interactions that facilitate collective remembering. Leaders use public spaces as theaters to host formal events linked to political events and ritual calendars that emphasized rulers’ power (Inomata & Coben, 2006; Moore, 1996b). Social solidarity was often encouraged through the audiences’ collective witnessing of performances, dancing, and state-sponsored feasts. These kinds of events produce collective memory through incorporation as described by Connerton (1989). However, formal events at smaller venues, including households, village centers, and neighborhood gathering places, bring communities together at a more local scale (Cap, 2012; Peuramaki-Brown, 2013; Yaeger, 2000b). These smaller venues promoted more personalized contact between attendees (Moore, 1996a) and consolidated collective identities through food sharing and active participation in rituals.

Additionally, city and village centers likely served as loci for interactions and collective remembering outside of periodic events. In the modern cities of Central America, plazas and public spaces are centers of public life where farmers, craftspeople, and traders exchange wares and gather to chat and gossip (Gade, 1976; Low, 1996). Ethnographers have pointed out that these kinds of informal interactions have important political implications (Low, 2000), and they provide the venue for collective remembering to take place. Among the ancient Maya, the recent identification of public plazas that were used as periodic and permanent marketplaces indicates that ancient Maya public gatherings were not just for formal performance and spectacle but also provided ample opportunity for trade and socialization, as is common in market settings (Freidel, 1981; King, 2015). Similarly, archaeologists interested in how identity is created relationally have pointed to the role fields, pathways, and neighborhoods played in facilitating inter-household interactions that form the foundation for communities (Ardren, 2015; Hutson, 2010). As I will discuss more, space limits who remembers together and facilitates the creation of communities based on a common understanding of the past. Just as proximity and frequency of interaction facilitate the convergence towards a collective memory, residents of distant locales interact less frequently, and their interactions may clarify divisions between communities based on differences in understandings of the past. Space plays a role in structuring how similarities and differences in collective memory develop and are acknowledged as part of the process of community and identity formation.

Monuments and urban space are durable reminders of the past that structure and influence collective remembering in addition to providing a venue for the person-to-person interactions that facilitate the production of collective memory. Recent work by archaeologists and social theorists interested in “post-humanism” or the “ontological turn” aim, in part, to credit materials, including the constructed landscape, with their ability to shape and influence people, places, agendas, ideas, and practices (Alberti, Fowles, Holbraad, Marshall, & Witmore, 2011; Appadurai, 1988; Gell, 1998; Hodder, 2012; Ingold, 2011; Knappett & Malafouris, 2008; Latour, 2005; D. Miller, 2005; Watts, 2013). Similarly, the ancient Maya built environment impacted how the past was remembered. I have already laid out how public space provided the venue for conversation. However, buildings, roads, and monuments were also markers dedicated to past kings, venues for past ceremonies and spectacles, and places where official business of the polity or village takes place. They inscribed the past—sometimes literally with texts—onto the landscape for all to remember (Alcock, 2002). The durability of urban space, stone monuments, and monumental architecture perpetuated knowledge of the past because their existence frames the kinds and content of interactions that take place. Indeed, the continuous presence of people within these spaces perpetuates their importance as markers of the past (Van Dyke, 2017).

Yet contra Latour (2005), humans and the constructed landscape are not equivalent actants—entities that modify the action of each other. Rather, monuments and monumental architecture were power-laden emblems of the asymmetrical relations between elites—who commissioned and lived among them—and commoners—who constructed and dwelt at their margins (Van Dyke, 2015). The construction of temples, palaces, and monuments reflected the motives, affiliations, and memories of those in power at a given moment (Golden, 2010; Iannone, 2010). However, self-aware interpretations and memory of durable monuments result from discursive conversations and other social engagements. As a result, these kinds of monuments become entangled with meanings, associations, and remembrances that may not match the meanings intended by those who built them (Meskell, 2003, pp. 49–53; VanValkenburgh 2017). Furthermore, the memory around places develops discursively through events and conversations. In this way, collective remembering empowers non-elites, who are less visible in the archaeological record, to reinterpret monuments based on their own relationships, socioeconomic, and historical contexts (Pauketat & Alt, 2003).

Variability in Memory: Communities and Alternative Narratives

Variability in collective memory can also be understood in terms of narratives across social class. Wertsch (2002, pp. 55–56; also 1998) identifies “narratives as one of a few different instruments we have for representing settings, actors, and events of the past.” In part because memory is constructive, narratives are dialogic, delivered, and conceived within the context of a particular teller or occasion of telling (Wertsch, 2002, pp. 59–60). The delivery of a narrative is cast in the agenda and perspective of a particular teller. As such, narratives propagated by dominant powers in public venues may be different than those told by common folk spoken in alleys or kitchens. Wertsch (2002, pp. 117–148) traces how narratives of the past in the Soviet Union persisted on parallel tracks—an official, propagandist narrative was propagated by the government, while a set of folk narratives were passed on illicitly in kitchens and dining rooms. The latter were valued precisely because government authorities suppressed it. Several of the other articles in this issue provide additional examples of how multiple narratives of the past can co-exist despite pressure to adopt an official understanding of the past (Klehm, 2017; Overholtzer & Bolnick, 2017; Pool & Loughlin, 2017; Van Dyke, 2017). Similarly, Maya rulers clearly aimed to propagate a specific narrative of their lives and their polities’ histories through inscriptions on stone and monumental construction in stone (Marcus, 1992; see also Pool & Loughlin, 2017).

What we do not know is how tightly commoner Maya narratives adhered to these official narratives. Were alternative narratives of the past maintained in households and rural communities? How tightly did official narratives hold the allegiance of commoners? If alternative narratives were maintained in hinterland communities, these can be considered a kind of resistance. James Scott (1990) provides one framework for understanding resistance. He differentiates between public transcripts—narratives and practices adopted in public contexts—and hidden transcripts—practices and narratives maintained solely in private contexts, often in opposition to dogma. Hidden transcripts are viewed as tools of resistance that critique official doctrines propagated by hegemonic powers. To archaeologists, public transcripts are evident in tools of official propaganda, including written histories, public monuments, impressive urban centers, and material markers associated with the state. In contrast, hidden transcripts are obscured by public transcripts and have subtler material markers that can include the maintenance of local symbols of itentity despite hegemonic pressure to adopt materials linked to the state. Alternative narratives are a form of hidden transcript because they exist privately and in opposition to official understandings of the past perpetuated by those in power.

I argue that alternative narratives are not just a possibility but also an inevitable product of the relational and constructive nature of collective remembering. Just as spectacle, public space, and collective remembering may result in different memories of the past across socioeconomic classes, settlement patterns facilitate differential intensity of interactions, whereby proximity may play a critical role in the development of mnemonic communities (Zerubavel, 1996, 1997). I follow Wertsch (2002, p. 67) in arguing that “collective remembering typically provides an essential basis for the creation and maintenance of groups—specifically, imagined communities.” In pre-industrial societies, like the ancient Maya, collective remembering is dependent on face-to-face interactions and is inherently framed by the spatial and material parameters that constrain interaction. In other words, individuals have greater opportunity to discuss the past—and agree or disagree—with those they interact with frequently. The social geographies that structure individuals’ daily lives—such as neighborhoods, villages, rural estates, and minor centers—may serve as natural fault lines for the construction of mnemonic communities, just as they do other kinds of communities explored in archaeology (Ardren, 2015; Hutson, 2010; Yaeger & Canuto, 2000).

At the community level, the perception of a collective past may be materialized through the construction of village centers or community gathering points associated with village life. Importantly, as long-lived units of social identity and interaction, a community’s collectively remembered past might not concur with official narratives or the alternative narratives maintained by nearby communities. At rural centers away from the polity capital, inhabitants would have been more able to safely express opinions contrary to the official narrative. Thus, diverging interpretations of the past can arise in villages and secondary centers where geographic seperation allows for the maintenance of alternative narratives.

A Methodology for Approaching Collective Remembering

From spectacle, to conversation, to monument, a collective remembering approach views memory as resulting from a constructive and relational process. This process results in ever-evolving remembrances of the past that often diverge between social groups. Of course archaeological evidence for memory typically does not clearly articulate multiple understandings about the past. Instead, evidence of memory work is laden with power relations and most often reflects the views of those with political capital (Mills & Walker, 2008a; Overholtzer, 2013; Schwartz, 2013; Van Dyke, 2017). Untangling alternative narratives in the archaeological record can be difficult, even when it is clear that stakeholders with different viewpoints were present (Pauketat & Alt, 2003).

My methodology for identifying collective remembering and alternative narratives relies on identifying three things in the archaeological record. First, I identify communities established through frequent interaction or commonality in media consumption. These mnemonic communities form the social venue wherein collective remembering takes place.

Second, I trace the specific history of this community to understand the events and circumstances that the community may be remembering. Additionally, identifying the community’s changing social, political, and economic contexts is critical to understanding how the community’s remembrances may change as specifics fade over time.

Third, material evidence of memory points to how alternative narratives are constructed and deployed. Material evidence for “memory work” (Mills & Walker, 2008b, pp. 4–5)—social practices that evoke the past—can be found both within mundane everyday contexts and public settings. Although evidence of memory work is quite common in the archaeological record broadly (chapters in Mills & Walker, 2008a) and among the Maya specifically (Gillespie, 2010), alternative narratives, by their very nature, are often clandestine and therefore more difficult to identify materially. That said, alternative narratives can be identified through at least two approaches. One approach looks for direct material evidence of resistance within private contexts, while the other looks to identify the repudiation of previous symbols of power during periods of political transition.

One approach traces material evidence for alternative narratives maintained within households and other private contexts (Overholtzer, 2013; Overholtzer & Bolnick, 2017). Similarly, commoner collective remembering may be evident in informal graffiti or rock art (Pool & Loughlin, 2017). Private or anonymous forms of material expression provide direct evidence for alternative narratives that existed contemporaneously with, and sometimes in resistance to, the dominant narratives of those in power.

A second approach focuses on periods of societal transition when power dynamics come into question. Scott (1990, p. 206) describes moments of “political electricity,” where hidden transcripts are expressed publicly for the first time, often in opposition to those in power. In the archaeological record, hidden transcripts and alternative narratives may be visible in the actions of commoners and the previously - subjugated during periods of political collapse and transition. As hegemony weakens, previously subordinate groups may feel emboldened to express themselves and to draw on alternative narratives as they rebuild. In this way, narratives that once were alternative may become dominant in new social circumstances.

I draw on this second form of material evidence as an indicator of alternative narratives present at Actuncan. The dynamic nature of Maya politics (Marcus, 1993; Martin & Grube, 2008) included multiple transitions in regimes and political systems evident in both the epigraphic and archaeological records. Within ancient Maya cities, monuments and public architecture are durable, material remnants of past regimes (Navarro-Farr & Rich, 2014; Stanton & Magnoni, 2008). During transitional periods when alternative narratives and hidden transcripts may be declared publicly (Scott, 1990, pp. 202–227), these old spaces provide a venue for communities and leaders to perform their feelings towards the past. The ancient Maya often left material traces of these newly dominant narratives through deposits of discarded (or intentionally placed) artifacts and through renovations and other modifications to the urban landscape (Iannone et al., 2016; Schortman & Urban, 2011; Strockett, 2010). At Actuncan, the Terminal Classic community left remains of ritual deposits and alternatively desecrated and reutilized old buildings within the site core. I draw on the material evidence of these actions to trace evidence for collective remembering and to argue that local alternative narratives played a critical role in legitimizing the site’s new political institutions.

Collapse, Reorganization, and Memory Among the Ancient Maya

The Maya collapse of the Late and Terminal Classic periods was a major sociopolitical transition and as such provides a particularly fertile time period for the investigation of collective remembering. During the process of collapse, the Classic Maya political system, in which regional city-states ruled by divine kings had competed with each other for prestige and control of economic resources, disintegrated (Demarest et al., 2004; Freidel & Shaw, 2000; Martin & Grube, 2008). Initially, the collapse was accompanied by regional drought (Hodell, Curtis, & Brenner, 1995; Iannone, 2014; Kennett et al., 2012; Yaeger & Hodell, 2008) and widespread warfare (Demarest et al., 1997; Inomata, 2003; Palka, 2001). Ultimately, the rejection of divine kings may have resulted from a failing faith in the ability of kings to ensure access to food, security, and other critical resources (Freidel & Shaw, 2000; Iannone et al., 2016).

Although the process of collapse and reorganization took place over the course of three centuries and varied regionally depending on the timing and course of local events (Aimers, 2007; Demarest et al., 2004), broad trends are evident in the demographic, economic, political, and ritual transformations that took place. Most dramatically, evidence points to a substantial redistribution of lowland Maya populations. Broadly, populations in the southern Maya lowlands slowly declined in the Terminal Classic period (Turner, 1990) as people emigrated to the northern Maya lowlands and to coastal and riverine settlements in Belize and the Petén Lake region (e.g., Masson, 1997; Pendergast, 1986; Schwarz, 2009). These locations were easily accessed by canoe and served as gateways to the increasingly important pan-Mesoamerican trade networks (Berdan, Masson, Gasco, & Smith, 2003; Sabloff & Rathje, 1975). Some sites located along the Caribbean coast and major river systems may have been stabilized by participation in Chichén Itzá’s expanding trade network (Andres & Pyburn, 2004; Andrews & Andrews, 2003; Kepecs, 2007; Masson, 2000; Masson & Mock, 2004; Pendergast, 1986; Sabloff & Rathje, 1975). In some areas, attachment to Chichén Itzá’s mercantile network was accompanied by the introduction of a new pan-Mesoamerican religious cult dedicated to Kulkulkan (D. Z. Chase & Chase, 1982; McAnany, 2012; Ringle, Gallareta Negrón, & Bey, 1998).

Despite the clear impact of pan-Mesoamerican interconnectedness on Maya economic and religious institutions, political and social transformations were much more locally determined. Evidence indicates that individual lowland Maya communities gained increasing autonomy and adapted on an individual basis based on community-level needs (D. Z. Chase & Chase, 2006; Masson, 1997). Indeed, even though the impact of new long-distance relationship is evident, transformations in material culture largely reflect the adaptation of ideas that were already present locally (Schwarz, 2009, 2013).

This local innovation is apparent in the variety of political solutions local communities adopted to quickly normalize social relations following the collapse of regal political hierarchies. In some communities, new charismatic leaders attempted to revive the symbols and stations associated with divine kingship (Tourtellot & González, 2004; Źrałka & Hermes, 2012). In most areas, however, new leadership schemes were likely anchored in community-level negotiations between leaders of remaining kin groups.

In the absence of royal hierarchies, this sociopolitical transition provided an opening for non-royals to express their views about the past in a public forum. At some sites, kings and royal families were sacrificed (Buttles & Valdez, 2016; Demarest, Quintanilla, & Suansnavar, 2016), inscribed monuments were defaced (Harrison-Buck, 2016), and royal palaces were sacked and buried (Iannone, 2005; Schwake & Iannone, 2016), clearly marking the rejection of local divine leaders. In contrast, at the site of El Perú-Waka’, the non-royal residents of the site converted a pyramid previously dedicated to that site’s dynastic line into a public shrine representing the remembered importance of this place to the site’s remaining occupants (Navarro-Farr, 2016; Navarro-Farr & Arroyave Prera, 2014; Navarro-Farr, Freidel, & Arroyave Prera, 2008). The repeated ritual visitation by community members reflects an alternative interpretation of this space developed through collective remembering by mnemonic communities previously invisible in the archaeological record. The major political changes that took place during the Terminal Classic period provided space for expressions of memory maintained by groups other than Classic period royals and elites.

Memory and Sociopolitical Change at Actuncan

The ancient Maya site of Actuncan was occupied for about 2000 years, from the initial settlement of the hilltop around 1000 B.C. to its final abandonment around A.D. 1000 (Fig. 3). Throughout its occupation, Actuncan was one of several villages and, later, urban centers that vied for control of the valley’s labor, rich agricultural resources, and significant trade route (Fedick, 1989; Garber, 2004; Laporte, Adánez, & Mejía, 2008).

Fig. 3
figure 3

Map of the Mopan River valley showing sites mentioned in the text

Full size image

Although initially occupied in the terminal Early Preclassic, Actuncan was first constructed as a capital during the Late Preclassic period when an earlier village was completely covered by a new, formalized site plan. At this time, Actuncan’s public architecture was constructed in two formal civic-ceremonial groups located on adjoining hilltops (Fig. 4). Actuncan South was a triadic temple group and the site’s ritual center. It housed a large, visually dominant triad of pyramids focused on Structure 1, the 28-m-tall pyramid mentioned at the beginning of this article.1 Evidence from looter’s trenches indicate that multiple Preclassic versions of Structures 1 and 5 were decorated with monumental polychrome plaster masks (McGovern, 2004). Several stone monuments were also found in this plaza, one of which (Stela 1) was carved in a style diagnostic to the Late Preclassic period with the image of a figure (Fahsen & Grube, 2005, p. 79) reminiscent of the sacrificing individuals painted on the west wall of Las Pinturas Sub 1-A at San Bartolo, Guatemala (Taube, Saturno, Stuart, & Hurst, 2010, pp. 12–29). These early artistic programs indicate that Actuncan was likely the seat of a Preclassic divine king in the model of those identified at Cerros and San Bartolo (Freidel & Schele, 1988; Saturno, 2009).

Fig. 4
figure 4

Map of the Actuncan site core indicating the location of structures and household groups mentioned in the text

Full size image

Actuncan South is isolated from the much larger Actuncan North by a deep ravine. The site’s only known formal causeway (or sacbe) stretches across the ravine connecting the two parts of the site. In contrast to Actuncan South’s single plaza, Actuncan North is organized into five plazas bounded by monumental structures that likely served a variety of ceremonial, administrative, economic, and residential purposes. In particular, the structures surrounding Plaza F form an E-Group complex—an ancient Maya ceremonial complex associated with solar observation (Aimers & Rice, 2006; Aveni & Hartung, 1989). These ritual complexes were critical to the establishment of Preclassic ceremonial centers (A. F. Chase & Chase, 1995; Doyle, 2012). Additionally, several long and narrow multi-room structures, including Structure 19a, were likely administrative in function (LeCount, Wells, Jamison, & Mixter, 2016; Mixter, Jamison, & LeCount, 2013; Reents-Budet, 2001; Terry, Fernández, Parnell, & Inomata, 2004). The multiple functions of these urban spaces reflect the multifarious roles urban centers played in early Maya society. Most of the monumental architecture within the Actuncan site core was constructed in the Late and Terminal Preclassic periods with renovations continuing into the Early Classic period (Donohue, 2014; McGovern, 2004; Mixter, 2016, Table 2.2; Mixter et al., 2013; Simova & Mixter, 2016). Overall, Actuncan’s architectural layout and evidence of early artistic programs (the carved stela and polychrome painted masks) indicate that this site was the seat of authority in the upper Belize River valley during the Late and Terminal Preclassic periods.

Actuncan’s period of political preeminence was the first of four phases in the site’s political history that provide a series of lenses that distorted local collective remembering. The unusual density of ceremonial centers within the Belize and Mopan River valleys2 points to significant competition between the leaders of these centers and produced a dynamic sequence of political shifts (Ashmore, 2010; Driver & Garber, 2004; Helmke & Awe, 2012). In the Mopan River valley specifically, the major centers of Actuncan, Buenavista del Cayo, and Xunantunich were rarely, if ever, simultaneous seats of authority. Rather evidence indicates that authority shifted between the three locations due to changes in local political fortunes and the interference of regional hegemons located outside the valley (Ashmore, 2010; Ball & Taschek, 2004; A. F. Chase, 2004; Helmke & Awe, 2012; Helmke, Awe, & Grube, 2010; Kurnick, 2013; LeCount & Yaeger, 2010a; Leventhal & Ashmore, 2004; Peuramaki-Brown, 2012). These political shifts and their impact on Actuncan are summarized in Table 1.

Table 1 Reference chart describing the chronology of Actuncan’s occupation
Full size table

Because of the residential settlement that established itself within Actuncan’s urban core during the site’s political apogee, Actuncan was a population center in addition to a ceremonial center. Two types of domestic structures—patio-focused groups and tall single-mound houses integrated into the urban plan—were founded during the site’s Preclassic apogee, pointing to the establishment of a class hierarchy with at least two tiers (LeCount, 2004; LeCount & Blitz, 2005; LeCount, Keller, & Blitz, 2011; Mixter, Fulton, Bussiere, & LeCount, 2014). Although the elite single-mound houses were quickly abandoned, the patio-focused groups proved to be durable residences, and the households that occupied them were witness to the political transitions that followed. From Actuncan’s apogee to abandonment, the site served a variety of roles in the local polity as a capital; hinterland village; secondary administrative center; and, finally, a capital again.

The second phase in this political history began during the Early Classic period when Actuncan failed as a political capital. Sometime during the third or fourth century A.D., nearby Buenavista del Cayo (located 4 km to Actuncan’s south) surpassed Actuncan in power and came to dominate the region (Ball & Taschek, 2004). Actuncan’s elite houses were abandoned and construction on its monumental architecture largely halted. Importantly, Actuncan was not abandoned—its commoner patio-focused groups remained occupied. Rather, the site’s monumental architecture and public spaces appear to have temporarily fallen into disuse, leaving Actuncan as a hinterland village situated among a degrading monumental core.

The third phase in Actuncan’s political history began with the establishment of Xunantunich (located 2 km to Actuncan’s north) as the preeminent regional power during the Late Classic I period (LeCount, Yaeger, Leventhal, & Ashmore, 2002). Evidence from the region clearly shows that Xunantunich eclipsed Buenavista del Cayo as the seat of local power (Kurnick, 2016; LeCount & Yaeger, 2010b; Leventhal & Ashmore, 2004; Peuramaki-Brown, 2012; Yaeger, 2008). At this time, Xunantunich seems to have utilized Actuncan as a secondary administrative center and, I argue, as a source of legitimization. A noble was installed at Actuncan within a multi-patio palace complex (distinctive for its small size), which served as both a residence and center of administration (LeCount et al., 2016; Mixter et al., 2013). Additionally, evidence indicates that some of some of the monumental architecture, which appears to have weathered during the Early Classic period, was refurbished at this time (Mixter, 2016, p. 299; Mixter et al., 2013).

Finally, during the Terminal Classic period, Actuncan was reestablished as the local center of political authority. As the Xunantunich’s divine kings declined in influence as part of the broader Classic Maya collapse, Actuncan’s residents initiated an urban revitalization marked by a renewed program of monumental construction, focused on the construction of a council house, and a reconceptualization of the city plan (LeCount et al., 2011; Mixter, 2016).

Between the end of Actuncan’s initial failure as a center of authority during the Early Classic period and Late Classic interest of Xunantunich’s rulers in Actuncan (a period of 200 to 300 years), very little evidence exists for ritual activity or monumental construction within Actuncan’s urban core. Given the apogees of nearby Classic period capitals, this gap in ritual activity likely reflects Actuncan’s subjugation during much of the Classic period. The dominant powers of this time period had a vested interest in suppressing local affiliation with Actuncan, instead redirecting public ritual to legitimize their own rule. In this context, the reestablishment of Actuncan as a center of political authority reflects the outcome of a long-term process of collective remembering. With the widespread collapse of divine kingship in the region, the Actuncan community reasserted itself as the local center of political authority bolstered by a locally-maintained alternative narrative of the past.

Settlement in the Belize River Valley: Small Communities with Long Memories

Perhaps owing to the availability of water and good farm land provided by the annual flooding of the Belize River and its tributaries (including the Mopan River), settlement in the upper Belize River valley is relatively uniformly dispersed, following pockets of the best soil along the valley margins (Fedick, 1989, 1995; Smith, 1998; Willey, Bullard, Glass, & Gifford, 1965). Research by the Xunantunich Settlement Survey indicates that households cluster into settlement zones that focus around a number of villages and minor centers strung across the Mopan River valley (Yaeger, 2010). Many of these minor centers, villages, and settlement clusters, which include Chan, Callar Creek, Chaa Creek, Rancho San Lorenzo, Dos Chombitos, and Nohoch Ek, were occupied as early as the Middle Preclassic period until the Postclassic period, indicating the deep antiquity of these outlying settlement zones (M. K. Brown, Awe, & Garber, 2009; Connell, 2000; Ehret, 1995; Kosakowsky, 2012; Robin, 1999; Yaeger, 2000a). Patterns of household growth indicate that the earliest established households controlled critical resources leading to the greatest stability. At Chan and San Lorenzo, the most intensively investigated of these settlements, the oldest households tended to be the last households abandoned during the ninth-century Maya collapse period (Ashmore, Yaeger, & Robin, 2004; Robin et al., 2012c).

Importantly, the settlement clusters spread across the valley are not homogenous in their arrangement or composition. Some, like Chan, Chaa Creek, and Callar Creek, consist of a variety of household clusters that radiate from a formal minor center (Connell, 2000, 2003; Kurnick, 2013; Robin, 2012, 2013; Yaeger, Kurnick, Dykstra, & Peuramaki-Brown, 2012). Others, like Nohoch Ek, likely served as the estates of rural elites (Taschek & Ball, 2003). Yet even modest villages, like San Lorenzo, and suburban neighborhoods, such as Buenavista del Cayo’s south settlement zone, were organized around modest non-residential structures that likely served as gathering points and community centers (Peuramaki-Brown, 2013; Yaeger, 2000b).

In their review of the principles that underlie community formation, Yaeger and Canuto (2000) focus on interaction as critical to the construction of communities. Although they caution that communities need not be spatially discrete, the kind of clustered settlements recorded in the Belize River point to frequent interactions and shared practices that often underpin communities. In fact, for the San Lorenzo community, Yaeger (2000b) argues that shared habitus and practices of affiliation associated with resource procurement and village-wide gatherings, such as feasts, emphasized the sameness of the village’s residents. The construction of community centers, such as the Chan ceremonial core and SL-13 at San Lorenzo, would have reified community connections as both centers of interaction, community gathering, and materializations of a common community past (Cap, 2012; Kosakowsky, Novotny, Keller, Hearth, & Ting, 2012; Yaeger, 2000b). At Chan’s central group, the repeated deposition of ritual offerings and burial of important ancestors point to the kind of events that would have structured the community identity around a common past (Kosakowsky et al., 2012). The discovery of caches and other deposits at Chan is a particularly striking example of community remembering because they were not likely placed by a ruler or ruling family. Instead, Cynthia Robin and colleagues (Robin, Kosokowsky, Keller, and Meierhoff 2014; Robin, Meierhoff, and Kosakowsky 2012b) have argued that power in Chan was shared by community members rather than centralized in the hands of a single individual. Community-initiated acts of collective remembering like these speak to the imagined quality (sensu Anderson, 1991) of the these communities; as rural settlements, these kinds of public displays result from the organization of residents who self-identify as part of a community fixed in a common past. While it is clear that members of hinterland communities were affiliating with larger polity capitals (Connell, 2010; LeCount, 2010a; Yaeger, 2000b, 2003), they also maintained their own local community identities, visible both in settlement patterns and subtle variability in material culture (Preziosi, 2003), that were rooted in relations often older than the political capitals that ruled them.

During its history, Actuncan was a major center, the Preclassic capital of the Mopan River Valley, and a community center for the households located in the site’s urban core. Next, I examine Actuncan’s founding and the trajectory of the households that remembered together for at least 1000 years.

The Preclassic Foundations and Residential Persistence of Actuncan’s Community

Since 2001, the Actuncan Archaeological Project, directed by Lisa LeCount, has investigated the growth trajectories of households located within Actuncan’s urban core to understand how these houses were impacted by, and implicated in, the site’s dynamic historical trajectory (LeCount, 2004; LeCount & Blitz, 2005; LeCount et al., 2011; Mixter et al., 2014; Mixter et al., 2013). Actuncan’s primary residential neighborhood is located to the north of Actuncan’s ceremonial core (Fig. 4). This neighborhood contains six patio-focused residential groups visible from the modern ground surface. Three of these household groups were founded during Actuncan’s apogee in the Late and Terminal Preclassic periods, while the other two were likely founded during the Early Classic period (Fig. 5). Importantly, our evidence indicates that all the households were occupied continuously through the end of the Late Classic period. Additionally, four of the six households continued to be occupied in the Terminal Classic period.3 These data point to the longevity and resilience of Actuncan’s commoner households in the face of a dynamic regional political landscape (Mixter et al., 2014).

Fig. 5
figure 5

Figure indicating the periods of occupation of Actuncan’s six investigated household groups (data from Antonelli and Rothenberg (2011), Hahn (2012), LeCount (2015), LeCount and Blitz (2001), LeCount, Blitz, and Kelso (2005), Rothenberg (2012), and Simova (2012))

Full size image

Similar to other settlement clusters located in the Mopan River Valley, Actuncan’s long-lived households would have formed a community simply by living within a circumscribed area for centuries. Kara Fulton (2015; also LeCount & Blitz, 2012) has argued that this community was held together in part by affiliative practices of feasting and ancestor veneration focused on Group 1, the largest and oldest household at the site. This pattern is similar to the kind of feasting hosted by the oldest and wealthiest households in nearby San Lorenzo (Yaeger, 2000b, pp. 131–133). However, in contrast to the other communities in the Mopan River Valley, the residents of Actuncan’s northern neighborhood were not served by a modest village center or gathering place. Instead, the Actuncan community was located immediately adjacent to a monumental architectural complex. As the members of the Actuncan community met on the way to and from the field or market, the site’s massive pyramids and broad public plazas would have shadowed them. These massive mnemonics would have structured their interactions and remembrances. Whether or not the Actuncan community continued to understand the original historical context of these structures’ construction and use, the continued occupancy of the site would have ensured that the monuments’ existence was not forgotten. Further, the continuous occupation of the Actuncan settlement would likely have tied local community identity to the remembrance of this landscape and to the stories, legends, or ancestors attached to these abandoned pyramids (see also Halperin, 2014; Stanton & Magnoni, 2008).

Importantly, the past as remembered by the members of the Actuncan community was not straightforward. When monumental construction halted at Actuncan in the Early Classic period, Actuncan’s elite households were abandoned, leaving only the occupants of the northern neighborhood to maintain the memory of Actuncan’s past greatness. Because of Actuncan’s remarkably monumental landscape, a distinct mnemonic community likely formed in the northern neighborhood around narratives explaining the remains of these Preclassic pyramids. Yet, over the neighborhoods’ 1000 years of occupation, the content of these narratives would have drifted through each retelling, not only impacted by mistakes but also influenced by the contemporary context of current events and competing narratives.

Actuncan Under Foreign Rule: Official Narratives in the Late Classic

During the Early and Late Classic periods, both Buenavista del Cayo and Xunantunich developed official narratives of the local past that were advantageous to their own power. As part of these polities, Actuncan’s community members would have been exposed to these official narratives. In this article, I primarily focus on Xunantunich rather than Buenavista del Cayo.4 Following Ashmore (1998, 2010), I argue that Actuncan’s antiquity was incorporated into the official narrative Xunantunich used to legitimize its authority. However, the official narratives of these local lords were not accepted by hinterland communities without question.

Xunantunich: Actuncan’s Late Classic Overlord

Beginning in the seventh century, Xunantunich grew from a hilltop occupied by a single platform and the remains of a few Preclassic dwellings to a great settlement featuring 3 plazas, many pyramids, and numerous stone monuments, several of which were carved with hieroglyphic inscriptions (Helmke and Awe, 2016a, b; Helmke et al., 2010, p. 98; LeCount et al., 2002). The dramatic rise of Xunantunich may have been spurred by the influence of larger regional hegemons and appears to have come at the expense of Buenavista del Cayo. Additionally, this shift in power was accompanied by substantial demographic shifts within the Mopan River valley. Evidence points to migration away from the settlements around the failing Buenavista del Cayo (Peuramaki-Brown, 2012; Yaeger, 2010). Otherwise, the population grew uniformly across the valley with new households established in most of the other old settlements (Robin et al., 2012c; Yaeger, 2010).

Xunantunich’s layout was designed to imply cosmological connections to the four corners of the Maya world and political connections to more powerful regional hegemons (Ashmore & Sabloff, 2002). The most important indicator of Xunantunich’s new power was the Castillo, a 39-m-tall pyramid complex (Leventhal, 2010). The Castillo was visible across the valley, just like Actuncan’s Structure 1. The construction and visibility of the Castillo edifice would have established the power of Xunantunich’s leaders and fixed their ability to secure regional labor in people’s memory. Furthermore, Virginia Fields (2004) has argued that the stucco friezes located at the summit of the Castillo reflect the site’s royal charter. Finally, the site’s carved monuments represent the most explicit propagation of an official narrative. The images and texts carved on these monuments recorded the official history of specific rulers’ deeds and geopolitical importance (Helmke et al., 2010). The monumental architecture and artistic programs in the Xunantunich city center form the foremost material manifestation of the site’s official narratives.

Xunantunich and Actuncan

In addition to material manifestations of legitimizing narratives within the Xunantunich site, Wendy Ashmore (1998) has argued that the rulers of Xunantunich claimed Actuncan as their ancestral home. Because Xunantunich was rapidly constructed as a royal capital, their claims for legitimacy were partially rooted in the creation of a narrative of descent from the region’s oldest divine kings. Several lines of evidence point to Xunantunich’s adoption of this official narrative. First, Xunantunich was constructed to the south of Actuncan, so that Actuncan would be to the site’s north, the direction where the ancestors reside (see Fig. 3; Ashmore, 1998). The choice to place Xunantunich in this directional relationship with Actuncan would have invoked Maya cosmological principles to emphasize the narrative that Actuncan was ancestral to Xunantunich (Ashmore, 1989, 1991). Second, a portion of a sacbe, a formal processional road, that connected Xunantunich to Actuncan was mapped by the Xunantunich Settlement Survey (Ashmore, 1998, p. 174). The construction of this sacbe connecting Xunantunich to Actuncan would have facilitated formal processions between the two city centers and is an indication that Xunantunich claimed Actuncan as part of its territory. Third, Xunantunich’s ancient name may point to an association with Actuncan. Christophe Helmke (Helmke & Awe, 2012, p. 68) has read the Xunantunich emblem glyph as the toponym Katyaatz Witznal, which he glosses as “clay-bearing mountain.” Xunantunich was constructed on top of a tall limestone hill; however, there is currently no evidence that this hill produced clay. Actuncan, on the other hand, was constructed on a T-3 alluvial terrace above the Mopan River. Because of the depth of this natural alluvial clay deposit, our excavations have never reached bedrock. This indicates to me that the rulers of Xunantunich may have borrowed the ancient name of Actuncan for their new polity—creating a direct and explicit reference between the two.

Additionally, evidence of construction at Actuncan points to direct attempts by Xunantunich to exert its authority over the Actuncan community. During the first part of the Late Classic period, Actuncan’s largest Preclassic administrative structure, Structure 19a, was converted into Group 8, a multi-patio household compound reminiscent of the large palace complexes inhabited by kings and elites at other Late Classic sites (Fig. 6; Mixter et al., 2013). However, the palace at Actuncan is oddly proportioned for a ruler or even an important elite. Even though the household includes three patios, all the constituent mounds aside from Structure 19a are quite low. In contrast, Structure 19a had a corbel-vaulted roof and sat on a 6-m-tall platform, all constructed centuries earlier. Based on chemical data retrieved from the floors of Structure 19a, we have argued that in the Late Classic period, it served primarily as an administrative space for hosting small events and storing ritual goods (LeCount et al., 2016). Based on these data, my colleagues and I (Mixter et al., 2013) have argued that Structure 19a and its associated dwellings were occupied by an administrator placed at the site by Xunantunich.

Fig. 6
figure 6

Map of Structure 19a and Group 8 layout showing location of excavations

Full size image

To secure their position within the Actuncan community, the members of Group 8 built a household that appeared much older than it was. They constructed Structure 22, a large shrine on the eastern side of the largest patio in the position usually reserved for ancestor shrines in Actuncan’s households (Fig. 6). However, our excavations revealed that the shrine held no ancestors (Mixter & Freiwald, 2013). Instead, Structure 22’s height and the Group 8’s multiple patios created an appearance of deep history and legitimacy when, in reality, the residents of this household were the newest members of the local community (see Overholtzer & Bolnick, 2017 for another example of a household exaggerating its antiquity). This act speaks to a very intentional attempt to construct memory and influence collective remembering. By occupying Structure 19a and building a multi-patio household—the only one at Actuncan—with an overly large but empty ancestor shrine, the residents of Group 8 created the appearance that they were deep-rooted members of local community while simultaneously differentiating themselves through their elaborate palace architecture.

In addition, there is evidence that the rulers of Xunantunich were directly interested in Actuncan’s most visible ritual architecture, its triadic temple complex located in Actuncan South. Following two to three centuries of neglect during the ascendency of Buenavista del Cayo, Actuncan’s monumental architecture was likely in poor condition. Indeed, excavations by Thomas Jamison (2013; Mixter et al., 2013) indicated that Structure 19a’s corbel vault was near collapsing prior to its Late Classic period renovation and refurbishment. Investigations of looter’s trenches by James McGovern (2004) found that Structures 1, 4, and 5 were all renovated in the Late Classic period. Additionally, the plaster plaza floor was refurbished. In addition to their height, these pyramids may have held interest to Xunantunich because of their location at the opposite end of the site’s sacbe from Structure 19a (see Fig. 4). This sacbe was likely used as a causeway for ceremonial processions that metaphorically connect the residents of Structure 19a and Group 8 to the triadic temple group. In this way, the residents of Group 8—and by proxy Xunantunich—may have claimed to anchor their authority in the community’s memory of Actuncan South. Although we have found little direct evidence for Late Classic period ritual activity in these spaces, the evidence for refurbishment suggests that the residents of Group 8 and the hegemons at Xunantunich saw value in revitalizing these old spaces.

Placing a local administrator at Actuncan would have benefited Xunantunich’s rulers by advertising their claim to the earlier site’s rulers. Part of this claim may have been to create the appearance that their vassal living in Group 8 was a well-established member of the Actuncan community connected to the site’s most visible old pyramid complex. Of course, members of the Actuncan community likely remembered the real impetus for the construction and occupation of Group 8 and may have resented the attempt by Xunantunich to rewrite their past.

Symbols of Subjugation and Resistance in the Xunantunich Hinterland

That said, at Actuncan and other Belize River valley settlements, there is a strong indication that the narratives propagated by Xunantunich were accepted, at least superficially. LeCount (2010a) has identified the Mount Maloney Black ceramic type as a clear marker of affiliation with the Xunantunich polity. This ceramic type appears in the Late Classic period and becomes the most common slipped pottery type in this portion of the Mopan River valley. However, at sites outside this region, the type is relatively rare (Gifford, 1976). In the Xunantunich center, Mount Maloney pottery is ubiquitous in ritual deposits, and sherds were used in the fill to protect the Castillo’s stucco friezes when they were covered during a renovation event (LeCount, 2010b). At hinterland settlements, Xunantunich’s rise to power is evident in a gradual transition from a ceramic assemblage dominated by red-slipped pottery to one dominated by black-slipped pottery including Mount Maloney Black. This transition likely points to communities’ decisions to affiliate themselves with Xunantunich rather than other contemporaneous capitals (Connell, 2000, 2010).

While the use of Mount Maloney pottery marks the adoption of Xunantunich’s official narrative, some evidence indicates that, for at least some members of the Xunantunich polity, this acceptance was largely a public transcript that obscured collectively remembered hidden transcripts. Although all households within the Xunantunich polity possessed some Mount Maloney Black ceramics, there is evidence from Chan that enthusiasm varied for this black-slipped symbol of Xunantunich’s rule. Interestingly, households founded during Xunantunich’s rule possess a high percentage of Mount Maloney Black pots, while Chan’s older households maintain a greater balance between red- and black-slipped pots (Kosakowsky, 2012; Robin, 2013, pp. 169–171). This indicates that older households and communities were actively maintaining their deep-seated, collectively remembered traditions. Maintaining larger quantities of red wares constitutes a subtle kind of resistance—a hidden transcript to the public transcript of black wares.

Similar to Chan, Actuncan was an ancient, continuously occupied community. In both communities, proximity to common public facilities enabled intensive intracommunity interactions. It is no accident that the oldest households limited their adoption of black-slipped ceramics. These are the households that have interacted the longest and developed the strongest local concept of community. Chan provides an example of not only how official narratives impacted the practices of local communities but also how these communities resisted. Although we have little evidence from Actuncan for direct resistance to Xunantunich during the Late Classic period, evidence from the Terminal Classic period, contemporaneous with the failure of Xunantunich, points to how subjugation by the Late Classic hegemon impacted collective remembering at the older site.

Evidence of Collective Remembering: Actuncan in the Terminal Classic Period

As part of the widespread phenomenon of the Maya collapse, during the ninth century, the divine kings of Xunantunich failed and disappeared. Evidence from the Xunantunich Settlement Survey indicates that the valley began to depopulate by the beginning of the Terminal Classic period around A.D. 780 (Ashmore et al., 2004). Even as Xunantunich continued to create history and attempted to manipulate memory through the erection of carved monuments, all its monuments were dragged into a single public plaza, a wall was constructed around it, and other public plazas were largely abandoned (Jamison, 2010). Simultaneously, evidence from other centers indicates that their communities gained new autonomy. Elaborate Terminal Classic tombs have been encountered at both Buenavista del Cayo and Cahal Pech that appear to point to claims of royal status—claims that would have contested Xunantunich’s hegemony (Awe, 2013; Helmke, Ball, Mitchell, & Taschek, 2008; Yaeger, 2008).

A different kind of political revival took place at Actuncan. As Xunantunich’s power waned, a new program of monumental construction was initiated at Actuncan. In particular, a large public platform was constructed as a novel civic center within the ceremonial core, and new smaller platforms were constructed within Plaza A (Mixter, 2016). These constructions marked a return of political authority to Actuncan, likely led by a council of community leaders. As previously discussed, the population of Actuncan’s community remained relatively stable during the Terminal Classic period when other settlements were experiencing depopulation. The households that remained had been occupied continuously since the Preclassic period (Mixter et al., 2014). Evidence from the Terminal Classic period indicates that the local households had formed a community based on centuries of interaction and common practices (Fulton, 2015). With the return of political authority to Actuncan, renewed ritual activity in association with the site’s Preclassic architecture provides evidence for collective remembering.

Evidence of Collective Remembering at Actuncan

Evidence for Actuncan’s collective remembering comes from my interpretation of three data sets. First, evidence for construction and smashed ceramic deposits within Actuncan South points to a renewal of ritual activity within the triadic temple complex. Second, AAP excavations on Preclassic monumental architecture in other parts of the site center have not encountered evidence for Terminal Classic ritual or reoccupation. Instead, excavations indicate that monumental architecture in Actuncan North was dismantled, not reused. Third, our excavations in Group 8 and on Structure 19a indicate that this space was ritually terminated and Structure 19a dismantled during the Terminal Classic period, after its utilization as a noble residence in the Late Classic period. Many of the patterns described could be explained using other frameworks and likely were the result of complex and multi-faceted social processes. Among these possibilities, I argue that these data sets demonstrate Terminal Classic collective remembering of both the site’s distant Preclassic apogee and, also, the more recent Late Classic period of political subjugation.

First, investigations by the McGovern (2004) and I (Mixter, 2016) in Plaza A indicate that the triadic pyramid complex was the venue for renewed ritual interest during the Terminal Classic period. Although refurbished during the Late Classic period, material remnants of ritual activity all date to the Terminal Classic period. The clearest evidence of Terminal Classic ritual in the triadic group comes from Structure 5, the eastern structure of the triadic groups, originally constructed during the Late Preclassic period. During investigations of Actuncan’s looter’s trenches, McGovern (2004, p. 159) described a thick deposit of burned ceramic material located on the floor of the summit building’s final version.

“A 0.6 to 0.75-m-thick layer of thousands of burnt Late Classic II and Terminal Classic sherds, broken but complete bowls, dishes, and vases, and charcoal resting on the floor and stairs…. They were obviously smashed and burnt in what can only be considered a termination ritual” (McGovern, 1994, pp. 112–113).

To estimate the deposits’ antiquity, McGovern (2004) collected over 6600 ceramic sherds from the wall of the looter’s trench. This a collection that reflects only a small fraction of the deposit’s original volume, some of which was excavated by looters and much of which remains in place today.5 Similar Terminal Classic period deposits have been identified across the Maya lowlands (A. F. Chase & Chase, 2004; Houk, 2016; Mock, 1998; Navarro-Farr & Arroyave Prera, 2014; Navarro-Farr et al., 2008; Stanton, Brown, & Pagliaro, 2008). Rather than the single ritual suggested by McGovern (1994, p. 113), to me the size and density of this deposit point to an accumulation of materials over time, reflecting repeated ritual depositions on Structure 5 by community members during the Terminal Classic period.

Additionally, excavations I directed in the summer of 2013 investigated three long, low platforms constructed in Plaza A of Actuncan’s main triadic group (Fig. 7; Mixter & Langlie, 2014). Ceramics recovered from these structures indicate that they were constructed in the Terminal Classic period. In front of Structure 8, the northern platform, we encountered a deposit of partially reconstructable serving vessels left on the building’s step (Fig. 8a). It is not clear if these were smashed in antiquity or left whole and broken after deposition. The ceramics in this deposit included at least two McCrae impressed dishes, a large unslipped jar with a flared lip and a fragment of an imitation fine orange vase (Fig. 8b). These ceramics are all locally diagnostic to the Terminal Classic period. Additionally, the deposit included about two thirds of a jade bead. The prevalence of serving vessels in this deposit and its location on the building’s step indicate that the deposit was likely the smashed remains of a meal and may point to repeated communal food consumption within the patio space created by the construction of these low platforms. To me, the presence of serving vessels in this deposit suggests that the rituals taking place in plaza were inclusive community events intended to cement a collective identity. Considered together, the deposits from Structure 5 and Structure 8 indicate that Plaza A was used as a venue for inclusive community-building rituals anchored in an explicit reverence of the site’s Preclassic apogee, symbolized by Structure 5.

Fig. 7
figure 7

Map of Actuncan South showing the location of Structures 1, 5, and 8

Full size image
Fig. 8
figure 8

Terminal Classic feature associated with Structure 8. a Photograph of the feature in situ on the edge of the Structure 8 platform. b Photographs of ceramics that are locally diagnostic to the Terminal Classic period from the feature including two different McCrae impressed dishes, a pie crust jar rim, and a partially reconstructed flared lip jar rim

Full size image

Second, we have encountered little evidence of Terminal Classic reverential ritual in public spaces outside of the triadic group. I argue that the choice of the triadic group as their primary locus of ritual derives from collective remembering. Importantly, all of these deposits are located in the shadow of Structure 1, the site’s tallest pyramid and most visible testament to Actuncan’s past greatness. By reinitiating rituals in this zone, the community is remembering the site’s early greatness and drawing on this past to legitimize its new power.

Why then was the triadic group specifically singled out for reverence and not other Preclassic structures such as the E-Group? Excavations on top of Structure 26, the E-Group’s eastern range structure, found little evidence of concerted Terminal Classic ritual engagement. There, evidence of Terminal Classic activities is limited to a few scattered sherds on the surface and perhaps the scattering of chert eccentrics on top of building collapse (Donohue, 2014). Our findings indicate that the E-Group had not been maintained for some time prior to the Terminal Classic period. On the contrary, the façades of all structures associated with the E-Group appear to have been systematically mined for stone during the Terminal Classic period. We found few stones in situ from the final staircases of Structures 23 and 26 and no evidence of cut stones in the collapse at the buildings’ bases (Donohue, 2014; Heindel, 2016; Simova & Mixter, 2016). Excavations into the nearby Group 4 indicate that cut stones were likely taken during the Terminal Classic period when they were used as fill in the construction of that monumental Terminal Classic platform (Mixter, 2016). Because the stones removed from Structures 23 and 26 appear to have been immediately reutilized in the construction of Group 4, I argue that this dismantlement indicates apathy towards these structures rather than a message-laden desecration. I argue that this treatment is evidence for a kind of unintentional forgetting—a “structural amnesia” that is inherent in collective remembering (Connerton, 2008). Overtime, the importance of the E-Group was forgotten, whereas the triadic group’s memory was perpetuated.

Third, the Terminal Classic Actuncan community also collectively remembered the period during which it was subjugated by Xunantunich. If Xunantunich established the residence of a vassal administrator at Actuncan to strengthen its claim to Actuncan’s past, then Actuncan’s residents explicitly rejected that connection after Xunantunich’s fall. Structure 19a, the five-room administrative building connected to the administrator’s residence, was abandoned by the Terminal Classic period. Sometime in the final days of the Late Classic period, the building’s corbel vault was intentionally destroyed, the building’s vault stones were placed in its doorways, and the rooms were filled and capped by a layer of plaster (Mixter et al., 2013). When plaster or sascab is used to seal a room, the act is typically considered a desecratory (Pagliaro, Garber, & Stanton, 2003) and, I would argue, represents the community’s ritual killing of the residence of their oppressor. This action would have served as an explicit rejection of Xunantunich’s official narrative and any claim its rulers might have had on Actuncan’s history.

Furthermore, the resumption of ritual activity in the triadic group reflects the emergence of hidden transcripts in the form of alternative versions of the past collectively remembered by the local community. There is little evidence of ritual activity around the triadic group following the ascendance of Buenavista del Cayo in the Early Classic period, presumably because Buenavista’s elites limited the Actuncan community’s formal ceremonial activities. Instead, the community was likely coerced or cajoled to attend events hosted at Buenavista del Cayo, and the triadic group was left to return to the forest. During this two- to three-century gap in use, association of these pyramids with specific rulers and events likely faded from memory.

Christina Halperin (2014) has argued that forested ruins were viewed by the ancient Maya as effigy mountains and the homes of ancestors. The evidence from our research in the triadic group supports this idea. The major deposits of Terminal Classic ritual material were encountered on Structure 5, the eastern structure of the triadic group. Additionally, the arrangement of the low platforms built in the triadic group plaza refocused the plaza towards Structure 5, away from the physically dominant Structure 1 (Fig. 7). In the Classic period in the Belize River valley, royal ancestors were often buried in pyramids on the eastern edges of plazas (Audet, 2006; Awe, 2013; Healy, Hohmann, & Powis, 2004). Ironically, neither Structure 5 nor the remainder of the triadic group likely contained the remains of Actuncan’s ancestors. During the Preclassic period when these buildings were built, burials were rarely placed in the pyramids of triadic groups (Hansen, 1998, p. 89). So, the Terminal Classic reverential deposits placed on Structure 5 may be referencing ancestors believed to live within the triadic group, even though Preclassic practices suggest that no human remains were originally buried within that pyramid.

The memory and associations connected to the triadic group changed and drifted over time, even as its monumentality retained the memory of its past importance. After the collapse of Preclassic Actuncan, memory of the triadic group’s importance faded as the tropical forest over took the mounds. It would have become associated with a generic past, tied to the Actuncan community identity. Xunantunich may have attempted to coopt Actuncan’s legacy to its own purposes in the Late Classic period; however, Actuncan’s residents seem to have collectively remembered their own version of their past. When the failure of Xunantunich brought autonomy, the Actuncan community treated the triadic group as though it was a sacred place that could be harnessed in support of the site’s Terminal Classic political legitimacy. It seems likely that the Actuncan community collectively remembered a local version of their past that did not match Xunantunich’s official narratives. In this sense, the deposits found around the triadic group represent the transformation of these hidden transcripts into public transcripts, now in the service of Actuncan’s new, post-royal authority.

Conclusions

During Actuncan’s 2000 years of occupation, the site’s political context shifted dynamically depending on the amount of authority held by local leaders. In contrast, the daily life of Actuncan’s households remained quite stable from their founding. Over many generations, the site was occupied by descendants of the original settlers who passed down the narratives they had heard from their parents, grandparents, and neighbors. Members of Actuncan’s northern neighborhood passed each other every day in the fields, forests, and plazas. They exchanged stories and opinions on the events of the day. From the Late Preclassic onward, the Actuncan community met under the shadow of pyramids, first as members of the capital city, then as subordinate to Xunantunich. Through these interactions, both with each other and with Actuncan’s monuments, the memory of the site’s apogee and the treatment of the local people by the Xunantunich state were collectively remembered.

The evidence of memory from the Terminal Classic period reflects Actuncan’s collective remembering at a particular moment in time. In the historical context of the Maya collapse, the local divine kings had fallen in disfavor, but Classic period cultural norms remained largely intact. Within this context, the Actuncan community constructed an understanding of the past that derived both from that immediate context and from the site’s collectively remembered past. The triadic temple group became the location of reverence for residents, while the Late Classic palace was dismantled and desecrated. The motivations for these actions were likely multifaceted and likely include the Actuncan community’s disillusionment with the recently failed rulers at Xunantunich, irreverence for their vassals who lived in Structure 19a and Group 8, and nostalgia for Actuncan’s period of power and authority.

Evidence indicates that the Actuncan community did not understand the original function of Preclassic architecture they occupied. After at least 300 years of subjugation, the Actuncan community probably did not remember the specifics of the site’s apogee. Perhaps the Preclassic kings had been unpopular tyrants. If so, this memory was not remembered. Instead, the pyramids and plazas of these early kings became the major ritual venue in the Terminal Classic period. In turn, I infer that the site’s royal past was invoked to legitimize the authority claimed by the community in the Terminal Classic period.

As far as we know, Actuncan’s Preclassic past was not recorded in inscriptions, monuments, or any other media. Instead, it was remembered relationally through the continued interactions of community members with each other and their urban landscape. Considering remembering within this relational frame provides an opportunity to look at how collective memory both forms communities and impacts political relations. As a hidden transcript, the local memory of Actuncan’s past formed a foundation for the return of authority to the site in the Terminal Classic period. Through the reverential and desecratory acts in Actuncan South and Structure 19a, this constructed version of the past became public and thereby gained authority.

For archaeologists, collective remembering provides a tool for understanding how memory works to define and differentiate communities over time. I have shown how a constructed and relational approach to collective memory provides a way to bridge studies that explore official memory production and those that model memory produced in household practice. Though memories may be produced through official propaganda or passed down in the privacy of kitchens, these collective memories are accumulated and transformed through the conversations and interactions that take place in the course of everyday life. The folk interpretations of events produced through these interactions are critical because they need not conform to official narratives. Instead, alternative narratives often exist as hidden transcripts, invisible from view until changes in political circumstances allow for their public expression. Purposefully considering how the common people interpret and remember events may improve archaeologists’ ability to understand cohesion, resistance, and the reciprocal relationships involved in exerting authority.

Notes

  1. 1.

    Structure 1’s 28-m height measures from the base of the substructure to the top of the structure. Although not excavated, my personal observation of the mound suggests that the highest point of the extant building is likely the surface of a corbel-vaulted roof that would have been originally topped by a roof comb. McGovern (2004) found that the final version of this building was constructed during the Late Classic period. However, it was built on the same substructure as the Terminal Preclassic penultimate version. Although McGovern only identified the foundations of Structure 1’s penultimate superstructure, the similar height of their substructures indicates that the Terminal Preclassic version likely was similar in height or just slightly shorter than the final version. Both would have been easily visible across the Mopan River valley.

  2. 2.

    Actuncan is located along the Mopan River just above its confluence with the Macal River, which produces the Belize River. The region surrounding the confluence, which includes portions of all three rivers’ valleys, is known as the upper Belize River valley. In this paper, when I refer to the Mopan River valley, I am speaking about the section of the Mopan River valley between the Guatemalan border and its confluence with the Macal River. As such, the Mopan River valley is a subsection of the upper Belize River valley.

  3. 3.

    It is important to note that the same households, Groups 2 and 3, with no evidence of occupation before the Early Classic are the households with no evidence of Terminal Classic occupation. One possible interpretation of this pattern is that Actuncan follows the same demographic pattern seen elsewhere in the valley, whereby the most recently founded households are abandoned first (Ashmore et al., 2004; Robin et al., 2012c). However, these two groups were also subject to the most limited excavation (LeCount, 2004; LeCount & Blitz, 2001), so the absence of materials from before the Early Classic and after the Late Classic may instead be the result of the limited nature of these excavations.

  4. 4.

    Although I focus on Xunantunich, it is important to note that Buenavista del Cayo also impacted Actuncan and attempted to integrate the Mopan River valley hinterlands by propagating its own official narratives. Most obviously, the rise of Buenavista del Cayo to prominence was contemporaneous with the end of Actuncan’s apogee. We do not yet have clear evidence how this transition in power happened; however, it seems likely that Actuncan’s political diminution was a direct result of Buenavista’s rise to power. Additionally, evidence from Buenavista and the secondary center of Callar Creek suggest that Buenavista’s expanding influence was accompanied by the distribution of a set of materials and symbols directly associated with the Early Classic capital. These include the distribution of polychrome pottery produced at Buenavista (Reents-Budet, Bishop, Taschek, & Ball, 2000) and objects inscribed with the “Buenavista Device,” a symbol that Ball and Taschek (2004, p. 159) argue was exclusively associated with the Buenavista polity (Kurnick, 2016). We have not yet recorded evidence of these materials or symbols at Actuncan.

  5. 5.

    Charcoal recovered from the deposit by McGovern returned a 14C date of A.D. 690–987 at the 2σ level (LeCount et al., 2002, Table 3). This time span matches the Terminal Classic date of the deposit attributed based on ceramic styles. This date was calibrated by the author using OxCal 4.2 (Bronk Ramsey, 2009) on the IntCal 13 calibration curve (Reimer et al., 2013) based on the conventional age of 1175 ± 60 B.P. (AA-31355) published in LeCount et al. (2002).