• 1 La Motta, Schiffer 1999; Schiffer 1987.
• 2 Cameron 1991; Stevanonvic 1997, p. 385; Tringham 2005.
• 3 Stevenson 1982; Souvatzi 2012.

1Abandonment processes constitute important transformative events in the life cycle of buildings and settlements and the life history of individuals and communities associated with them. Abandonment phenomena determine evident changes in the building’s structures through processes of degradation and destruction or practices of maintenance and repair.1 They also define changes in the social and economic settings of those groups who decide to abandon, because, together with the practices of construction and occupation, settlement abandonment constitutes a segment of a process through which social identities, roles and statuses are negotiated, displayed or maintained.2 When approaching the study of archaeological contexts, the examination of abandonment sequences and the analysis of the practices involved in the ‘closure process’ of a single structure or an entire settlement can offer important insights to reconstruct dynamics of natural and anthropic origins which intervene in the transformation of the archaeological record. Acknowledging the significance of abandonment processes in the analysis of archaeological contexts is fundamental to reconstructing the social, cultural and economic trajectories occurring in ancient communities and the dynamic relationships between their social and natural environment.3

• 4 Swiny, Rapp, Herscher 2003; Swiny 2008; Crist 2021, p. 77.
• 5 Webb, Frankel 2013.

2In the studies of abandonment sequences, a frequent misconception is that destruction events are catalysts for abandonment. In this paper, we want to stress that destruction does not necessarily imply abandonment and that the choice to reconstruct or leave is always based on functional and socio-cultural motivations. In the archaeological record of Bronze Age Cyprus, evidence has been detected of rapid destruction events, which have been identified as principal triggering factors of definitive abandonment processes in some of the main settlements investigated on the island, e.g. destruction by seismic events at Early Bronze Age Sotira-Kaminoudhia4 or by an earthquake followed by fire at Ambelikou-Aletri.5 The present paper aims to review the destruction sequences of the investigated Early and Middle Bronze Age contexts on the island to scrutinize the available data to possibly attempt a characterisation of the abandonment dynamics at these settlements and possibly define if these processes were gradual, as part of a long-term project, or rapid and unplanned. To accomplish so, the paper will focus on floors and related occupation residues, including artefacts, and destruction deposits comprising collapsed walls and roofing structures. The question this paper wants to answer is whether floors were buried slowly or quickly and – in the case of rapid burial – if this was due to natural catastrophic events or to intentional collapse caused by humans. If we admit the occurrence of rapid destruction events, like earthquakes and conflagrations, were these the main triggering factors in the abandonment of the involved settlements? The reconstruction of these dynamics provides significant data which can help to disclose processes and possible motivations behind the abandonment of these Bronze Age Cypriot settlements.

• 6 Lazar et al. 2020.
• 7 Jusseret, Langohr, Sintubin 2013.
• 8 Frankel, Webb 2012; Bombardieri, Amadio 2021.
• 9 Amadio, Bombardieri 2019; Bombardieri, Amadio 2021; Villani 2022; Villani, Tripodi forthcoming.

3In approaching the analysis of Bronze Age contexts in Cyprus, each case study will be examined by considering a series of evidence which can inform on dynamics of buildings and settlements destruction. The evidence examined and the indicators taken into account in this review are based on recent studies on destruction sequences at Middle Bronze Age Tel Kabri6 and at Late Minoan IIIB Crete7 and on fresh examinations conducted on Bronze Age contexts in Cyprus.8 Five major Bronze Age settlements of the island are included in this analysis, based on the larger availability of published data: Early Bronze Age Sotira-Kaminoudhia, Early-Middle Bronze Age Marki-Alonia, Middle Bronze Age Alambra, Ambelikou-Aletri and Erimi-Laonin tou Porakou (henceforth Erimi‑LtP) (fig. 1). Data deriving from combined macro and micro analyses conducted at Middle Bronze Age Erimi-Laonin tou Porakou9 are reassessed here based on ongoing examinations of residual artefacts.

• 10 Frankel, Webb 2012, p. 479.
• 11 Frankel, Webb 1996; Frankel, Webb 2006.
• 12 Frankel, Webb 2012, figs. 4, 7.
• 13 Frankel, Webb 2006, p. 43.
• 14 Tomka, Stevenson 1993, p. 192.
• 15 Frankel, Webb 2006, table 4.14 b‑c.
• 16 Frankel, Webb 2006, table 4.14 c.
• 17 Frankel, Webb 2012, pp. 480‑483.

5The analysis of architectural units, features and assemblages at the Early-Middle Bronze Age site of Marki-Alonia indicates that buildings of the settlements were in a continual process of transformation, as individual rooms or areas of the building compounds were renovated, reorganised, rebuilt, demolished or left as standing ruins.10 The examination of stratigraphic information concerning the depositional history of individual units and compounds11 suggests that no evidence of tilted or displaced walls occurs at the site. Instead, a superimposition of structures, with newer walls built on top of the remnants of an older one, indicates continuous modification of rooms, buildings and courtyards within compounds.12 Sequences pertaining to the final abandonment of the structures are characterized by degraded materials, including deposits of stone rubble fallen from decaying walls and mudbricks decomposed by natural weathering and anthropic erosion. Burnt deposits have been evidenced in the depositional history of some of the buildings within the settlement, e.g. Units IX‑11, X‑11, phases D‑E.13 However, they represent only patchy evidence, perhaps related to the preceding abandonment and following clearing of the area before the subsequent reconstruction. Floors do not show marks or breakages left by collapsed constructional material or fallen objects, suggesting a gradual degradation of the wall structures and a possible spoliation of the residual materials within buildings. Micro residues of occupation, e.g. charred materials and ash within fireplaces, rarely occur within settlement compounds, indicating that the majority of the structures were cleaned before the inhabitant left, hence excluding any sudden collapse or destruction. The last phases of the site – G, H and I – show the gradual disuse of the settlement area, with buildings in ruins while others were still in use. Scholars have underlined how scavenging and reuse by people still living in a settlement play a key role in the distribution of artefacts.14 In the case of Marki-Alonia it is interesting to note that the majority of complete and almost complete vessels does not belong to the last phase I but to E‑1, F‑1, G‑1 and H‑1.15 For example, 40,6 % of the diagnostics vessels from phase G‑1 are complete or are composed of more than 2/3 of their shape.16 This points to the fact that they could be still usable when they were left in the rooms. In addition, well-preserved vessels are represented by small, easily transportable bowls. If still usable, these objects would have had the possibility to be recovered by people still living on the site when parts of the settlement were abandoned. These data are significant of the dynamic process of transformation at the settlement and emphasise how functional motivation alone is often not enough to explain choices of reuse or disuse. Data from Marki-Alonia indicates that no evidence of localised catastrophic events occurred at the settlement; instead, the entire site went through multiple phases of renovations, reconstruction, use and abandonment.17

• 18 Swiny, Rapp, Herscher 2003, p. 53.
• 19 Swiny, Rapp, Herscher 2003, p. 54.
• 20 Soren, Lane 1981.
• 21 Swiny, Rapp, Herscher 2003, pp. 53‑54.
• 22 Swiny, Rapp, Herscher 2003, pp. 50‑52. A mni of 8 individuals has been identified in settlement con (...)
• 23 Swiny 2008, p. 45.

6Data retrieved at the Early Bronze Age site of Sotira-Kaminoudhia are completely different from the evidence observed at Marki-Alonia. Analysis conducted on architectural structures in the three plots investigated at the site – Area A, B, C – indicates the occurrence of collapsed walls with well-coursed tumbles, as identified in Units 1 and 6, Area A and in Unit 12, Area B. Contextual analysis indicates that this collapse-type generally pertains to longer, limiting walls; instead, the shorter wall stretches degraded differently without retaining their original shape. According to Swiny, this evidence may be indicative of a strong seismic event, which principally damaged larger and heavier structures, like limiting walls WAE and WC of Unit 6, Area A.18 Evidence of conflagration, identified as thick layers of charred matter mixed with degraded building materials, has been recovered in Units 17, Area C and Unit 6, Area A. In the specific case of Unit 6, these burnt and ash deposits sealed completely the inventory of the building – which includes also objects showing signs of exposure to fire – and led Swiny to hypothesise a possible conflagration as a consequence of the earthquake episode.19 In support of this idea, he also argues that most earthquakes affecting Cyprus have their epicentres south of the island, close to an active fault zone;20 the proximity to the possible epicentre would explain the destruction horizon registered at Sotira.21 Faulted floors or breakages – which generally characterise destruction horizons by earthquakes – do not occur on occupation surfaces or have not been recorded. The excavator suggested that the dynamic of the walls collapse, the occurrence of extended traces of fire and the presence of skeletal remains within five rooms – Units 6, 16, 18, 22 and 44 – are all indicative of a seismic event and that this caused unexpected destruction of parts of some buildings, sealing some objects on the floors.22 This hypothesis appears to be reinforced by the occurrence of a group of almost complete vessels within some of the building units of the settlement, whose location within the rooms suggests that they were in situ and that they were sealed by a sudden collapse of the structure. However, except for a few cases, in situ finds are rare23 suggesting the possibility that people recovered most of the still usable artefacts from the buildings before leaving the settlement. In this regard, it is important to stress that not all the units in Area A, B and C show traces of rapid destruction; instead, the stratigraphic record indicates that the majority of buildings excavated display evidence pertaining to gradual natural collapse. In all these cases the destruction – rapid or gradual – appears to not have been followed by a re‑occupation and reconstruction.

• 24 Coleman et al. 1996.
• 25 Georgiou 2008.
• 26 Sneddon 2015; Sneddon 2019; Sneddon et al. 2022.
• 27 Coleman et al. 1996.
• 28 Sneddon 2019.
• 29 Georgiou 2008.
• 30 Sneddon 2019, p. 10.
• 31 Coleman et al. 1996, pp. 327‑331.

7At the Middle Bronze Age settlement of Alambra-Mouttes, no traces of coursed tumble, as indicators of walls collapsed by a seismic event have been recorded in Area A,24 B25 and C.26 In Area A, deposits pertaining to the collapse of wall structures in all the seven buildings investigated by Coleman,27 comprise blocks and rubbles mixed with lenses of degraded mudbricks, suggesting a slow and gradual deterioration of the structures. A different scenario instead characterises Unit 1 in Area C. Here, extended deposits of burnt materials have been identified by Sneddon,28 who also refers that evidence of ash and charred matter has been also recognised in Area B.29 The occurrence of this rather extended destruction horizon led him to hypothesise destruction by a conflagration, possibly due to a forest fire. The idea of a rapid destructive event is supported by stratigraphic data which indicate that the burnt deposits of Unit 1 in Area C sealed completely the underlying occupation surfaces, including the original inventory of the building. This domestic unit, likewise Sotira-Kaminoudhia, shows a case of human interment associated with peculiar objects, including two ingots, which were placed on the deceased’s chest (fig. 2).30 It is interesting to note, that also buildings of Area A returned various artefacts, which were left on the floor of the buildings. Their occurrence is unusual, considering that this area of the settlement did not suffer catastrophic events – as demonstrated by the complete absence of indicators of rapid destruction – and occupation surfaces and deposits were not sealed by the sudden collapse of the wall structures.31

• 32 Webb, Frankel 2013, figs. 4.8‑4.11.
• 33 Webb, Frankel 2013, p. 62.
• 34 Webb, Frankel 2013, pp. 62‑69.
• 35 Webb, Frankel 2013, p. 62.
• 36 Webb, Frankel 2013, p. 64.
• 37 Webb, Frankel 2013, pp. 170‑171.
• 38 Webb, Frankel 2013, p. 62.
• 39 Webb, Frankel 2013, see the hypothesised reconstruction of Area 2 on the frontispiece and descripti (...)

8The abandonment sequences of two areas (1 and 2) investigated at the Middle Bronze Age site of Ambelikou-Aletri have returned different stratigraphic evidence. In Area 1 data pertaining to the collapse of the wall structures are constituted by deposits of irregular blocks mixed with sediments partially deriving from the degradation of perishable wall and roofing structures and partly pertaining to deposits accumulated after the abandonment of the site.32 Area 2 instead – the so‑identified pottery workshop – shows the occurrence of thick deposits of charred and ashy matter, onto which built‑up materials collapsed, including partially intact mudbricks, which were preserved possibly because baked by the exposure to high temperatures.33 Unit I of Area 2 is particularly interesting for our analysis. The room returned various complete or almost complete artefacts.34 Largest vessels were found in their position of use, while the majority of the objects stored in the room have fallen on the floor during the destruction.35 The presence of objects with traces of use, as well as newly manufactured vessels produced just before the rapid destruction of the unit,36 suggests that part of the area was destroyed by a rapid event which sealed the artefacts on the floor. The presence of small, not utilitarian objects – such as the only complete figurine discovered to date in a prehistoric settlement on the island –,37 and the occurrence of vessels left in their context of use, support the hypothesis of an accidental event. Webb and Frankel sustain the idea of a conflagration as a consequence of a seismic event,38 however, no further evidence of sudden destruction has been observed in Area 2. Given the presumed perishable nature of structures of Area 2,39 the hypothesis of an accidental conflagration originating from an uncontrolled fire could be taken as an additional possibility and could better explain the localised burnt horizon identified in the workshop. In fact, no direct indicators of destruction by fire or by an earthquake have been identified in the neighbouring Area 1.

• 40 Bombardieri 2021, pp. 8‑9.
• 41 Bombardieri 2017.
• 42 Villani, Tripodi forthcoming.
• 43 Christofi, Stefani, Bombardieri 2015; Bombardieri 2017, p. 3.

9Archaeological investigations conducted at the Middle Bronze Age site of Erimi‑LtP have indicated the occurrence of two main phases (B and A) of occupation. Contemporaneously, the excavations in the residential area of the settlement (Area B, B2, T2, T3, T5) and in the workshop complex (Area A), provided information on sub-phases B2, B1, A2 and A1, which have been identified by few structures which were truncated and modified by successive construction works.40 Concentrations of refuse from the ancient phase B have been discovered in pits and basins disused and sealed during subsequent occupation phases,41 suggesting that the majority of waste materials were discarded in designated areas outside the domestic and productive space.42 When the site was abandoned, inhabitants actualised a series of practices, which provide some interesting information to suggest the conditions and the length in which they decided to leave the settlement. The absence of resettlement and reoccupation after the abandonment has been fundamental to partially preserve the structures from modification and destruction in the post-abandonment phases. The area was sporadically occupied during the Hellenistic and Roman periods, as indicated by surface ceramic evidence in the settlement area and the presence of a few tombs in the Vounaros funerary cluster.43

• 44 Amadio 2019.
• 45 Villani 2022; Villani, Tripodi forthcoming.

10The technique adopted in the construction of Middle Bronze Age Erimi, with buildings carved into the bedrock floor for approximately 0.60‑0.70 m, contributed to preserving the majority of deposits and artefacts from erosion and transportation.44 In addition, a fire event, which destroyed a group of units in the workshop complex at the end of their occupation, produced a rapid burial of the structures and relative floors and ensured the preservation of the objects from scavenging or reuse. Ongoing analysis of the artefactual assemblage, integrated with previous examinations and micro-applications on structures and deposits within buildings of the workshop complex, has provided significant data in the definition of the abandonment processes at Erimi.45 In this area of the settlement, 14 rooms have been investigated, of them, four are wholly burnt (SA I, SA III, SA V and SA XII), five are partially burnt (SA IIa, SA IIb, SA IV, SA XIII and SA XV) and other five are not burnt (SA VI, SA IX, SA X, SA XI and SA XIV) (fig. 3).

• 46 Villani, Tripodi forthcoming.
• 47 Villani 2022.
• 48 Villani 2022.
• 49 Villani 2022.
• 50 Amadio, Boaretto, Bombardieri 2021, p. 302.

11Analysis of the artefacts’ assemblages defined how the majority of residual objects left on the floors of burnt and unburnt units are characterised by big dimensions and heavy weight, suggesting functional motivations behind the decision to leave them out and not take them away.46 The majority of the artefacts recovered in the workshop complex are concentrated in partially or wholly burnt rooms; minor is the number of artefacts within unburnt units. It is interesting to note that only a limited number of almost-complete vessels and still usable artefacts occur in unburnt units; they have been predominantly identified within burnt units. The fact that these unburnt units were almost empty – especially if compared to the large assemblages within burnt rooms – and the total absence of charred materials within them, may indicate that these spaces were already abandoned when the conflagration destroyed the other units of the workshop complex.47 Further important in the reconstruction of the dynamic of settlement abandonment are the processes of reduction and de‑functionalisation which have been identified in some of these unburnt spaces; these processes may support the idea that the abandonment of the workshop complex was conducted as a gradual long-term plan. Particularly interesting is the case of unit SA IIa, a small building located in the central area of the workshop complex, in between two burnt units, SA I and SA III. The bad preservation of the almost complete artefacts recovered within SA IIa – characterised by breakages and repairs in antiquity, and macro use wear – and the presence of primary refuses, provided information to indicate that, in its final phase, this unit was used as a provisional discard.48 Analogously, SA X, a small annex of unit SA V, was filled with more than 130 calcareous blocks of various sizes, suggesting its use as a space to dump building materials, which were possibly dismissed during the progressive abandonment of contiguous units.49 As specified before, the majority of the artefacts from the workshop complex are concentrated in partially or wholly burnt rooms; however not all the units destroyed by the fire returned concentrations of objects. SA V, the largest room discovered at the site (ca. 54 m²),50 returned only a few almost complete artefacts, one spindle whorls and some ground stones.

• 51 Amadio, Boaretto, Bombardieri 2021, p 13.
• 52 Amadio, Boaretto, Bombardieri 2021, pp. 315‑317; Villani, Tripodi forthcoming.

12Fundamental in the definitions of the abandonment processes at the settlement is the presence of a group of small, still usable artefacts recovered within some of these burnt units; e.g. unit SA IV, which returned four intact bowls piled in front of the door and sealed with a layer of silty clay.51 The location and preservation of these complete and still usable objects suggested a deliberate deposition – possibly ritual (?) – before the arson of the room.52

• 53 Frankel, Webb 2012.
• 54 Frankel, Webb 2006.

13Data presented in the previous sections indicate that, except for Marki-Alonia, the other examined contexts show evidence of rapid destruction events occurring in some of their structures during the final part of the settlements’ life history. The stratigraphic record from Marki-Alonia, instead, evidences the complete absence of indicators for sudden destruction. Even when relatively intact assemblages are postulated, the material is often worn, fragmentary and incomplete.53 The presence of multiple phases of renovations and abandonment and the gradual reduction in the use of the buildings provide unequivocal data to suggest a planned abandonment process with forms of intra-site abandonment behaviours, involving the gradual disuse of some of the building compounds and settlement spaces.54

• 55 Bombardieri, Amadio 2021.
• 56 Rapp 1986, p. 368.

14More ambiguous are the data extrapolated from the analysis conducted at Sotira-Kaminoudhia. The evidence summarised in Table 2 indicates that in general terms, the settlement of Sotira-Khaminoudhia shows indicators of a sudden collapse of the structures, which has been interpreted as a consequence of a seismic event. However, only 6 units – Units 1, 6, 16 in Area A, Unit 12 in Area B, Units 22 and 17 in Area C – out of 41 returned evidence which may indicate rapid destruction, and among these, only Unit 6 shows all the indicators pertaining to a possible earthquake event: displaced walls, burnt deposits sealing the room inventory, including a human interment, broken in situ objects. It is also interesting to note, that collapsed and burnt structures are not contiguous with each other, as we may expect in the case of a catastrophic event which destroyed a settlement to the point that the site was subsequently abandoned. As it has been already pointed out in a previous paper focused on the analysis of destruction events at these Bronze Age contexts of the island,55 even in the circumstance of intense earthquakes with the consequent collapse of the mudbrick super-structure, only moderate damage to the stone structure is produced,56 thus permitting fast rebuilding of the structures to a usable condition. The complete absence of evidence for reconstruction after the destruction events, which affected only a limited part of the total number of buildings investigated at Sotira-Kaminoudhia, leads us to question whether the supposed earthquake event was the principal triggering factor for the definitive abandonment of the settlement or if other possible causes need to be taken into consideration. The presence of multiple human interments and the occurrence of localised evidence of destruction could also suggest possible conflicts, which may have caused the devastation and subsequent abandonment of a few structures within the settlement, but not of the entire settlement. Reconstructing the causes of these localised destructions at Sotira-Kaminoudhia is not straightforward and would require more supporting evidence to confirm if the demise of some of the units within the settlement was due to natural events or anthropic processes. Nevertheless, stratigraphic data indicate that individual units are characterised by rather dissimilar depositional sequences, suggesting different life histories. This can suggest that while some of the units were destroyed and deserted, others were possibly still occupied and were potentially abandoned at a later moment. We propose here to see evidence of rapid destruction at Sotira-Kaminoudhia not as an indicator of a fast and unplanned abandonment process which involved the whole settlement but as a segment of a longer-term plan, which was presumably organised in advance and gradually accomplished.

• 57 Bombardieri, Amadio 2021.
• 58 Amadio, Boaretto, Bombardieri 2021.
• 59 Peltenburg 2003.
• 60 Peltenburg 2003, pp. 116‑118.

15Middle Bronze Age Alambra shows a similar scenario with buildings of Area A indicating a natural degradation of the structures and Unit 1 of Area C presenting evidence of conflagration, which has been interpreted as sudden destruction by a forest fire. Because stratigraphic information from Unit 1 is still limited, it is not possible to confirm if the sudden collapse of this structure was caused by a natural catastrophic event or not. As it has been noted,57 it is unlikely that the intramural burial recovered within Unit 1 can be the victim of a sudden collapse. The body, in fact, appears to have been buried with some personal objects (the set of cooper ingots) – evidence which is scarcely compatible with an unexpected death. This intramural burial, instead, supports the idea of a planned action, conducted intentionally. Based on this datum, it is possible to speculate that the conflagration can potentially be connected to the burial as part of a closure ritual. While this practice has never been identified in Bronze Age contexts of the island – with the possible exception of the workshop complex at Erimi –58 analysis conducted at the ceramic neolithic site of Agios Epiktitos-Vrysi indicated the possibility of ritual practices connected to the abandonment of the settlement houses involving the deliberate placement of assemblages on the building floor.59 Stratigraphic evidence resulting from the analysis conducted in Area A at Alambra indicates that all the seven buildings investigated returned still usable objects left in situ, whose placement does not necessarily respond to any functionalistic or practical motivation. Alambra cannot equate to Vrysi in the number of objects, and we are not proposing here to interpret the objects retrieved at Alambra as secure indicators of ritual practices associated with the abandonment of the structures. But we want to stress that for more fully comprehending major episodes of house demise it is necessary to consider in the analysis not only functionalistic explanations because abandonment practices are always culturally mediated and not exclusively based on economising notions in which functional or valuable items are curated.60

• 61 Webb, Frankel 2013, p. 225.
• 62 Webb, Frankel 2013.

16The case of Ambelikou-Aletri represents a rather different example because evidence of destruction and abandonment derives from the analysis of a productive, non-domestic context. As mentioned above, the proposed idea of rapid destruction by a seismic event in the pottery workshop (Area 2) should be possibly re‑evaluated considering that no traces of earthquake have been identified in Area 1. The fact that the burnt horizon appears to be extended on more than one unit in Area 2 could support the idea of an accidental and uncontrolled fire affecting contiguous structures. The high number of artefacts recovered, including newly manufactured vessels and small non-utilitarian objects and the absence of evidence of reconstruction after the destruction can suggest a rapid abandonment process as also proposed by Webb and Frankel.61 They indicate that the absence of pottery recognisably later in date than Middle Bronze Age Cyprus, both from the trial trenches excavated by Dikaios and by the tombs at Theotokou suggests that the whole area was abandoned at this time.62 Unfortunately, the lack of data from a possible-related settlement limits the possibility for a broader reconstruction of the abandonment dynamic.

• 63 Bombardieri, Amadio 2021.
• 64 Amadio, Bombardieri 2019, p. 7; Amadio, Boaretto, Bombardieri 2021; Villani, Tripodi forthcoming.
• 65 Amadio 2019.
• 66 Villani 2022.

17More data are available from the coeval workshop complex at Erimi‑LtP. As it has been presented above, Erimi represents a fortunate case study to analyse abandonment practice in prehistoric Cyprus due to the exceptionally good preservation of depositional sequences within buildings. The rapid destruction, which has been documented in 9 out of 14 units in the workshop complex, appears to not have been consequential to natural disasters but it is the result of anthropic practices, as indicated by the localised nature of the fire evidence, by the lack of episode of reconstruction after the destruction, by the unusual assemblages recovered within burnt buildings compared to the unburnt ones.63 The presence of some easily transportable objects on the floors of the burnt structures suggests a deposition not conditioned by functional motivations, but possibly related to the intentional closure made by the inhabitants and possibly associated with a form of rituality connected to the abandonment.64 The possibility that the observed destruction event could have been related to episodes of violence by external attack or internal strike cannot be substantiated by other evidence, e.g. the presence of possible victims in the settlement perimeter; the evidence of sudden destruction also in the buildings of the residential area. We do expect further data from the domestic buildings of the settlement to foster current interpretations and reconstruction. The integration of stratigraphic analysis and microarchaeological applications with artefactual examinations has provided data to indicate that not all the buildings were rapidly buried, but instead, others went through many phases of transformation, rebuilding and occupation, as testified in some cases, by the occurrence of multiple floor sequences and by the unusual placement of artefacts within buildings.65 These data suggested the site of Erimi was abandoned in a planned and gradual process66 and that not only functional motivations played a key role in the decision to leave or take-away still usable artefacts.

• 67 Cameron, Tomka 1993, pp. 3‑7.
• 68 Webb, Knapp 2021.
• 69 Crewe 2017, p. 146.

18In their overview of abandonment studies, Cameron and Tomka conclude that rapid single-event permanent abandonment occurs only under relatively unique conditions67 because, in the majority of the circumstances, the abandonment of buildings and settlements tends to be a pre-organised planned process, which requires time and cooperation to be fully accomplished. At the regional scale, the abandonment of Middle Cypriot settlements rather than being a single event was a long-term process with local specificities. As indicated by Webb and Knapp68 some of these settlements were abandoned during Middle Cypriot II, including Marki, Alambra and Ambelikou. Other sites instead were abandoned at the very outset of Late Bronze Age Cyprus, such as Erimi‑LtP or later on in the Late Cypriot IA, such as Early-Middle Bronze Age Kissonerga-Skalia. Here the final decline of the settlement corresponded to a decrease in activities followed by a gradual abandonment of the inhabited area.69

• 70 Webb, Knapp 2021.
• 71 Webb, Knapp 2021.
• 72 Webb, Knapp 2021.

19At a settlement scale, it is important to underline that in none of the cases studied, has the destruction of buildings and settlement areas been followed by reconstruction and resettlement. While we could not trace precise motivations behind the definitive abandonment of these Bronze Age Cypriot settlements, we tried to identify broader processes behind this decision. Middle Bronze Age Cyprus was not a period of uninterrupted peace and stability;70 substantial transformations in the socio-cultural and economic structure of these prehistoric Cypriot communities were certainly in place, as demonstrated by new data deriving from ongoing excavations and a reassessment of legacy data.71 Communities and social structures of Middle Bronze Age Cyprus were complex, mobile and interconnected.72 In this transformative social environment, increasing social competition among communities and groups within these communities and the possible emergence of tension and conflict might have been responsible for localised destruction events, as documented in some of the contexts analysed, e.g. Alambra and Erimi‑LtP. On the other hand, the choice of a definitive abandonment – as recorded in all the Middle Bronze Age settlements examined – may be seen as a communal decision, which was possibly conducted as a cooperative effort. The variegated sketch deriving from this study is representative of the multiple biographies of settlements and communities living on the island at the end of the Cypriot prehistory. Reconstructing these micro-histories is determinant in analysing the transformations, of social, economic and environmental nature, which were in place during this period and which conditioned the choice to stay and reconstruct or to abandon and relocate.