Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Introduction

For the last few years, archaeologists at Flinders University, Adelaide, and staff at a former military hospital have been on the trail of an elusive structure: a subterranean air raid shelter (or series of shelters) built in 1942 to protect the lives of doctors, nurses, and their soldier patients in the event of a Japanese air raid. The search has attracted enormous public interest, as well as the attention of conspiracy theorists, convinced that the air raid shelters conceal military surplus and other secrets. The project has involved professional archaeologists, students, hospital staff, patients, veterans, and members of the local residential community. However, despite the investment of resources in documentary research, oral history collection, geophysical survey, and excavation, archaeology has revealed few material traces of the shelters, stranding them largely in people’s imagination.

This is not necessarily a negative. In the uncertainty of their absence, the shelters generate stories with egalitarian ease: everyone’s recollection has an equal chance of being right. Every person who approaches the archaeologists or hospital administrators has something to offer the project. The process of inviting stories has deepened the connections people feel to the hospital and the place, reconnecting them to the wartime landscape of Adelaide. Unlike northern Australia, Adelaide never suffered air raids, and so the shelters remained unused for their original purpose. Enhanced by a public archaeology program in the present, the archaeological quest for the missing air raid shelters has created a “heritage community” of disparate people, united by their common desire to see these much-remembered and imagined structures.

In Australia, the advent of legislation protecting Aboriginal material culture and requiring consultation with Aboriginal communities, coupled with Aboriginal critiques of colonial research practice, has created a robust and genuinely community-oriented archaeology, still contentious and under constant renegotiation, but indicative of far-reaching changes to the traditional power structures of archaeology. Historical archaeologists, dealing primarily with the material culture of European and other settlers in Australia, have had less incentive to engage with the community as equal partners. In part, this is because the ethical dilemmas of investigating other cultures in colonial societies are largely absent (Greer et al. 2002: 267). In part, it is also due to a general acceptance by the Euro–Australian community of the expertise of archaeologists supported by their institutional authority. But, as Greer et al. (2002) point out, communities are increasingly challenging this authority in the process of assigning significance to places in heritage management. In this paper, we explore some of the facets of doing archaeology in the community through the air raid shelter project at the Repatriation General Hospital (“the Repat”) in Adelaide (Fig. 11.1). Wartime experiences created potent memories and new communities from fractures in the social order; our project can be seen as linking communities of the past and present through individual and social memories.

Fig.  11.1
figure 1_11

The location of the Repatriation General Hospital and the Field School site

Full size image

Theorizing Community Archaeology

In the archaeological literature, community archaeology and public archaeology tend to be used interchangeably; however, they are not necessarily the same thing. The term “public” is often used to describe an extension to an otherwise “pure” academic endeavor. In keeping with the notions of pollution that “pure” implies, it also carries connotations of the vernacular. “Public,” in this sense, sets up a dichotomy between academic and popular that highlights the standard arc toward professionalization followed by many disciplines throughout the twentieth century. It was not until the advent of systematic heritage legislation in the 1960s and a concomitant increase in the number of academy-trained archaeologists working as public servants in the employ of federal, state, and local governments that the practice of archaeology was forced to shift outside its academic context. As a result, early cultural heritage management practice in Australia was the first activity to be labeled “public archaeology” (e.g., Sullivan 1984: v; Witter 1979), tapping into the necessity for public service archaeologists to interact with a nonspecialist client base. This tag was short-lived, however – perhaps because the tensions between professionally warranted archaeologists as stewards of a government-protected past and the people who variously desired to use, collect, or demolish that past only served to highlight the divide. More than three decades later “public archaeology” still implies a deliberate attempt to bridge a schism and could be interpreted as a “dumbing down” of archaeology to appeal to the masses.

Community archaeology, on the other hand, is slightly different: an archaeology where the research agenda is driven by the community. But what exactly is the community? Early sociological literature equated communities with residence in a geographic location, and compared rural or premodern communities (often theorized as ideal) with urban communities in which kinship was no longer a defining factor (e.g., Wirth 1964 [1938]). In Marxian theory, the capitalist social order alienates people from themselves and from the community (Megill 1970: 390): powerful economic and industrial forces operate against the formation and maintenance of communities at the local scale (Baumann 2001:144; Kaufman 1959: 8). According to Nisbet (1953: 7), this decline “has made ours an age of frustration, anxiety, disintegration, instability, breakdown, and collapse” – in other words, the postmodern condition.

For Appadurai (1996), one of the main symptoms of this decline is a growing sense of rootlessness as modern identities become less territorialized and the uncertainties of globalization break down formerly entrenched relationships with place. Within postmodernity, the development of faster transportation and telecommunications have exponentially expanded the space in which communities can exist (Kaufman 1959: 9), transforming this space into something outside of a literal geo-graphy. The opportunities this creates encourage people to live increasingly imagined lives, tempered by an imaginary sense of locality and distance (Appadurai 1996). There is a strong element of contemporary identity which reaches out to the past for its direction, but within this framework it is just as likely to be an “invented nostalgia,” a connection to a past that never existed. In this context, anything which strengthens communities and their relationship to place is seen as having social benefits in terms of individual well-being.

The importance of common action and collective goals emerges as a central facet in many explorations of community. Where groups form to protect or highlight local heritage, collective action is clearly important. But action is not necessarily a critical factor in the communities which emerge around archaeological activity in a local heritage place. People may have a shared attachment to a place or a memory, they may influence each other’s understandings of the past, and they may value the support provided by the shared experience; but the communities which archaeologists identify as their clients do not fit easily into the classical sociological definition. In undertaking the air raid shelter project, we found ourselves examining how the community was constituted by the collective act of remembering.

The History of the Air Raid Shelters at the Repat

Plans for a new Military Hospital in Adelaide, South Australia, began in 1940. Located in a largely domestic and agricultural suburb, the Repat was not in an area considered “vulnerable” – that designation was reserved for the city center and the more heavily industrialized areas to the north. Even so, the Repat was identified as a hospital for air raid casualties, and had a responsibility to protect its staff and patients. Most air raid precautions fell under the heading of passive air defense (PAD) – any means for protection or the prevention of injury that did not involve attack. The Commonwealth committee responsible for all military base hospitals around Australia considered Passive Defence late in 1941, and decided on a standard set of regulations, including, among other things, slit trenches for patients and staff, adequate drainage for trenches and emergency lighting (Hospital Administration Committee 1941–1942).

By January 1942, construction of air raid shelters at the Repat was imminent, although not immediate, since a general shortage of labor complicated matters. The builders and architects had been having a great deal of difficulty finding labor for the work already commissioned: it was wartime, many able-bodied men had enlisted, and only “natural-born British subjects of good character” were permitted to be employed on “Secret Defence Works” (AP567/1 1941/4). By the middle of March, it was decided that the only possible course was to use the military’s own engineering services. The shelters were subsequently built by military labor in late March 1942. The Lieutenant Maintenance Officer further elaborated on their construction: “[c]onditions were that the Architect was to supply Trestles (88 in all), the necessary roof timbers and 300 sheets of Galvanised Corrugated Iron, for riveting. This material has been delivered. The Garrison Engineer to supply all labor, curved iron for the roofs and the necessary sandbags” (105 Adelaide Military Hospital War Diary, App. I, McWaters to Laybourne-Smith, 24/3/42). No other information on the construction of the shelters has been located: both sets of architectural drawings for the trenches are missing from the archived set of hospital plans originally held by Works and Services (which otherwise are complete), and, although the architect refers to a letter describing their construction when enquiring how much they cost (NAA D618/99 M127, Laybourne-Smith to LeMessurier, 14/3/42), this, too, is absent from the file.

Clearly, none of the extant written descriptions refer to anything other than a covered trench: there is mention neither of reinforced concrete nor of internal fittings. The only known photograph (Fig. 11.2a) of the shelters shows one entrance between Wards 1 and 3 to have outwardly flaring concrete walls supported by sandbagged retaining walls, with the whole feature covered by a low earth mound. A set of 1943 surveyor’s sketches show the shelter as a long U-shaped trench accessed via two concrete paved entrances behind the wards, with two smaller trenches providing rear exits (Sketch Book 1180 of Property and Survey Branch of the E&WS, GRG53/166, Unit 48) (Fig. 11.2b). The bulk of the trench system is located east of the wards and behind the then-new cyclone fence cordoning off the open field. No other connecting rooms or trenches are shown.

Fig.  11.2
figure 2_11

(a) The only known photograph of the air raid shelters, c1942, showing the entrance between Wards 1 and 3; (b) The only known plan of the shelters showing the trench layout

Full size image

Little is known of the ultimate fate of the shelters. There is no documentary evidence for their removal or demolition, or the disposal of any contents. Given that the trenches presented constant problems of drainage and undesirable behavior (immediately after their construction the shelters had become the focus for “a congregation of undesirables” engaging in “drinking and unseemly conduct with…women” [105 Adelaide Military Hospital War Diary, AWM 52 11/2/20, 14/4/42]), it is unlikely that they would have been left open beyond their immediate use during wartime. An oral history recorded with a former earth moving contractor noted that in “around 1960” he had been responsible for delivering truckloads of fill to the area of the entrances, and later a D4 dozer and driver. Since his responsibility ended at this point, he was unaware of exactly what the bulldozer driver did, but assumed that he had been employed to push dirt into or over the entrances. A close reading of aerial photographs of the shelter area suggests that they were still visible up until at least 1950, but that after 1958 no obvious surface features remained. It seems likely that the entrance paths to the shelters were bulldozed to the fence line when this area was landscaped sometime between 1949 and 1958. While the entrances were definitely on hospital-owned land, the bulk of the shelter complex lay in the adjoining field. The sale of this land for private housing from 1954 onward caused further, and finally total, obliteration of any surface traces of the air raid shelters.

The Archaeology of the Air Raid Shelters

The archaeological program at the Repat combined a range of techniques in search of the air raid shelters. Following detailed geophysical survey, seven trenches were excavated by hand, six excavated using a backhoe, and a further nine locations probed using a high-pressure water drill. The excavations were incorporated into an 8-day field school that was open to the staff and patients of the Repat, as well as the general public. More than 40 students participated in the field school and between 80 and 100 visitors became involved in the week-long archaeological program.

Despite the scale of excavations, and the depth of both the backhoe excavations and drill holes (over 2 m), only one trench (Trench 7) revealed any traces of the shelters. This was located at the rear of Wards 1 and 3, in a location initially excavated in 2004. This location revealed one of the original shelter entrances, the very same entrance as in Fig. 11.2. Trench 7 revealed a sloping concrete path running west–east, leading to two low, L-shaped retaining walls. Beyond the path was a gravel floor overlying a thin lens of builder’s sand. The walls of the shelter were largely formed from the natural hard clay subsoil into which a timber framework had been placed, although only one section of this remained in situ at the edge of the path (Fig. 11.3). We infer that the interior floor of the shelter tunnel was timber, possibly slatted to enable water to drain through the layer of gravel to the bottom layer of sand. The tunnel walls were approximately 1.8 m (6 ft) high, and clearly continued east under the fence line of the adjoining property.

Fig.  11.3
figure 3_11

The only material traces of the air raid shelters: a concrete path, low retaining walls and the remnant of a single jarrah beam

Full size image

We know from historical documents that the trenches were constructed from a framework of jarrah overlain by corrugated iron, with curved corrugated iron sheets for the roof. The archaeology of Trench 7 clearly showed that any such superstructure had been completely removed prior to the trench being backfilled, at least on the hospital side of the fence. We also know from the historical record that air raid shelters in southern Australia were an extremely short-lived phenomenon. In August 1942, after a flurry of shelter-building across southern Australia that had only begun in late 1941, the Prime Minister ordered no more shelters be built and any projects to construct them stopped so that resources could be focussed on the more vulnerable northern part of the country (Lazzarini 1944).

Oral Histories of the Air Raid Shelters

Since the inception of the Repat project, we have collected more than 25 oral histories relating to the air raid shelters. There are three sets of oral histories that reflect a range of interactions with the site: the first from people who played in the air raid shelters as children in the 1940s and whose memories are filtered through the perceptions of childhood; the second from original staff of the Repat who worked there in its first three decades of operation (i.e., 1941–1943, and the late 1950s–1970); and the third from current or recently retired staff who worked at the Repat after 1970. A fourth set of interviews revealed former staff with no knowledge of the shelters, although some had stories of other supposed underground areas in the hospital. There is not enough space here to detail the components of each oral history, nor the many points at which they united and diverged. Suffice it to say that the earliest accounts of the shelters – remembered by those who had played there as children – described dark tunnels, sometimes with branching corridors and possibly interconnecting, with shelving and/or benches along the walls. The tunnels and entrances were located at the rear of the wards, although some informants also described rear entrances/exits slightly further to the east in what was then an open field.

The account of a former orderly employed at the Repat between 1941 and 1943, who was given the task of searching the air raid shelters for Absent Without Leave (AWL) patients, i.e., patients who had left the wards without permission, is the earliest account of an adult’s experience. He remembered the existence of only one shelter which he described as very dark, about 8-feet wide and resembling a dugout. Other early staff did not encounter the shelters until the 1950s, and recount exploring the complex in more detail. A former gardener and groundskeeper who began work at the Repat in 1958 described two entrances: one between Wards 1/3 and one between Wards 5/7. The 5/7 entrance was sealed by a metal door flanked by brickwork, with a ramp that descended into the tunnels. To him the tunnels seemed to be set deep underground, with kerosene lanterns still hanging from the ceiling. Another former orderly who worked at the Repat after the war and up until the 1970s described four separate shelters – one per ward block and without connecting tunnels in between – although he himself only entered “a couple.” He described each as a long tunnel, again with kerosene lights inside. The most unusual oral history from this period was provided by another former nurse who recounted entering the shelters in the early 1950s. His story described an extensive complex made of reinforced concrete, containing an underground operating complex with at least two fully set up theaters and two wards – one for recovery and one for a general ward, with a few beds in each.

All respondents with direct knowledge of the site described the shelters as consisting of a tunnel, or network of tunnels, with junctions and turns, and at least two entrances marked by low earth mounds. As generalized as this is, this is the most reliable evidence to come from the oral history program, since there is widespread agreement from the earliest accounts through to the latest and irrespective of the age of the informants when they encountered the site. Any attempt to flesh out details beyond this, however, immediately raises inconsistencies.

In terms of construction materials, for example, accounts (in chronological order) ranged from timber, or dirt reinforced with timber, to reinforced concrete, concrete and brick, and curved concrete. While many early accounts referred only to small moveable items or fixtures inside the shelters, two oral histories referred to an underground operating theater, although only one informant described seeing this for himself. The other recounted this as a story that had been told to him, describing an area with two old-fashioned fixed operating tables, and brighter lighting than the rest of the complex. While there is only one claim for an eyewitness to an operating theater, three other interviewees referred personally and separately to seeing beds inside the air raid shelters. One described these as “more like stretchers,” some of which were moveable, others fixed, while another described seeing steel-framed bunk beds, bed pans, bottles, and instruments inside a shelter between Wards 1/3, when routine maintenance work accidentally broke through a buried brick wall. All other references to the contents of the shelters are anecdotal stories circulated among hospital staff rather than direct observations. One former nurse, for example, related stories told to him by older staff:

We were told very clearly by some of the older senior nurses when I first started here that if needs be they could use them [the underground operating theaters] because everything was set up, the instruments, the trolleys, the trays were all there. Apparently when they were sealed all the instruments had been covered in grease, wrapped in oiled cloth and put above whatever height so they could all just be taken off, … there would be no rust, they’d just be cleaned and used (Repat Interview #1, 24/8/05).

By the time current staff became aware of the site, no one had any direct experience of the shelters, although stories about their size, form, and content were still circulating. All of these later oral histories essentially describe the hospital folklore that has been constructed around the site, showing how the details have changed while the core of the story remains. These accounts describe three shelters, with entrances between Wards 1/3, 3/5, and 5/7, a rear tunnel located as far east as Goodwood Road and encountered during construction for the Service (Gas) Station on the corner of Daws and Goodwood Roads, with areas designed to accommodate 300 beds and containing a fully equipped surgical suite. Lined with corrugated iron and with curved ceilings, the shelters were also described in these interviews as being reinforced with concrete. One account described the shelter as having electricity, but with lights that only came on when the door was closed, analogous to the door of a refrigerator. The most extreme account came from a female former nurse who was interviewed during the field school. Her oral history referred to a shelter made from stainless steel, with an 8-in. thick stainless steel door, “compact with food and bedding … So you could always grab a blanket and curl up … it was … just huge inside” (Repat Interview #25, 23/9/07). According to her, the entrance was wide enough to accommodate 5–6 people abreast, with a wooden floor, an 8-in. thick stainless steel door operated by a push button, and air vents to circulate fresh air.

Most of the elements of these later oral histories do not tie into any known information from documentary or archaeological sources. In addition, many are highly unlikely given what we know about the circumstances of the shelters’ construction and use. During the war, for example, stainless steel was an extremely valuable material in great demand for the munitions industry, and it is highly unlikely that it would have been used in any part of an expedient and quickly constructed air raid shelter at the Repat. Furthermore, given that PAD measures were deliberately incorporated into the construction of both the above-ground operating theater and the wards (the theater in particular was designed with a splinter proof wall and ceiling protection), there would have been no need to duplicate such facilities underground. Combined with this, the chronic shortage of labor throughout the Repat’s construction would have made the installation of an underground operating theater impossible, and there is certainly no paper trail for such a feature in any expenditure records.

There are several other interesting patterns to the oral histories. Some aspects can be matched to what we know of the shelters from historical documents, such as the use of timbering, the location of the entrances, the mounding above the entrances, the directions of the tunnels, the junctions and right angles, and the interconnection of the tunnels into a U or H shape. Despite this, the accounts clearly become more elaborate over time. Early accounts from the 1940s describe only tunnels of dirt and timber with limited, if any, contents. The first mention of alternative construction materials arises in the early 1950s, along with mention of the fully functional underground operating theater. The operating theater story that resurfaces again as hearsay in the late 1950s/early 1960s has become a fixture of Repat lore by the 1970s and 1980s, only to be further elaborated in response to the archaeological excavation in 2007. Most interestingly, at least three eyewitnesses claimed to have seen beds or other equipment inside the shelters. Given that these particular stories come from three separate people, they cannot be immediately discounted as hearsay, and it is not possible at this stage to reconcile the content of these stories with either the historical documents or the archaeological evidence.

Does the Truth Matter?

In one sense, it is possible to argue that some facets of the oral histories cannot be true in any objective sense (i.e., the use of stainless steel for walls and doors), while others must be considered highly unlikely (an underground operating room and underground wards for 300 patients). In another, it is also true that local narratives such as these tell us less about objective history and “more about how people construct their sense of place and cultural identity” (Bird 2002: 526). In this sense, people do not necessarily believe the stories (at least not in the literal sense this usually implies, i.e., by having seen or experienced it themselves), but nevertheless use them to constitute a small thread in the complex construction of their cultural identity (Bird 2002: 543). This is particularly important when those identities are tied to places, since one definition of a sense of place is a collective memory and tradition in the locality (Featherstone 1993: 177). These stories bind members of the same community together: once they begin to hear the stories, they are becoming a member of that community; when they begin to pass them on to others, they are consolidating their place within it (Lippard 1997: 50).

A key means by which narratives continue to have effect is through their performance: being told and retold, often with accompanying tests, ordeals, or journeys that transform the generic elements of the narrative into individual personal experiences (Bird 2002) – what Bird refers to as “the legend trip.” The same former nurse who recounted stories told to him by older staff recounts a classic version of the legend trip:

I didn’t go in there because he was telling me terrible stories. … [y]ou wouldn’t believe the sort of stories we used to get told – [that] there are probably some bodies here from the war, [that] this is where they put the people [who] died that shouldn’t in surgery, all these sort of things that they tell young people when they first start in a hospital. And you never really knew what was true and what wasn’t, and you didn’t want to really take the chance, just in case there were some bodies! Stories that we were told when I first started here from the orderlies that had been here included that … there were four theaters down there, [and] the corridors were wide enough to hold a hospital bed and have another hospital bed pass them at a push. That there were ancillary rooms as well, that they were all fully equipped with instruments for surgery, with trolleys, supposedly [they] had kerosene lanterns, I think, rather than any electricity … They could hold 300 and something patients between three units at any one time (Repat Interview #1, 24/8/05).

It is important to note that stories of the shelters do not stand alone: one thing that the oral history program clearly revealed was the volume and extent of stories referring to a range of underground places at the Repat. We collected numerous references to tunnels between buildings, underground storage areas, secret holding cells, underfloor sleeping areas, and hidden cache locations. These are enduring aspects of hospital folklore that create a subterranean “hidden” Repat to mirror the functions of the everyday workplace. Repat staff members Ken Mayes and Darren Renshaw personally followed up each of these leads, as well as investigating every possible hatch or manhole noted across the Repat complex. None proved positive: reputed “tunnels” turned out to be stormwater drains, “cellars” were exposed as crawlspaces for service access, and many areas claimed to have entry hatches showed no signs of any underfloor features. If tunnels had been installed, it is reasonable to assume that they would have been built before or during the main construction phase (1940–1942) and therefore would be indicated on existing plans and specifications. With the exception of Laybourne-Smith’s plans for the shelters, all construction specifications for the Repat have been located and none show any evidence for underground areas. Tunneling would also have involved considerable expense and labor, neither of which is present in any expenditure reports. Despite their proliferation and resilience, none of these stories could be substantiated as true.

Memory

The archaeology of the two World Wars has perhaps engaged with memory more than any other kind of archaeology, although much of this has been concerned with memorials and other commemorative places (e.g., Blades 2003; Saunders 2001, 2003). In a sense, air raid shelters, like all passive defense measures, are a memorial to the civilian survivors so often overlooked in standard treatments of war. Air raid shelters reflect a common home front experience: fear of attack, the inability to fight back, the reliance on reaction, and the drawing of civilians – by definition noncombatants – into the arena of war. As such, they were an acute focus of community interest at the time of their construction and helped link people through shared fear – people interviewed for the Repat project recalled many civilian defensive projects across Adelaide, as well as a general sense of unease about what could happen once Japan entered the war. This original home front community is dwindling, however, as the personal experience of WWII recedes ever farther from the present. Increasingly, the shelters are becoming the focus for a different kind of community, united in the construction of memories about these structures rather than the actual experience of them.

There is a burgeoning interest in the operation of memory at the individual and community level, particularly “social memory” or “collective memory” as first defined by Halbwachs (1992 [1925]). Halbwachs argued that individual memory did not exist outside the social contexts that allowed it to be recalled, contra the Freudian idea of the unconscious as a repository for all past experiences (Olick 1999: 335). Social memory is inevitably bound up with the constitution of communities and their identity. The processes behind the construction of collective memory and its many social manifestations range along a continuum, from authorized public memory (orthodox memories from recognized authorities), to social memory (vernacular memories created spontaneously within a group of people), and social myth (memories of events that never occurred) (Delle 2008). We know from contemporary sociologies of memory (e.g., Degnen 2005), anthropologies of landscape (e.g., Ingold 1993) and archaeologies of identity (e.g., Delle 2008) that absence can be as potent as presence. As Kuchler (2001: 62) argues, “we still customarily conceptualise the memorial’s value as residing in the object or parts of the object, rather than the mental resources created through the object’s disappearance.”

A key factor here is that the veracity of memories is not at stake; there does not need to be an actual experience to realize memory. As Kansteiner points out (2002: 189), millions of North Americans share a limited range of stories and images from the Holocaust despite their lack of direct experience. Kelly (1995) has explored the intersection of memory with place:

Places are held in sites by personal and common values, and by the maintenance of those values over time, as memory. As remembered, places are thus conserved … This conservation is at root psychological and, in a social sense, memorial. But if places are held inside us, they are not solipsistic, since they can be held in common. At a given threshold, our commonly-held places become communities … (Kelly 1995: 142).

Degnen’s (2005) work in the English village of Dodworth is particularly relevant to our experience with the air raid shelter project. In talking to the “Doduthers,” she became aware of the importance of places that no longer existed as effective landmarks for remembering people and events in the “memory talk” of the village. She found herself able, even as the villagers, to move through space with an awareness of an alternate yet present landscape of people and places which were invisible until invoked by the memory talk. Degnen’s wide-ranging exploration of memory, place, and identity highlights the value that Western culture typically places on presence rather than absence: “The irony here is how absences and erasures, which are inversions of physical and present, become loci of memory, or, rather, remain as loci of memory, despite their connection to the physical reality being interrupted” (Degnen 2005: 740–741, emphasis added). What Degnen did not consider were the memories that were also social myths: the imaginative elaborations continually woven around absent places and invisible things. Social myth is the vehicle through which memories do, indeed, “become” loci, while also connecting to the loci that “remain” through social and public memory.

Delle’s (2008) experience with the Underground Railroad is even more pertinent. Like the tunnels that are supposed to exist beneath the Repatriation Hospital, the Underground Railroad has generated countless stories of secret rooms and subterranean access ways. Delle’s excavations at the Parvin homestead in Pennsylvania proved that no such tunnels existed; nonetheless, the story had such a hold on the local community’s imagination that unrelated archaeological features uncovered during the excavations were co-opted to support the social myth. One of the main strengths of social myth is that it has little problem surviving in the face of contradictory evidence (Delle 2008), causing the Underground Railroad to exist simultaneously in all three forms of collective memory: through the authorized histories that have become the accepted narrative (public memory); through the spontaneous social memories preserved by descendant communities (social memory); and through the folklore that persists despite all evidence to the contrary (social myth). In a similar vein, the Repat’s air raid shelters occupy more than one niche. The orthodox memory of the shelters is their raw and streamlined trench shelter form: the simple galvanized iron and timber trenches that can be reconstructed from the documents and archaeological evidence. As social memory, the trench shelters embody the collective experience of those who lived, worked, and played in the Repat landscape during WWII, although in this realm the form and content becomes less fixed and immutable. At the same time, certain elements of the shelters have become pure social myth: their size, stainless steel or reinforced concrete construction materials, operating theater(s), and recovery wards all appear to be wholly imaginary. Paradoxically, it is these elements of social myth that have the most impressive durability, one that field work, compounded by the inability of archaeology directly to contradict them, and the many documented oral histories that elaborate them, have probably only served to reinforce, rather than diminish.

Conclusions

Why do the air raid shelters seem to invite the telling of stories? At a local level, it may be connected to the particular type of community epitomized by the Repat. Hospitals, as Blankenship and Elling observe (1962), support separate, semi-isolated communities:

Each is the hub of a network of loyalties, commitments, and values which are specific to it … [with] volunteers, staunch adherents in the community, loyal workers, and devoted doctors, nurses and board members. This type of commitment … is generated by the organization itself as a cooperative, purposeful system with minimum reference to the larger community and power structure (Blankenship and Elling 1962: 267).

Analysis of oral histories shows that the shelters were far from forgotten in the aftermath of the war. Stories about them circulated among staff who joined the Repat much later, contributing to a sense of community that extended beyond the initial military constitution of the hospital into the present day civilian community. After the war, as the hospital became centered on veterans and the process of repatriation, connections to the past became more and more central to the process of defining its identity. The invented nostalgia generated across 60 years of Repat culture became a vehicle for constructing a popular cultural sense of place that bonded people to each other and to their workplace locality.

At a more general level, memory is vital to all forms of conceptualizing heritage, and the values given to the past in the present. If Appadurai’s assessment of the fragmentation inherent in modern communities is accurate, then one of the key factors in strengthening a sense of community is fortifying their relationship to place. Irrespective of whether this involves “real” social memory or the invented nostalgia of social myth, community archaeology has the potential to forge such ties in a way that few other practices can. In this sense, then, social myth is as vital a force for community strength as any other kind of commemorative practice, although its distance from the orthodoxy of the archaeological past brings back echoes of the academic/public divide. Darvill refers to this very personal structure of knowledge as contemplative knowledge: “beliefs and understandings that provide a basis for attachment to a place, time or event and … which contribute to a sense of identity and a place in the world” (Darvill 2007: 449, 451). He notes that people often are completely uninterested in other categories of knowledge, such as narrative knowledge, typified by the big picture truths of archaeology. Instead, people seek a different experience for their own personal reasons, however much we might wish to tell them otherwise or have them believe a different story. We echo Delle (2008) when asking, should archaeologists debunk such social myths when they function to unite communities?

Invisible, undiscovered, the Repat’s air raid shelters have become a symbol that produces multiple meanings for many different people. Like the Loch Ness monster, the myth of the air raid shelters has been sustained by rare and contradictory sightings: we are left with the task of defining their character from half-remembered glimpses and parts that must then stand in for the whole. When certainty is removed, places can shift freely into social myth, as they become tied to the creative process of “dwelling in the landscape” (Ingold 1993) and the ways in which stories “allow listeners to place themselves in relation to specific features of the landscape, in such a way that the meanings may be revealed or disclosed. Stories help to open up the world, not to cloak it” (Ingold 1993: 171, emphasis in original). From this perspective, the presence of actual air raid shelters is not vital for demonstrating the success of the project. The void left by the shelters is the space where we can value the contributions of all equally. The shelters conceal Schrödinger’s cat, at this moment neither alive nor dead. To excavate the shelter categorically would be to collapse the wave function of memories, stories, images, into a single particle: measurable certainly, but ultimately less rich, and perhaps less meaningful, than the state of uncertainty.