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

Regional collection · Malta & Gozo

People, stone
and changing seas.

An island’s history begins beyond its present shoreline. Follow the finds from early sea crossings through farming, monumental buildings and Mediterranean trade.

The earliest secure published presence

Latnija · about 8,500 years ago

Tools, hearths and food remains place hunter-gatherers here before the known farming settlements. Their presence required a sea crossing: both indexed land-bridge chronologies put separation thousands of years earlier.

The 2026 correction concerns the regional comparison model. The authors report that the pre-farming interpretation remains unchanged.

Past sea levels · Evidence and reconstructions

The coastline is part of the question.

The flat atlas and lightweight globe retain generalized modern coastlines. The terrain atlas offers a separately labeled Malta–Sicily shoreline scenario on fixed modern bathymetry. Use local relative sea level, bathymetry and their uncertainties together. Global mean sea level alone cannot resolve local land movement, sediment change or erosion. Reconstructed exposed land does not demonstrate human use or a migration route.

The flat atlas provides period-specific landscape notes. The terrain atlas now offers a regional, time-linked shoreline scenario, with a separate sea-level sensitivity band and a modern-coast reference. It uses fixed modern bathymetry, not a fully reconstructed ancient landscape.

Explore Malta–Sicily’s changing shoreline ↗

Geomorphology and model

Ice Age connectivity

Last Glacial Maximum · c.26,000–19,000 years ago

Published seismic interpretation identifies the western Malta–Sicily land-bridge palaeocoastline at present seabed depths of 119–131 m; its eastern margin is less certain. These are feature depths, not a sea-level confidence interval.

A reconstructed land connection does not demonstrate people crossed it. The window is a broad period index, not a dated lifetime of the bridge.

Bathymetry plus global sea-level model

A modelled disconnection

About 14,200 years ago · 2026 model

Blinkhorn and colleagues combine GEBCO 2019 bathymetry with the Lambeck et al. sea-level reconstruction and place disconnection after about 14.2 ka.

This is model-dependent. It differs from the c.13 ka overview in the Latnija study; the two estimates are not bounds of one confidence interval.

Published palaeogeographic interpretation

An alternative bridge chronology

About 13,000 years ago · 2025 overview

The Latnija study describes submergence of the connection around 13 ka, well before its reported human occupation.

Retain this estimate separately from the 2026 model. Neither estimate directly dates a human journey.

Archaeology interpreted against palaeogeography

Latnija required a sea crossing

About 8,500–7,500 cal BP · occupation overview

Both indexed bridge chronologies place separation before Latnija. The excavation authors infer a substantial open-water crossing, roughly 100 km along their proposed route.

The route, boat and individual voyages were not recovered. Modern straight-line distance is not the reconstructed journey. No exact local sea-level height is assigned to this entire interval here.

Roman-period local relative sea level

Furlani and colleagues infer sea level 1.36 ± 0.10 m below present from archaeological indicators and comparison with modelled curves. The repository abstract does not specify the confidence convention.

A local period estimate, not a global curve or the age of a shipwreck. Exact marker dates, indicative ranges and elevation datum require the full marker audit before time filtering or interpolation.

Read source S107 ↗

The scenario documents its curve, datum assumptions, resolution and sensitivity. Local relative sea level, vertical land movement, erosion and sediment changes still require further work. Shoreline exposure is environmental evidence; human activity still needs its own dated finds.

Sixteen entry points

Read the find alongside its date.

A structured first survey of Malta and Gozo from Pleistocene environmental evidence to later antiquity. Sixteen entry points cover major evidence classes; this is not an exhaustive inventory of every excavation, object or publication.

Showing 16 records

SITE-026 · First people

Latnija

About 8,500–7,500 cal BP · occupation overview

What was found
Limestone flakes, hearths, ash deposits, wild fauna and collected marine shells.
What was dated
Tables report 32 charcoal dates, one successful bone date and 49 shell dates, plus five failed bone attempts. Shell calibration depends on phase-specific charcoal comparisons; two model inputs differ from Table S2.
Reasonable inference
A published pre-farming presence requiring a substantial sea crossing.
Keep this distinction
The approximately 8.5–7.5 ka span is the authors’ occupation overview, not a single confidence interval. Conflicting inputs remain unresolved; the ending is poorly constrained. Older natural-deposit charcoal is not automatically human evidence.

SITE-027 · Before and after arrival

Għar Dalam

Pleistocene fauna; Neolithic cultural layer around 7,400 years ago

What was found
Extinct island fauna in older deposits; pottery and domestic-animal evidence in the upper cultural layer.
What was dated
Heritage Malta gives faunal deposits older than 160,000 years and an approximately 7,400-year-old human settlement assessment.
Reasonable inference
The cave records environmental history and later human use in different contexts.
Keep this distinction
Ancient animal bones do not demonstrate equally ancient people. The older “earliest settlement” description is superseded island-wide by Latnija.

SITE-028 · First farmers

Santa Verna

About 7,400–7,300 cal BP · dated wheat overview

What was found
Early settlement deposits followed by a later temple. A wheat seed is an identifiable early-farming indicator.
What was dated
Wheat grain UBA-31042, Triticum cf. dicoccum, Context (90): 6412 ± 44 conventional radiocarbon BP; published calibration 5474–5320 cal BCE (95% CI), FRAGSUS Table 4.1.
Reasonable inference
Directly dated crop evidence supports farming by this period.
Keep this distinction
Direct crop age does not date its final deposition, pottery facies or temple. Context (90) contains material of different ages; this is not a date for the first farmer’s arrival.

SITE-029 · Settlements and temples

Skorba

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

What was found
Domestic structures, yards, pottery, animal bone and figurine fragments, followed by two megalithic structures.
What was dated
Excavated sequence and ceramic phases; this survey uses a broad 4th–3rd-millennium BCE index for the later monuments only.
Reasonable inference
Domestic life can be studied alongside monumental activity.
Keep this distinction
The monument band is not the age of the earlier huts or proof of continuous occupation across every phase.

SITE-030 · Settlements and temples

Ġgantija

Late Neolithic · institutional overview 3800–2000 BCE

What was found
Two multi-apsed structures within a shared outer wall; differentiated limestone use and painted plaster fragments.
What was dated
Heritage Malta supplies a broad 3800–2000 BCE assessment. Revised regional ceramic phases provide a different level of resolution.
Reasonable inference
Large-scale planning, stoneworking and coordinated labour.
Keep this distinction
This broad institutional span is not a dated construction interval. “Bronze Age” in UNESCO’s short introduction conflicts with its detailed prehistoric synthesis and the site’s Neolithic attribution.

SITE-031 · Settlements and temples

Ħaġar Qim

3600–3200 BCE · older institutional construction overview

What was found
Apsidal buildings, very large limestone blocks, stone and clay figurines and a solstitial illumination feature.
What was dated
Heritage Malta’s historical overview gives 3600–3200 BCE; this does not directly date individual blocks or provide a newly calibrated interval.
Reasonable inference
Substantial architectural skill and symbolic activity.
Keep this distinction
Figurines do not by themselves identify a named deity, social system or universal fertility cult. Building phases need their own chronology.

SITE-032 · Settlements and temples

Mnajdra

3600–2500 BCE · older multi-building overview

What was found
Three principal buildings around a forecourt, with seasonal sunlight relationships in the south structure.
What was dated
The institutional summary places the earliest structure at 3600–3200 and the south structure at 3150–2500 BCE.
Reasonable inference
Architectural choices can be measured against the horizon and the Sun.
Keep this distinction
An observable alignment is not, by itself, proof of intended ritual meaning, a complete calendar or a navigational school.

SITE-033 · Settlements and temples

Ta’ Ħaġrat

4th–3rd millennia BCE · broad monument index

What was found
Two structures made from local Upper Coralline Limestone; the older building was modified for the smaller addition.
What was dated
Architectural sequence and ceramics from before and after construction; no individual determination transcribed here.
Reasonable inference
A changing built place, with use extending beyond one construction event.
Keep this distinction
Earlier pottery beneath or around a monument does not directly date every wall.

SITE-034 · Settlements and temples

Tarxien

Late Neolithic construction; Early Bronze Age cremation reuse

What was found
Four structures, spiral and animal reliefs, a large skirted statue fragment, hearths and a later cremation cemetery.
What was dated
Building sequence and cultural phases; broad 4th–3rd-millennium BCE construction index here. Heritage Malta places cremation reuse after 2000 BCE.
Reasonable inference
Changing access and activities within a major ceremonial complex.
Keep this distinction
The statue’s identity and exact ritual are interpretations. The later cemetery is not the original construction date.

SITE-035 · Burials and people

Ħal Saflieni Hypogeum

c.4000–2500 BCE · UNESCO use chronology

What was found
Three levels of rock-cut chambers, collective burial deposits, red-ochre decoration, ornaments and the Sleeping Lady figurine.
What was dated
UNESCO’s detailed synthesis assigns use to c.4000–2500 BCE. The Heritage Malta page prints 4000–1500 BCE; the disagreement is retained.
Reasonable inference
A complex mortuary setting with architecture echoing above-ground monuments.
Keep this distinction
Resonance is observable; intentional acoustic design and ritual effects need further evidence. Traditional burial totals are estimates, not a modern census.

SITE-036 · Burials and people

Xagħra Circle

Late Neolithic burial sequence; modelled Tarxien end 2375–2255 cal BCE

What was found
Collective burial deposits, skeletal remains and three sufficiently informative genomes in the 2022 study.
What was dated
Bayesian chronology: site’s overall Tarxien presence ends at 2375–2255 cal BCE (95%). DNA specimen age ranges come from chronology models.
Reasonable inference
The sampled people show restricted ancestry and elevated runs of homozygosity.
Keep this distinction
Three genomes do not represent everyone on Malta. Genetic effective population size is not a census; limited gene flow is compatible with some trade.

SITE-037 · Routes and disputed dating

Malta’s cart ruts

No universal numerical date adopted

What was found
Paired grooves in exposed limestone, with variable depth and relationships to other landscape features.
What was dated
Morphology and relative stratigraphic/contextual relationships; not a directly measured age for every track.
Reasonable inference
Transport-related wear or construction can be tested against individual ruts and their settings.
Keep this distinction
Similar shapes need not prove identical age or function. Coastal position alone cannot date a rut without local sea-level and erosion history.

SITE-038 · Later prehistory

Borġ in-Nadur

3000–700 BCE · multi-period overview

What was found
Neolithic remains followed by domestic adaptation, Bronze Age huts, silos and fortification.
What was dated
Institutional overview combining distinct phases; the long band is an index of reported use.
Reasonable inference
Settlement organisation changed after the major temple-building tradition.
Keep this distinction
The band does not demonstrate uninterrupted occupation, or that all structures and silos are contemporary.

SITE-039 · Mediterranean connections

Tas-Silġ

Prehistoric monument; later Punic, Roman and Christian sanctuary

What was found
Megalithic remains, later foundations, pottery, coins, animal bone, votive inscriptions and architectural reuse.
What was dated
Different excavation areas preserve different phases; sealed Tarxien deposits in A differ from residual Bronze Age pottery in A/B.
Reasonable inference
Astarte/Juno identifications in later periods have textual and inscriptional support.
Keep this distinction
Reuse of a place does not establish an unchanged religion since the Neolithic. Absence of a phase in one trench is not absence across the whole site.

SITE-040 · Mediterranean connections

Domvs Romana · Melite

c.50 BCE–50 CE · house’s reported heyday

What was found
Mosaic pavements, domestic architecture and a curatorial record of an elite Roman dwelling.
What was dated
Heritage Malta dates the house’s heyday to the latter half of the first century BCE and earlier half of the first century CE.
Reasonable inference
Evidence of Roman-period domestic life and connections to wider artistic traditions.
Keep this distinction
The date belongs to this house’s assessment, not the beginning or end of Roman Malta.

SITE-041 · Mediterranean connections

Xlendi Phoenician shipwreck

Early 7th century BCE · ceramic typology

What was found
A submerged cargo containing several ceramic-container types and grinding stones, documented by photogrammetry.
What was dated
The preliminary report dates amphora types to the first quarter of the seventh century BCE.
Reasonable inference
Direct material evidence of maritime transport.
Keep this distinction
Cargo typology does not independently identify every crew member, the precise voyage or the hull’s building date. The map marks only a broad offshore area.

Evidence still to connect

Questions that guide the next audit

How early were people present?

Latnija changes the earliest secure published record. Older fauna, charcoal in underlying deposits and disputed remains need separate arguments for human association.

How did people cross and stay?

Reconstruct past coastlines before estimating distance. The 2026 population study favours repeated maritime contact under its ecological assumptions; it does not directly observe the number of voyages.

Where did the early travelers come from?

After the Latnija occupation audit, compare contemporary Sicilian communities and alternative Mediterranean contacts. Keep departure coast, contact network and ancestry separate; possible routes and tool similarities do not by themselves establish origin.

Did settlement, farming, burial and monuments change together?

MT-SEQ-001 finds different schedules with periods of overlap. Compare local occupation breaks, dated crops, construction-related samples and distinct mortuary episodes; none is an island-wide cultural switch.

Did households move before communal activity concentrated?

Taċ-Ċawla shows domestic activity before and during the monumental period. Wider redistribution is plausible, but a household relocation-first sequence remains unproven. Next compare Santa Verna’s final functions with additions at already active Ġgantija and Xagħra.

What did the monumental buildings do?

Ġgantija’s construction comparison now separates walls, finishes, roofing hypotheses and later approach works. Next test how phased entrances and forecourts changed access and gathering, comparing Santa Verna and Tarxien.

Why did the temple tradition change?

Compare climate stress, demography, land use, contact and social reorganisation. Published models differ between abrupt closure and a longer transition; one destruction event has not been demonstrated for the entire archipelago.

Which tracks belong to which period?

Cart ruts require individual formation and dating histories. A rut at today’s shore is not automatically Ice Age evidence.

Coverage still to expand

Kordin III, In-Nuffara and deeper Taċ-Ċawla coverage. Selected Taċ-Ċawla contexts and spatial interpretations are now compared in MT-REORG-001. Its complete assay/context appendices remain unaudited; further Kordin III and In-Nuffara work remains open.

Għar Ħasan and earlier Palaeolithic claims. Investigate original imagery and specimen histories against modern reassessments; do not assign old ages from appearance.

Other tombs, dolmens, catacombs and submerged sites. Extend the inventory with individual excavation reports. Community-authored histories and attributed oral traditions should retain their own provenance, rather than becoming archaeological dates.

The source register distinguishes full articles, supplements, abstracts and metadata-only records. The latest cart-rut journal article still needs a full-text audit; no universal rut age is plotted. The Latnija sample-to-layer ledger now connects all 82 reported dates and five failed attempts to their published contexts, with source discrepancies retained. The field-records follow-up adds published context correlations, hearth analyses and a faunal spatial archive. Original laboratory certificates and the complete context-to-hearth mapping remain unlocated. Related comparisons: the Latnija occupation audit, the origins and motives comparison, the deer comparison with Sicily, the settlement, farming, burial and monument comparison and the settlement reorganization comparison.