PUBLIC RESEARCH PREVIEW Research in progress · uncertainties and source access are recorded throughout.

SITE-029 · 2026-10-29

Skorba

Early Neolithic settlement; later Ġgantija- and Tarxien-phase structures

MaltaGozoSettlements and templesarchaeologysea level
Working research. Assessments and source-access limitations are retained from the research record. This entry has not completed its public-release review.

What has been found

Domestic structures, yards, pottery, animal bone and figurine fragments precede the megalithic buildings. Earlier occupation, first temple construction, rebuilding and later reuse require separate dates. [S84 · Heritage Malta. Skorba Prehistoric Site.] [S83 · Malone et al. (2020)]

Original construction evidence: West before East

MT-SKORBA-001, October 29, 2026. The original Museum Department reports provide a physical basis for the relative order: the East Temple was fitted against an existing West Temple wall, partly removing it and partly incorporating it. This is stronger evidence than ordering the buildings by shape or their oldest pottery alone. It contradicts the East-before-West order in the published Sirius chronology compared here. It does not supply an exact construction year. [S369 · Museum Department, Malta. Report on the Working of the Museum Department for the Year 1…] [S358 · Barratt, R. P. (2022)]

What the excavators actually reported

EvidenceOriginal contextWhat it establishes / limitation
West Temple wall potteryThe 1961 report assigns construction to Ġgantija from pottery within the walls.Construction-related ceramic attribution; no direct assay of initial construction. [S368, PDF p. 3]
West’s early surfacesOnly a few points were examined; clay in the court and west apse, some torba in the east apse.Partial first-floor observations, not complete coverage. [S368, PDF p. 3]
East against WestThe 1962 report describes removal of older wall for the front apse and incorporation of another part into the second apse wall.West wall already existed when this East building was constructed. [S369, PDF p. 4]
East’s attributionThe completed excavation identifies four apses and a niche, with Tarxien pottery from floors and walls.Contextual phase attribution agrees with the relative order; not an exact calendar date. [S369, PDF p. 4]
Earlier occupation beneath EastClay floors attributed to Ġgantija were exposed below its original floor level at the southeast corner.Earlier domestic activity under the temple; complete underlying plans were not recovered. [S369, PDF p. 4]

The decisive report is available directly: Museum Department report for 1962, PDF page 4. Its covering letter is dated May 25, 1963. “1962” identifies the reporting year. The 1961 report, PDF pages 3–4 records the preceding campaign.

These are checked readings of contemporary excavation reports, not a new field inspection. The 1966 monograph, detailed plans, stone identifiers and notebooks remain unaudited. That limits reconstruction of the exact contact; it no longer leaves the reported reason for West-before-East unknown.1

Construction, alteration and abandonment are different events

West was substantially altered in the Tarxien phase: its inner apse was partitioned, with a doorway and additional installations. Later, Tarxien Cemetery-phase occupants subdivided it again. The latter is a separate Bronze Age phase, not another name for the Neolithic Tarxien phase. [S368 · Museum Department, Malta. Report on the Working of the Museum Department for the Year 1…]

Trump later suggested that East fell out of use while West remained occupied. His more explicit account calls this a hint, based on Tarxien rubbish up to 40 cm thick over East’s floor before a roof fire, and cites the 1966 monograph, pp. 9–10. We retain the observation separately from the inference of overlapping occupation. It does not establish the exact interval or continuing ritual use in West. [S371 · Trump, D. H. Xewkija and elsewhere: new thoughts on old sites. Malta Archaeological Review 8 (issue labelled 2006/2007; volume copyright 2011)] [S366 · Trump, D. H. (1981)]

What the radiocarbon and soil dates can tell us

BM-143 is charcoal, not a dated stone or identified initial foundation deposit. The original laboratory report lists Sample 9 at 4380 ± 150 radiocarbon years BP, associated with Tarxien material. FRAGSUS Table 7.1 subsequently labels its context “West temple floor” and gives 3500–2550 cal BCE, 95% CI. The two sources supply different parts of that provenance chain. This is a reported calibration, not one recomputed here. The laboratory’s printed 2430 BC is an uncalibrated expression and must not be used as a calendar construction date. [S370, p. 5] [S83, p. 238]

The renewed excavation also matters because it identifies mixed terrace deposits containing older material. An older seed redeposited during terrace building does not directly date that building event. Its OSL results concern soil formation and a buried land surface, not temple construction. The 2016 context register is now checked, but that trench is not a new excavation of the East–West wall junction. [S83, chapter 7 and Appendix 7]

An additional laboratory-ID discrepancy is preserved in the ledger: the original 1968 report and FRAGSUS Table 7.1 assign the 5000 and 5140 BP results oppositely to BM-145/BM-147. Neither is used to date temple construction here; neither source is silently corrected.2

Consequence for the Sirius proposal

Independent orderPublished Sirius predictionResult
Existing West wall → construction of East against itEast 8750 BCE → West 5300 BCEReversed at the reported physical contact. [S369 · Museum Department, Malta. Report on the Working of the Museum Department for the Year 1…] [S358, Table 1, p. 4]

The predicted dates remain mediated through Barratt’s table attributing them to Reedijk; the complete book’s calculation table has not been checked. A revision proposing an earlier building beneath either temple would need positive evidence identifying its foundations and surviving sightline. Earlier settlement alone does not establish an earlier temple. The reported wall relationship tests these particular buildings, not every possible prehistoric structure on the site.

Honest label: “West-before-East supported by the original reported wall contact; Neolithic phase assignments supported by construction-related pottery. Exact initial construction dates remain unassigned. The compared Sirius chronology reverses this relative order.”

Did adding East and remodeling West reorganize access?

Follow-up MT-SKORBA-002: supported in part. West acquired more strongly marked and potentially closable internal boundaries. The developed complex also has a more elaborate access arrangement in a published analysis. The evidence inspected does not synchronize West’s alterations with East’s construction or identify which people used particular spaces.

The plan and the individual changes

Trump’s excavation plan is accessible in FRAGSUS Figure 7.3, p. 229. Its legend distinguishes Ġgantija, Tarxien and Tarxien Cemetery elements, with separate symbols for features in walls and floors. This is a later reproduction of the original excavation plan, not an independent re-excavation. Crucially, Bronze Age subdivisions cannot be counted as Neolithic access controls. [S83 · Malone et al. (2020)]

FeatureEvidence for changeMinimum inference
West’s terminal apseA later cross-wall leaves a central doorway; the 1961 account records rebates, perforations and bar holes.Access could be narrowed and a barrier secured. Its material and rules of use are unknown. [S368; S366, p. 130]
West’s lateral apsesAdded steps; the eastern apse received a doorway and porthole slab.Movement crossed more explicitly defined thresholds. The east apse within West is not the separate East Temple. [S368 · Museum Department, Malta. Report on the Working of the Museum Department for the Year 1…]
Installations near these boundariesAdded altar-like fittings and reuse of dismantled decorated pieces.Activity settings were rearranged, but neither the performed actions nor their participants are identified. [S368 · Museum Department, Malta. Report on the Working of the Museum Department for the Year 1…]
East’s additionFour-apse building fitted against and into West’s earlier wall.The complex gained another building. Removal of wall for an apse does not itself demonstrate a connecting doorway. [S369 · Museum Department, Malta. Report on the Working of the Museum Department for the Year 1…]
East’s later floor depositRubbish preceded the reported roof fire.Depositional use changed before burning. West’s simultaneous occupation remains Trump’s qualified inference. [S371 · Trump, D. H. Xewkija and elsewhere: new thoughts on old sites. Malta Archaeological Review 8 (issue labelled 2006/2007; volume copyright 2011)]

The physical capacity to close an inner doorway is supported more directly than the claim that an elite permanently excluded most people. A closable boundary could also regulate occasions, sequences of participation, storage or preparation. These are alternatives to investigate, not recovered functions.

An existing access analysis tests the larger pattern

Bonanno, Gouder, Malone and Stoddart (1990) explicitly model Skorba in their Figure 5 and discussion on p. 200. Their published comparison changes from two access levels and six spatial units in the Ġgantija scheme to three levels and sixteen units in the Tarxien scheme. This supports increasing spatial elaboration in their representation of the complex. [S373 · Bonanno, A., Gouder, T., Malone, C. and Stoddart, S. (1990)]

Those figures are source-reported model results, not new measurements. The diagram distinguishes enclosed spaces, niches and an outside reference point. It does not mean sixteen rooms, three storeys, increased population or a measured volume of traffic. The aggregate change cannot be allocated entirely to West’s partition: additions, niches and other represented spaces also contribute. The figures remain attributed because a complete node-to-excavation-feature crosswalk has not been recovered.3

There is a second useful insight in Grima’s analysis: counting spaces alone can miss what happens at their boundaries. A broad opening changed into a closable doorway can alter admission and visibility even if it still connects the same two rooms. Grima discusses Skorba’s cross-wall and lateral thresholds, but also presents an alternative to exclusive priestly access: members of the wider community might have participated selectively or on particular occasions. Neither social interpretation follows uniquely from the plan. [S372, pp. 50–54, 63]

What the evidence can and cannot order

RelationshipAssessment
West’s original arrangement → its later screen/doorwaySupported by the reported excavation sequence. [S368 · Museum Department, Malta. Report on the Working of the Museum Department for the Year 1…] [S366 · Trump, D. H. (1981)]
Existing West wall → East built against itSupported by the reported contact. [S369 · Museum Department, Malta. Report on the Working of the Museum Department for the Year 1…]
West’s screen/doorway ↔ East’s constructionUnordered within the evidence inspected. Both are attributed to Tarxien; no shared construction seam, floor tie or narrowly dated joint event was identified.
East floor use → rubbish deposition → roof fireReported deposit sequence; exact duration unknown. [S371 · Trump, D. H. Xewkija and elsewhere: new thoughts on old sites. Malta Archaeological Review 8 (issue labelled 2006/2007; volume copyright 2011)]

Result: a change in the organization of access is supported. A single coordinated renovation, a prescribed procession, permanent elite exclusion and immediate replacement of West by East remain unestablished. Several successive projects within the same ceramic phase could produce the final plan. A phase-composite access diagram is not proof that every represented space operated together at one moment.

What would discriminate between the remaining explanations?

The most useful missing evidence is now a feature-level crosswalk linking the original plans, sections and find records:

  • Does a floor or construction layer physically connect West’s new partition to East’s addition? That could order or synchronize the projects.
  • Do securely located floor deposits differ across the new thresholds—for example, preparation debris on one side and repeated placement of selected objects on the other? That could test different activities without assigning them in advance.
  • Does West have a use deposit demonstrably contemporary with East’s abandonment? That would test continued parallel use more firmly.

Existing records offer a practical route: the Maltese Temples Landscape Project reports laser scanning of the standing remains, re-excavation beside the temple and recovery of material from old spoil. The public project account does not provide the raw scan or a phase-specific interior finds distribution. Spoilheap finds cannot recover their original room positions simply by being newly collected. [S374 · Bournemouth University. Maltese Temples Landscape Project. Institutional project accoun…]

Honest label: “Access within West became more explicitly bounded and potentially controllable; increasing complexity of the combined complex is supported by a published access model. Coordination of the two building projects and the social purpose of the changes remain unresolved.”

For the astronomical comparison, see the Sirius evidence audit. Interior alterations are another reason to distinguish a building’s general orientation from a particular historic viewing position and aperture; no new astronomical fit is claimed here.

Follow-up MT-SKORBA-003: model and catalogue availability

The earlier note that laser scans had not been found needs narrowing. Heritage Malta’s 2007–08 annual report says 3D models of Skorba and Ta’ Ħaġrat were completed in February 2008 and prepared in X3D for eventual publication. A 2013 technical review identifies laser scanning as the method planned for the Skorba model. Heritage Malta later exhibited a virtual model, and its institutional Google Arts & Culture exhibit lists high-resolution texture-mapped models of the Skorba megalith, central temple and hut. These are real digital records of the standing site, but the sources inspected do not expose raw point clouds, excavation-stage surfaces, phase-coded floors or registration against Trump’s context plan. A CARARE proposal described georeferenced point clouds as “in preparation”; that is not evidence they were subsequently published. [S375 · Heritage Malta. Annual Report 2007–2008, Med-Archeo-Sites project, PDF pp. 46–47.] [S376 · Stanco, F. and Tanasi, D. (2013)] [S367 · National Museum of Archaeology, Malta. The Megalithic Temples of Malta. Google Arts & C…] [S377 · Heritage Malta. Annual Report 2024, Ta’ Ħaġrat and Skorba conservation and documentatio…]

The archive lead is more promising for find provenience. Heritage Malta’s 2024 report says the artefact catalogue made for Skorba and Ta’ Ħaġrat in 2008 was retrieved from CDs for entry into its digital system. The report does not describe fields in that catalogue or publish an interior distribution. The publicly documented MTLP spoilheap sample and Trench M excavation also cannot locate finds on either side of the temple thresholds. [S377 · Heritage Malta. Annual Report 2024, Ta’ Ħaġrat and Skorba conservation and documentatio…] [S374 · Bournemouth University. Maltese Temples Landscape Project. Institutional project accoun…]

Result: no floor or construction layer was identified in the material reviewed that links West’s internal partition to East’s addition. No securely located find distribution across the altered West thresholds was recovered. The digitized catalogue is now the best concrete next archival test: determine whether its records preserve original room, floor, context or threshold-side locations and can be cross-walked to the original plans. A modern surface scan alone cannot recover buried stratigraphy or provenience removed in the 1960s.

Revised honest label: “West-before-East is supported by the reported wall contact. West’s subsequent access alterations and East’s addition are both assigned broadly to the Tarxien phase, but no direct stratigraphic link orders or synchronizes them. Public 3D models document standing remains; raw scan access and phase-specific interior find locations remain unverified.”

Evidence files: research/data/malta/skorba-construction-evidence.json, skorba-access-evidence.json and skorba-archive-model-audit.json. The full 1966 monograph and its detailed plans remain unaudited, although its site plan is checked through the FRAGSUS reproduction. No atlas interval or assay record changed.

Chronological comparison: Settlement, farming, burial and monument use separates sample ages, modeled episodes and construction-related deposits.

Return to the Malta research index · Sea level and past coastlines · Locate this record

Footnotes

  1. The 1961 report tentatively suggested a second trefoil building; after clearance, the 1962 report identifies four apses and a niche. This is an explicitly recorded progression from preliminary interpretation to completed excavation. The earlier suggestion is not treated as a second structure. The broader typological three-to-five-apse development discussed by Trump in 1981 is distinct from this reported wall contact. [S368 · Museum Department, Malta. Report on the Working of the Museum Department for the Year 1…] [S369 · Museum Department, Malta. Report on the Working of the Museum Department for the Year 1…] [S366 · Trump, D. H. (1981)] ↩

  2. S370 p. 5: BM-145 = 5140 ± 150 BP; BM-147 = 5000 ± 150 BP. S83 Table 7.1: BM-145 = 5000 ± 150 BP; BM-147 = 5140 ± 150 BP. Assignment remains unresolved here pending laboratory/context records; these are conflicting representations of existing assays, not four independent dates. BM-143 agrees between these two sources. [S370 · Barker, H. and Mackey, J. (1968)] [S83 · Malone et al. (2020)] ↩

  3. S373 Figure 5 (p. 198) and text (p. 200) were visually checked. Its outside point is an analytical reference, not by itself proof of a surviving common courtyard or passage joining East and West. The analysis draws on the earlier excavation record; it is not another independent observation of the same building events. S83 Figure 7.3 is likewise a reproduction of Trump, not a second original survey. These dependencies are retained in the ledger. ↩