Both folders inventoried completely; 13 works read in targeted passages; neither a full-corpus reading nor a complete North African archaeological synthesis.
Selected Maghreb and Libyan comparisons; Egypt, Nubia, prehistory and medieval archaeology are not systematically covered.
Feasible as an evidence-linked learning and research reader within the existing GIS. The original prompt correctly joins theory and empirical practice, but needs a bounded corpus, claim-level citations, specific image questions and an explicit distinction between regional analogy and Carthage evidence. It cannot recover lost stratigraphy or certify survey accuracy from visitor photographs.
What would make an archaeological interpretation testable?
A useful archaeological discussion begins by making its inferential steps visible. A photograph records the present appearance of exposed surfaces from one viewpoint. Calling those surfaces a house, workshop, sanctuary or repaired theatre introduces an argument about activities and history. Theoretical perspectives help formulate that argument: a processual question might concern resource use, a contextual question social meaning, and a material-practice question the organization of work. None licenses an answer without relevant controls. Hodder and Hutson explicitly connect meaning, agency and history; their position should be read as an argued approach, not a universal definition of archaeology. At Carthage the same wall can support several questions, but a wall’s appearance does not independently answer them all.
A101 · Theoretical interpretation and applicationBinford and Schiffer are often reduced to a single warning against treating ruins as a frozen moment. Reading their exchange shows a more substantive disagreement about inference. Binford insists that apparently disturbed records remain objects of explanation, rather than failed versions of exceptionally preserved sites. Schiffer argues that analysts must establish how particular floor assemblages were formed before equating their patterns with activities or household inventories. These positions share a concern with inference but are not interchangeable. The Carthage reader uses both: displaced material and later occupation are historically meaningful, while their formation must be investigated before they become evidence of a specific use. Their American case studies supply methodological questions, not an empirical history of Carthage.
A102 · Attributed theoretical debateFor each image region, write a description that another observer can check, then propose alternatives and specify their predicted evidence. “Large upright blocks beside smaller masonry” is more reproducible than “a Punic wall.” A construction hypothesis might predict continuous bonding; a repair hypothesis an inserted joint or altered facing; a modern conservation hypothesis documentary evidence of rebuilding. Failure to see a junction in an oblique image is not proof that the junction is absent. The image exercise therefore records hypotheses as unresolved and keeps the next observations explicit. Agreement between annotators can improve description, but it does not independently establish a date. A source must explain which part of the argument it supports, and its geographic and evidential scope must remain attached.
A103 · Implemented research protocolLimit: These proposed checks are not fieldwork already performed.
What happened between ancient use and today’s photograph?
A wall’s survival can involve construction, maintenance, demolition, burial, robbing, excavation and consolidation. These processes do not affect every location equally. The Byrsa mission identifies early graves, workshops, Punic housing and a later Roman monumental platform, giving a site-level historical framework for differential survival. That account cannot identify the exact wall in IMG_4029 without a plan match. A preserved slope room and a removed summit building have different archaeological biographies even when both are linked to redevelopment. The photograph records the latest visible condition of the exposure; it contains neither the removed deposits nor the complete pre-excavation sequence. Later rebuilding is consequently part of the evidence to study, rather than merely an obstruction to an imagined pristine Punic view.
A201 · Site synthesis and qualified image applicationMills’s Carthage discussion gives an empirical counterpart to formation theory. At Bir Massouda, ceramic building material occurs in roof collapse, robbing and makeup; older material can be redeposited beside later structures. Manufacture, first use, reuse and final deposition are distinct events. A fragment’s typological date may constrain its production but cannot alone date the formation of the fill that contains it. Nor can a heterogeneous assemblage be called a household inventory simply because it lies within walls. Record a find’s context, the context’s relationship to a surface or cut, and its proposed formation history separately. The resulting disagreement is productive: an early-looking tile in later fill can test reuse and disposal practices without forcing the whole context into an early phase.
A202 · Published Carthage example and methodological synthesisIMG_3999 shows pale stone steps with cracks, dark joint material and a modern railing. Those are observations of the current surface. They do not separate ancient stone from resetting or establish the age of the dark material. A published theatre account reports extensive modern adaptation, but it does not date every photographed block. Conservation records, before-and-after images and a fabric survey would be needed to distinguish those histories. The same care applies to exposed wall tops: a neat cap can affect the visual impression of courses without revealing the original height. In the model workbench, surface capture, reconstructed completion and stratigraphic interpretation should have different source records. A scan of a conserved ruin documents a modern condition as well as ancient fabric.
A203 · Direct image observation and research questionLimit: No photographed repair is independently dated.
What can masonry show about building, repair and labour?
In IMG_4029, large dressed uprights and smaller irregular stones are visible, with different textures and joints. IMG_3964 provides another view of large vertical elements and smaller coursed masonry. These descriptions invite comparison with stone-chain-and-panel construction, conventionally called opus africanum. Hanoune’s opening discussion emphasizes its heterogeneous forms, a modern term and examples in different places and periods; he also notes reconstruction at Dougga. It is therefore a comparative vocabulary, not a dating instrument or proof of an ethnic identity. Neither photograph has been matched here to a numbered building plan. The case exercise asks which faces meet, where a contact is obscured, and whether the observed fabric belongs to one operation, an alteration or a presented reconstruction.
A301 · Image observation and cautious typological comparisonLimit: Opus africanum remains a comparison; no feature date or ethnic attribution is asserted.
Bente and Berthold use chaîne opératoire to connect intentions, craft skills and technological processes with material analysis. Applied to architecture, this suggests questions about quarrying, selecting blocks, dressing faces, mixing binders, erecting supports and maintaining a building. A photograph may reveal face treatment, arrangement or a candidate joint; it cannot directly recover every stage or the organization of labour. Differences might reflect local materials, structural requirements, repair, availability or deliberate display. To discriminate between these explanations, combine a phased fabric survey with relevant material characterization, quarry comparisons, tools and construction records. Do not make the conceptual chain itself a discovered historical sequence. It is a research framework whose individual links must be supported and whose missing links remain open.
A302 · Theory-to-fieldwork proposalLimit: No petrography, binder chemistry or experimental replication was conducted in this build.
Lancaster offers a particularly instructive Carthage example. She reports that trace-element analysis identified Sardinian provenance for scoria from the Antonine Baths, revising an earlier preference for Sicily. For Leptiminus she discusses pumice linked to Pantelleria through mineralogical evidence. Those material results and the proposed transport histories have different evidential strengths. Her suggestion that building material travelled with established millstone or pottery traffic is an economic interpretation, not a photographed cargo manifest. The reader can connect fabric to regional supply questions while keeping laboratory identification, navigation possibilities and shipment hypotheses separate. Dark stone in the user’s Byrsa photograph is not assigned Sardinian provenance: it is a different, untested object. This distinction makes architecture an entry into empirical economic archaeology without replacing analysis with resemblance.
A303 · Attributed material-analysis result and transport interpretationHow do deposits, buildings and political dates relate?
Lucas distinguishes primary relative chronologies, including stratigraphy and typology, from derived periodizations and calendrical frameworks. For this application, “later than” is a relation; “Roman” is a classification; “after a coin minted in 425” is a conditional chronological constraint. They should not occupy the same database field. A relation may remain valid without any absolute date, and a broad phase may include several depositional episodes. Measurements of height above a room floor describe geometry, not elapsed time. A political event can provide a historical reference while leaving individual structures in use. The timeline therefore keeps ranges and alternatives, while the sequence diagrams retain the basis of each connection. Interpretive clarity depends on preserving these differences even when a uniform coloured period map would look simpler.
A401 · Chronological method and implementation interpretationLézine’s Room 18 record supports heights above the room floor for ash and charcoal, red clay, and earth with charcoal streaks. It reports another charcoal horizon at +1.06 m without a thickness, followed by deposits with an unspecified upper extent. The reader and 3D workbench scale only the reported vertical intervals. Their horizontal display dimensions are arbitrary, and an undocumented interval remains visibly open. This is not a measured reconstruction of the whole room, a sea-level profile or an absolute chronology for its deposits. The example can teach how an empirical record becomes a constrained visual model: each part of the graphic has a source, an extent and an explicit unknown. An uncertainty is not an invitation to finish the shape.
A402 · Published deposit description and bounded visualizationMills’s initial Bir Massouda BM00S2 table supplies a named archaeological grouping, with overlapping Archaic and Late Archaic bands and a Vandal label beginning around 400. These ranges are preserved in the history reader rather than aligned mechanically with the capture of Carthage in 439. The table’s categories organize an investigation; they are not a universal stratigraphy for every Carthaginian area. A crosswalk between excavators would require context concordances and an account of how phases were constructed. Until those exist, the database should retain source-local identifiers, alternative correlations and gaps. This also separates a 146 BCE historical destruction from the formation date of every layer containing Punic material. Residual debris can move, and different rooms can have different afterlives.
A403 · Legacy-source audit and proposed data practiceWhat does spatial analysis need before it can explain behaviour?
IMG_3951 shows a masonry enclosure below the surrounding ground, loose stones, vegetation and water farther behind it. A water installation is a possible research question, not a secure identification. A storage space, altered structure or modern excavation exposure can produce some similar visible features. Koloski-Ostrow’s Italy-based sanitation study combines structures, drainage connections and written evidence to evaluate use; the methodological lesson is to inspect connections rather than infer a latrine or cistern from shape alone. For Carthage, a candidate needs a matched plan, original surface and lining, inlet and outlet relationships, and a phased record. Present water in the surrounding scene cannot establish an ancient water level or ancient drainage performance. Capture conditions require documentation; rainfall is not inferred from this image.
A501 · Image observation and Italy-based methodological comparisonLimit: No sanitation or cistern identification is made.
Kaiser defines access depth through streets and turns, while questioning assumptions in some applications of space syntax. His discussion helps turn vague labels such as “main street” into explicit analytical choices. At Carthage, however, the modern GIS contains park paths, roads and approximate site anchors; these do not constitute the ancient circulation network. A phase-specific analysis would need the excavated street segments, known entrances, possible links, and a sensitivity analysis for missing edges. A closed doorway can change a room graph without changing the room’s outline. Apparent visibility in an oblique photograph is similarly different from a calculated viewshed with measured elevations. Numerical output becomes informative only when the graph’s construction and uncertain connections can be inspected.
A502 · Spatial methodology and feasibility limitThe surviving service structures of the Antonine Baths invite a different spatial question: how architecture organized work, heat, water and access within one building. Yegül and Favro discuss the elevated bath block and its service platform, and label their axonometric as a reconstruction after earlier scholarship. A vertical separation between service level and bathing rooms can be functional within a construction rather than a stratigraphic succession from Punic to Roman occupation. Before modelling circulation, identify which floor levels were contemporary, which doors are inferred, and which paths belong to later reuse. Missing vaults and upper rooms constrain the result. The existing Room 18 deposit model answers a different question and cannot be treated as a complete architectural section of this platform.
A503 · Published architectural interpretation and model distinctionWhose categories shape a “Punic” or “Roman” view?
A building erected under Roman rule can involve local patrons, craftspeople, materials and practices as well as imperial institutions. Rind emphasizes that villas have diverse architectural, economic and ownership arrangements, and that provincial examples do not simply replicate an Italian model. Her definition concerns extraurban complexes, which should not be transferred automatically to Carthage’s conventionally named Roman Villas district. A mosaic, column or construction technique does not identify the language, legal status or ethnicity of everyone who used a place. To investigate identity, connect particular inscriptions, objects, patronage records and spatial practices within their contexts. The question should allow discrepant experiences of empire and changes through time, rather than rank sites by an assumed percentage of “Romanization.”
A601 · Theoretical synthesis and qualified regional comparisonNielsen’s regional chapter is useful for its bath evidence and admission that the source distribution is uneven. Its larger claims about nomadic populations and the success of Romanization are interpretive framing, not observations that a wall can verify. Its own note on private bathrooms at Kerkouane cautions against collapsing public bath institutions into all bathing practices. Yegül and Favro’s reconstructed bath drawing is also useful without requiring adoption of every historical statement in their regional narrative. The reader flags their reference to Carthage being salted and withholds that story as evidence; the Ridley article is listed as a full-text follow-up because only its bibliographic record was accessible here. Source criticism selects and assesses particular arguments rather than declaring a publication entirely reliable or entirely worthless.
A602 · Historiographic auditLimit: The inaccessible Ridley text is not presented as having been read.
Dyson describes how French military and scholarly activity in North Africa made connections between Roman antiquity and modern colonial ambition, and how Delattre’s Carthage investigations intertwined archaeology and ecclesiastical projects. These histories help explain why some monuments became especially visible and documented. They also require careful handling: the motives of an excavator do not automatically falsify each find, while poor recording can limit the relationships that can now be reconstructed. The application should retain excavation campaign, recorder, museum custody, restoration history and changing site terminology. Tunisian partners and institutions belong in source authorship and interpretation, rather than appearing only as locations for foreign work. The relevant task is to reconstruct the production of archaeological knowledge alongside the ancient history it seeks to explain.
A603 · Research historiography and provenance practiceWhich comparisons clarify the site, and which would mislead?
North Africa is an analytical region, not one archaeological sequence. A maritime metropolis, a rural estate, a Cape Bon town and a Saharan oasis may be contemporaneous while differing in preservation, environment and political experience. The comparison cards therefore name the specific question each example can help investigate and state what cannot be transferred. Kerkouane is useful for asking how less-overbuilt Punic architecture survives, Leptiminus for testing building-material supply, and Sabratha for monitoring a coast. None provides the missing plan of a photographed Carthage building. Mattingly’s 2023 publisher outline foregrounds regional and social differences and African agency; it is retained as a reading priority. The book’s description and contents were inspected, not its full argument or complete evidence base.
A701 · Comparative framework and reading priorityThe 2024 Utica chapter summary links Greek pottery, other ceramic traditions, faunal material, radiocarbon evidence and neutron activation analysis to a deposit sealing a named well. It illustrates how several evidence streams bear on chronology, provenance and practices. The feast is the authors’ interpretation of an assemblage, not a direct view of people dining. The published title uses Well 20017 while the summary prints 200017; this mismatch is kept in the source register. Its calibrated intervals are not turned here into one exact date, and a dated deposit is not automatically the first settlement event. Applied at Carthage, this means keeping sample, context, measurement, model and historical inference distinct. A bowl’s stylistic origin is also different from the identity of its user.
A702 · Published summary and methodological comparisonLimit: Full chapter, sample determinations and calibration model not independently audited.
The Libya training study describes automated change detection followed by field campaigns involving the Department of Antiquities. It also discusses Arabic-language teaching and local expertise. The lesson for this Carthage build is practical: a changed pixel or a new annotation should initiate a documented check rather than become an ancient-feature label automatically. MarEA’s Sabratha study offers a coastal comparison, with the observation date and shoreline method retained. Such monitoring addresses present condition; reconstruction of an ancient shore needs additional geomorphological and chronological evidence. The regional interface consequently keeps survey methods, scholarly interpretation and conservation monitoring in separate roles. It links to published records without importing coordinates, feature identities or vulnerability claims that have not been verified for Carthage.
A703 · Published regional methods and proposed applicationHow can a model explain its evidence rather than hide its gaps?
IMG_1987 is a modern MAGON wine panel whose mosaic illustration is credited to El Jem / Musée de Sousse. It is not evidence that the photographed place was Magon’s estate or that the Magon Quarter lies beside the panel. IMG_3955 presents a museum model: the visible roofs and complete arrangement are features of that representation, not directly photographed ancient buildings. Both images are useful for studying how archaeological knowledge is communicated. Record the caption, credited author, date and institution if visible, then distinguish the source for the representation from evidence for the ancient reconstruction. A compelling image can carry a good interpretation, but learners should be able to follow its evidential chain and see which parts remain conjectural.
A801 · Source-critical image readingLimit: No remote publication plate is copied; model scale and reconstruction documentation remain unresolved.
The supplied 264 photographs cover four suggested visit areas, with many close views and modern displays. Screening every image achieves collection coverage, not representativeness of Carthage’s archaeology. Orton’s discussion of purposive and random sampling makes this distinction concrete. A visitor can choose a striking wall, an accessible viewpoint or a sign; the selection probability for every building or deposit is unknown. Twenty similar photographs of one monument are not twenty independent discoveries. The interface separates the bounded photo corpus from the wider site histories, and offers comparative records for areas without securely matched images. A future survey should define its target population, sampling units, visibility conditions and omissions before calculating feature frequencies or using absence in these photos as archaeological absence.
A802 · Sampling critique and collection-scope auditThe image exercise is a place to formulate an argument, not an expert certification mechanism. It saves a local draft with image identity, candidate interpretations, proposed checks and a versioned discussion reference, and exports those fields for review. The existing GIS annotation tool allows a point or rectangle to be linked to visible features. Review should evaluate the cited evidence and reasoning, preserve disagreement and distinguish an accepted description from an accepted date or historical interpretation. A future 3D reconstruction should likewise disclose surveyed surfaces, inferred geometry, missing elements and the reasoning behind completion. That approach allows the public to enrich visibility while giving researchers a recoverable account of how a claim changed. Unknowns remain first-class research results.
A803 · Implemented contribution workflow and development proposalLimit: Worksheet drafts are stored in this browser; they are not reviewed excavation records.
The institutional account contrasts a Punic town without a later Roman city built over it with more transformed sites, and identifies coastal protection measures.
Compare how overbuilding and coastal exposure affect which architecture remains legible.
Limit: A different preservation history; neither an untouched photograph of antiquity nor a template for missing Carthage houses.
A named well-sealing assemblage combines ceramics, fauna, radiocarbon and NAA. The feasting interpretation belongs to the authors.
Link samples and dating arguments to contexts before relating them to wider historical chronologies.
Limit: Summary only; Well 20017 / 200017 mismatch unresolved; deposit dating does not establish a city foundation year.
Hanoune discusses stone-chain-and-panel masonry at Dougga and notes that the Capitol example was substantially reconstructed.
Compare fabric vocabulary while checking conservation history.
Limit: No formal identification or date of the Carthage walls follows from this parallel.
Lancaster discusses pumice in the East Baths and a mineralogical attribution to Pantelleria.
Distinguish empirical material provenance from an interpretation of transport and supply.
Limit: No tested specimen in this photograph collection; the source cannot identify every pale or porous stone.
The publication joins standing buildings, buried sequences, AMS dating and concordances between excavation context systems.
Use its record structure as a model for reconciling earlier and newer Carthage investigations.
Limit: Chapter openings only; no stratigraphic table transcribed and no Garamantian period scheme applied to Carthage.
MarEA uses an image-series approach to investigate shoreline change and coastal condition.
Keep acquisition dates, mapping methods and uncertainty visible in a Carthage coastal study.
Limit: Modern monitoring does not reconstruct an ancient coast or establish a Carthage erosion rate.
Automated change detection was followed by Department of Antiquities field checking; the article discusses Arabic-language training and local expertise.
Use contributed observations to guide checks and support collaboration with local specialists.
Limit: These published campaigns did not validate this Carthage application or its image interpretations.
Open source
Targeted passages read · printed pp. 195–196, 202, 205
Debate about inferential goals and the archaeological record; mixed deposits remain researchable. Comparative theory, not Carthage field evidence.
Open source
Targeted passages read · printed pp. 18–20
Reply to Binford: household floor assemblages require formation-process analysis. Southwestern examples are not a Carthaginian deposition model.
Open source
Targeted passages read · Copyright / ; printed pp. 156–157
Meaning, agency and history; contextual interpretation. Filename says 2004, but the inspected copyright page records first print publication in 2003.
Open source
Targeted passages read · printed pp. 3–5
Relative sequences, periodization and calendrical dating; object manufacture and deposition dates must be distinguished.
Open source
Targeted passages read · printed pp. 14–15
Purposive and random sampling, finite populations and model assumptions. A visitor photograph collection has no demonstrated probability sample design.
Open source
Targeted passages read · printed pp. 30–32; preface pp. xxiii–xxiv
Carthage scoria provenance, Leptiminus pumice and transport hypotheses. Printed p. 30 visually checked. Provenance analysis is stronger than visual resemblance; co-shipping remains a hypothesis.
Open source
Targeted passages read · printed pp. 487, 492, 546, 549
Bath platform and an explicitly reconstructed axonometric. The broad historical framing on p. 492 is critically flagged: its salt story is not adopted as archaeological evidence.
Open source
Targeted passages read · Title / ; Introduction / (unnumbered)
Regional villa definition, economic and architectural diversity, and unequal documentation. Filename 2012 is the series number, not the publication year. visually checked.
Open source
Targeted passages read · printed pp. 84–85
Historical bath synthesis with source-distribution caveats and private bathing at Kerkouane in p. 84 n. 1. Its generalizations about nomadism and Romanization are treated as historiography, not adopted. Printed p. 84 visually checked.
Open source
Targeted passages read · Preface p. xvii / ; printed p. 3
Combines structures, drainage connections, texts and inscriptions. Italy-based methodological comparison; no Carthage installation identification is transferred.
Open source
Targeted passages read · printed pp. 52–53
Defines access depth and questions theoretical assumptions of plan analysis. Useful analytic vocabulary; modern park routes are not ancient street networks.
Open source
Targeted passages read · printed pp. 42–43
Production processes, materials analysis, experimentation and falsifiable hypotheses. OCR is imperfect; terminology is summarized, not silently corrected into a quotation.
Open source
Targeted passages read · printed pp. 60–61, 63
French colonial archaeological institutions, military identifications and Delattre’s Carthage work. A critical history of research, not grounds to discard every older observation.
Opening page read · Full accessible opening page, printed p. 29; article pp. 29–39
Variable stone-chain-and-panel construction, modern terminology and rebuilt Dougga fabric. No date or ethnicity inferred from one photographed wall.
Published summary read; full chapter not accessed · Publisher Summary, chapter pp. 363–397; online 30 May 2024
Assemblage sealing a named well, pottery provenance and NAA. Title uses Well 20017; summary uses 200017. Identifier discrepancy retained; radiocarbon ranges not simplified into a foundation year.
Description and chapter openings read; tables not transcribed · Published book description and accessible chapter openings: ch. 4 pp. 117–134; ch. 21 pp. 615–662
Standing-building and buried sequences, AMS dating and concordance of context systems. Illustrates a publication architecture; no Jarma context is imported into Carthage.
Article sections read · Methods and Case Study 2: Shoreline change assessment, Sabratha, Libya; online 12 July 2023
Coastal condition monitoring and shoreline change. Modern image-series methods are not measurements of Punic coastlines.
Article sections read · Abstract; Training workshops and case studies; Training feedback and outcomes; Libyan Studies 56, pp. 117–127
Change detection tested by Libyan fieldwork; Arabic-language training and in-person expertise. Publisher date: 11 November 2025; field activities: 2023–2024.
Publisher description and contents only · University of Michigan Press, Fall 2023 rights catalogue, : description and contents
Regional and social diversity, discrepant identities and African agency. A current reading priority, not a full-book evidential authority in this extension.
Institutional site account read · Description; Outstanding Universal Value; Integrity; Authenticity
Contrast in preservation and modern coastal protection. Heritage summary is not an excavation-context register; claims of unchanged appearance require qualification.
Bibliographic record only · Publisher bibliographic record; Classical Philology 81.2, pp. 140–146
Research lead for the salt-story audit, not full-text support. The reader withholds the salt claim rather than reconstructing an inaccessible argument.
Indexed excerpts only; full PDF retrieval failed · Search-index excerpts of figs. 5–9 and discussion on printed pp. 44–45
Kerkouane / Byrsa plans and room-access graphs; upper walls and windows unresolved. Author and article title remain unverified; retained as a targeted follow-up, not attributed or used for a detailed claim.
Open source
; printed pp. 3, 5–6, 70–73; Room 18 record p. 70 visually checked · ; printed pp. 3, 5–6, 70–73; Room 18 record p. 70 visually checked
Architectural study of the Antonine Baths, including an explicit warning about incomplete plans, service-floor circulation, Room 18 deposits and competing collapse interpretations. Heights are local historical measurements, not current surveyed GIS elevations.
Open source
; printed pp. 11–12, 16–17; Table 1.3 visually checked · ; printed pp. 11–12, 16–17; Table 1.3 visually checked
Separates BM00S1 basilica from BM00S2 sondages, primary collapse from redeposition and context analysis from phase grouping. Table 1.3 is reproduced as a named legacy scheme, retaining its overlapping bands. No photograph-level fabric provenance is inferred.
Overview section · Overview section
The agency explicitly describes limited ancient survivals and modern adaptation for performances. Present seating cannot be treated wholesale as Roman fabric.
Research account and bibliography; excavation campaign 1974–1991 · Research account and bibliography; excavation campaign 1974–1991
Institutional account of archaic burials, workshops, Punic housing and Roman monumental redevelopment. Its 1846 cathedral date conflicts with Hoyos’s late-nineteenth-century account; that construction date is not adopted.
Project account and composite image caption, especially Trench 2 · Project account and composite image caption, especially Trench 2
Provides a published floor-over-cistern relation and identifies earlier investigations. Its statement of continuous occupation is a project synthesis, not proof that every trench or building was continuously used.
Bounded image audit; screening and individual inspection distinguished · Existing atlas audit: 264 contact-screened images and four full-resolution checks. This archaeology reader: six additional display-derivative inspections, recorded separately.
Screening is not feature identification. Camera-area suggestions, depicted objects and visible materials are separate fields. No phase is assigned from appearance.