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

Latnija · Following the records

From a date
to a place in the ground.

The public record now connects more of the excavation. Original laboratory certificates and the complete dated-specimen-to-hearth register remain unlocated.

24hearth-analysis samples
7published context correlations
10charcoal IDs with wood identifications
What the follow-up establishes

The reporting form and investigator’s account confirm that context sheets, plans, photographs, photogrammetry and sample databases were created. Published figures and a newer faunal archive provide parts of that record. They do not yet identify which dated fragment belongs to each individual hearth position. S112 · S116

01 · Associations we can support

The feature and the dated fragment need separate identities.

Feature-to-analysis association

Basal hearth in square N2

Extended Data Fig.4b locates the 24 phytolith/FTIR samples in a basal Phase V combustion structure. Table S12 groups 61–66 as controls, 67–74 as thermal-impact samples and 75–84 as combustion residue.

Limit: These are analytical sample numbers, not MAMS laboratory numbers. The inspected figure and tables do not identify the dated-charcoal ID belonging to each sampled hearth position.

Feature and formation evidence

Bed 12 and the substrate beneath it

Supplement section 2.2 interprets the loose ash of Bed 12 and reddened sediment beneath it as an in-situ burning horizon. Extended Data Fig.6 places micromorphology sample MM6 beneath Bed 12; Fig.8 describes thermal alteration below the burning deposits.

Limit: MM6 is a sediment block for microscopy, not a radiocarbon determination. Bed numbers, excavation contexts and laboratory IDs need an explicit crosswalk before a numerical age can be attached to this feature.

Formation-history qualification

Bed 10 includes reworked burning material

Micromorphology samples MM3 and MM5, the contact geometry and lens b support reworking and infilling beneath parts of Bed 10. The supplement also allows local in-situ burning at the southern end.

Limit: Charcoal in ash-rich sediment need not date the deposition of an intact hearth. The formation history cannot be assigned to every Phase V sample without its recorded position.

Shared-context association

Contexts 034 and 048 have pollen analyses

Methods and Extended Data Fig.4a identify the pollen sampling contexts as 034 and 048, both ash-rich Phase V deposits. Table S2 places charcoal 54765 in 034 and 54780 in 048.

Limit: This links analyses at the context level. It does not make the pollen samples identical to the dated charcoal fragments or demonstrate one burning event.

02 · Relative position before numerical age

The Harris matrix supplies explicit context relationships.

A Harris matrix describes the stratigraphic order and correlations recognized by the excavators. In the selected N2 branch below, the published matrix places 039 above 040; 043 and 046 occupy separate branches above 047; 051 lies below. Their radiocarbon specimens are linked by Table S2. This is relative stratigraphy, not a timescale or a diagram of a single hearth.

Selected N2 context relationships with charcoal sample IDsContext 039 above 040, branching to 043 and 046, converging on 047, then underlying natural context 051. Laboratory identifiers come from Table S2, not the matrix itself. Positions and gaps are schematic.Context 039Charcoal 62546Context 040Charcoal 62547Context 043Charcoal 54769Context 046Charcoal 62548Context 047Charcoal 62549Context 051Charcoal 62551
Authored selection from Scerri et al., Extended Data Fig.3, with Table S2 laboratory IDs. Context 051 is Phase VI natural sediment; the others are Phase V. Unshown branches remain in the full matrix. View the complete published matrix ↗

The equality signs below are transcribed from that matrix. Matching or correlated contexts do not make their samples the same object, prove contemporaneity, or establish an intact hearth.

For example, bone 71688 and charcoal 65878 are both reported in context 143. The matrix correlates 143 with 038, which contains charcoal 54771. That strengthens the context trail; it does not resolve the bone’s conflicting laboratory/model values or assign all three specimens to one fire. Inspect 71688 →

03 · Analytical samples are not extra dates

What the 24 hearth samples actually show.

Samples 61–84 belong to the phytolith and FTIR investigation of the basal Phase V hearth in square N2. FTIR identifies mineral signatures; it does not provide a radiocarbon age. The labels below describe where analysts sampled within the structure.

Control substrate

61–66

6 analytical samples

All six are classified as non-burned clay with geogenic calcite.

Thermal-impact sampling zone

67–74

8 analytical samples

All eight clay entries are “Non burned”; three have pyrogenic calcite. The discussion considers subtle spectral alteration and contact or postdepositional mixing.

Combustion residue

75–84

10 analytical samples

Four clay entries (81–84) are classified “Burned”; most calcite is pyrogenic. Sample 81 has an unresolved notation difference within its row.

C, TI and CR are sampling-zone labels. Table S12 lists all TI clays as Non burned; the text discusses subtler spectral changes and possible contact or postdepositional explanations for pyrogenic calcite. Four CR rows (81–84) are classified Burned, but 81 also prints Cl* (nb); both fields are preserved. These results are not additional dates.

Inspect all 24 Table S12 rows
SampleZoneMinerals as printedClay category as printedCalcite origin
61CCa, Cl (nb), QzNon burnedGeogenic calcite
62CCa, Cl (nb), QzNon burnedGeogenic calcite
63CCa, Cl (nb), QzNon burnedGeogenic calcite
64CCa, Cl (nb), QzNon burnedGeogenic calcite
65CCa=Cl (nb), QzNon burnedGeogenic calcite
66CCa, Cl (nb), QzNon burnedGeogenic calcite
67TICa, Cl (nb), QzNon burnedPyrogenic calcite
68TICa, Cl (nb), QzNon burnedGeogenic calcite
69TICa, Cl (nb), QzNon burnedGeogenic calcite
70TICa, Cl (nb), QzNon burnedPyrogenic calcite
71TICa, Cl (nb), QzNon burnedGeogenic calcite
72TICa, Cl (nb), QzNon burnedGeogenic calcite
73TICa, Cl (nb), QzNon burnedGeogenic calcite
74TICa, Cl (nb), QzNon burnedPyrogenic calcite
75CRCa, Cl (nb), QzNon burnedPyrogenic calcite
76CRCa=Cl (nb), QzNon burnedGeogenic calcite
77CRCa, Cl (nb), QzNon burnedPyrogenic calcite
78CRCa, Cl (nb), QzNon burnedPyrogenic calcite
79CRCa, Cl (nb), QzNon burnedPyrogenic calcite
80CRCa, Cl (nb), QzNon burnedPyrogenic calcite
81CRCa, Cl* (nb), QzBurnedGeogenic calcite
82CRCa, Cl (b), QzBurnedPyrogenic calcite
83CRCa, Cl (b), QzBurnedPyrogenic calcite
84CRCa, Cl (b), QzBurnedPyrogenic calcite

Ca: calcite; Cl: clay; Qz: quartz; b/nb: thermally altered/not thermally altered. Sample 81 retains the printed asterisk and “nb” alongside the “Burned” category. These values have not been silently reconciled.

This refines the earlier summary: micromorphology supports heated substrate beneath the burning horizon, while the FTIR results vary by sample and analytical indicator. A zone named “thermal impact” is not, by itself, a positive test for burned clay in every sample. S106 · section 3.2 and Table S12

Ten dated charcoal IDs with identified wood taxa

Table S13 supplies individual identifications for these ten specimens. “cf.” and broad categories remain as printed; no taxon is inferred for the remaining charcoal dates.

Laboratory IDSquare / contextIdentification
65878LM3 / 143Pistacia sp.
65862M3 / 167Pistacia sp.
65864L3 / 172Pistacia sp.
65863M3 / 175angiosperm
65865L3 / 177Juniperus/Tetraclinis
65866L3 / 182Juniperus/Tetraclinis
65867L3 / 183angiosperm
65868M3 / 184Juniperus/Tetraclinis
65879N3 / 190cf. Pistacia
65874MN3 / 198angiosperm

04 · A newly located field-derived archive

Faunal positions are available. The laboratory crosswalk is still missing.

The supplementary archive of the 2026 faunal preprint contains spatial CSVs, excavation grids and other shapefiles. Its principal CSV has 2,868 rows and 1,015 unique UID strings. Rows repeat classifications such as mammal, bird, burned and crusted; those counts are not an audited number of individual finds.

Rows repeat specimens across classifications; some UID strings have coordinate or square variants. Neither row count nor unique UID count is presented as a verified number of individual finds. A separate file contains 46 refit rows grouped into 20 sets.

Coordinate and identity checks come first

Columns named Longitude and Latitude contain projected metre coordinates, not longitude/latitude degrees. Their names are retained in the source inventory; they are not fed into the world atlas. The accompanying code and projection file specify EPSG:32633 (WGS 84 / UTM zone 33N), established by the accompanying R code and .prj file.

The three inspected CSVs contain no 71688, 65865, MAMS or CAIS identifiers. Matching a context number to the dating ledger is a contextual association, not a specimen identification.

The preprint also reports a human identification by collagen fingerprinting. That is not an identification of bone 71688: no crosswalk to that dated specimen was located, and no new direct human date is assigned here.

05 · What was found and what was not

The document trail remains auditable.

The original April 2025 supplement and both archived model releases retain the two conflicting inputs. This extends their version history; it does not decide which measurements are correct.

Not located in inspected public sources

Laboratory certificates for 71688 and 65865

S2 provides reported results; model code provides different inputs. Neither is an original certificate. CAIS and CEZA attribution still needs the submission-ID crosswalk.

Retrieved and compared

April 2025 supplementary PDF

The earlier 111-page Cambridge copy repeats 6100 versus 8100 for 71688 and 7367 versus 7376 for 65865. The discrepancy therefore predates the updated supplement.

Retrieved and checked

Original and later model archives

Zenodo v0.2.0 (14192393) and v1.0.0 (16893452) retain the disputed inputs. v1.0.0 site code matches the pinned GitHub revision byte for byte; neither archive supplied original certificates.

Creation documented; complete register not located

Context sheets, plans and field database

The reporting form and investigator account describe these records. Published figures provide context relationships, selected positions and sections, but not a complete laboratory-ID-to-hearth-position register.

Retrieved and indexed

Basal-hearth analytical tables

S12 supplies 24 mineral-analysis rows and Fig.4b locates their analytical samples. S13 identifies wood taxa for ten radiocarbon charcoal IDs.

Retrieved; partial field-derived dataset

2026 faunal spatial archive

The preprint supplementary ZIP contains faunal coordinates, refits and excavation shapefiles. The inspected CSVs lack a crosswalk to the radiocarbon laboratory IDs.

Paper retrieved; interactive model not inspected

Laser-scanned site model

The 2025 conference paper links a public outreach mesh derived from 11 scans. This is a survey of site geometry, not a substitute for sequential excavation records.

Restricted full text

Marine-fauna dissertation

Only institutional metadata and abstract were accessible. No claims about uninspected specimen or context tables are made.

Search coverage and limits · 2026-10-28
  • Nature article, reporting summary and updated supplement
  • Cambridge original supplementary PDF
  • Pinned GitHub complete file inventory and relevant model files
  • Zenodo v0.2.0 and v1.0.0 ZIP inventories and site inputs
  • Published Harris matrix, chronological context figure, hearth photograph and section/micromorphology figures
  • 2025 3D survey paper
  • 2026 faunal preprint and supplementary spatial/code archive
  • University of Malta dissertation metadata
  • Targeted public web searches for Latnija laboratory reports, certificates, 71688/65865 and field/context registers

“Not located” describes this bounded search. It does not mean the records do not exist or cannot be shared by their custodians. Archive inventories and downloaded-file hashes are retained with the research data.

06 · The remaining records request

A specific request is ready for the excavation team.

The paper names Eleanor Scerri, James Blinkhorn and Nicholas Vella as corresponding authors. The reporting form places specimens at the University of Malta at the time of reporting, with eventual transfer to Heritage Malta planned; a current accession or custodian should be confirmed. No request has been sent.

Resolve the two laboratory inputs

  • Original reports for bone 71688 and charcoal 65865, including conventional ages and uncertainties, report dates and any revised certificates.
  • Laboratory accession numbers, submitter IDs and field specimen IDs: especially the bone listed under the S2 MAMS heading but attributed to CAIS.
  • For the bone: taxon/element, pretreatment, collagen yield and quality criteria (including C:N where recorded); for charcoal: pretreatment and identified taxon.

Connect dated specimens to individual features

  • Relevant context sheets, sample-register rows, specimen coordinates and survey datum for all 33 site-model inputs, prioritizing the Phase V N2 sequence and context 143.
  • Mapping among laboratory IDs, field IDs, excavation contexts, bed numbers and hearth/combustion-feature IDs; primary ash versus redeposited ash or interface/burrow status.
  • Annotated plans or photographs showing the dated fragments relative to the basal N2 hearth, FTIR samples 61–84 and micromorphology blocks MM2/MM3/MM5/MM6.

Identify the authoritative archive and selection history

  • Current repository/access route, version and reuse terms for context sheets, sequential photogrammetry and the laboratory submission register.
  • Explain how the sample-selection/exclusion workflow in Methods relates to the reporting form’s “No data was excluded” statement, and clarify the 2 mm versus 20 mm recording thresholds.
  • Any field-UID-to-laboratory-ID mapping for the newer faunal spatial archive, plus clarification of repeated classifications and coordinate/square variants.
Read the prepared request

Subject: Latnija: laboratory documentation and links between dated samples and hearth contexts

Dear Professors Scerri and Vella and Dr Blinkhorn,

I am developing Ancient History, a source-attributed research repository. I am documenting Latnija’s sample-to-context relationships and would appreciate guidance on the authoritative records and any conditions for citing or reusing them.

Table S2 lists bone 71688 as 6100 ± 30 radiocarbon BP, while the supplied OxCal code uses 8100 ± 30; charcoal 65865 is 7367 ± 19 in S2 and 7376 ± 19 in the code. These differences also occur in the earlier Cambridge supplement and archived model. Could you clarify the intended values and share the original laboratory reports or any corrected records? The bone’s laboratory accession and its relationship to the S2 identifier would be especially helpful.

I am also trying to connect the dated specimens to individual hearths. Are the relevant field sample-register rows, context sheets, plans and coordinates available, with a mapping from laboratory and field IDs to contexts, bed numbers and combustion features? Priorities are the Phase V N2 sequence and context 143, and their relationships to the analytical samples 61–84 and micromorphology blocks MM2/MM3/MM5/MM6.

If these records are held elsewhere or require permission, please direct me to the appropriate archive or custodian. I would also welcome any existing crosswalk between the radiocarbon samples and the newer faunal spatial dataset. I will retain the published alternatives and avoid assigning individual hearth ages while these links remain unresolved.

Thank you for your time and for making the published analyses available.

Return to the complete dating ledger →