The Physical Referent of Tevah: A Via Negativa Application in Mythology
Tevah 的物理指称：一次 Via Negativa 在神话学的应用

Introduction

Tevah (תֵּבָה) is a Hebrew word appearing in Genesis 6:15, commonly rendered as "ark." The text specifies the object's attributes: dimensions of 300 × 50 × 30 cubits, construction in gopher wood, pitch coating inside and out, one side door, a window, three internal levels. Together these attributes constitute a description of an object. Whether this object has a physical referent in the world, and if so how one might identify it, is a question that admits methodological treatment.

This paper belongs to Application Series 1 of SAE Mythology. The theoretical foundation of SAE Mythology is laid in Myth as Remainder Intuition (DOI: 10.5281/zenodo.18917796), which addresses the subject-function dimension of myth — how myth responds to the 5DD–6DD remainder intuition of individual-boundary preservation in the face of death. The present paper addresses a different dimension of mythic narrative: the physical-referent dimension. The two dimensions are complementary, not competing. A single mythic object (such as tevah) can simultaneously carry subject-function (as the vessel of "individual boundary cannot close" in a flood-annihilation narrative) and physical reference (as possibly corresponding to some real large-scale construction in history).

Methodologically, this paper uses Via Negativa (Methodology VII, DOI: 10.5281/zenodo.19481304) as its primary filtering tool. Two candidate methodological tools surfaced during the work and are recorded in Appendix E as input to the SAE methodology series.

This paper does not claim to have located the physical referent of tevah. The output of methodological work is a filtering result, not a proof. Whether the filtered candidate (Area D at the Tell Fara site) is truly the physical prototype of tevah exceeds the authority of methodological filtering and requires future archaeological fieldwork to adjudicate. This paper demonstrates the methodology itself and presents the filtering result for future reference.

---

Section 1 · Scope of Commitment

The first discipline of methodological work is to specify what it commits to and what it does not. Before entering any specific analysis, this paper declares its commitments.

The paper commits to: systematically excluding implausible physical-referent candidates using Via Negativa; describing the structure of whatever candidates remain; demonstrating the complete filtering process so that readers may independently verify; honestly annotating all uncertainties and gaps in the material.

The paper does not commit to: proving that any specific candidate is the physical prototype of tevah; claiming that one version of the Genesis flood narrative approaches "historical truth" more closely than another; offering theological or religious adjudication concerning whether tevah existed; overstepping methodological authority to offer conclusions that require archaeological fieldwork support.

This boundary matters. Via Negativa is an exclusionary tool. The output of an exclusionary tool is "the candidate hardest to exclude," not "truth." A lack of clarity about the boundary misleads readers and unfairly burdens the tool.

A further expression of the commitment scope: the entirety of this paper is based on publicly accessible materials. The author has not conducted fieldwork, consulted unpublished archives, or obtained unreleased data from excavation teams. This is not a methodological shortcoming; it is a boundary declaration. One advantage of Via Negativa is its relative tolerance for incomplete information — it produces meaningful output under information constraints. This paper leverages that property while honestly annotating the resulting limitations.

---

Section 2 · The Stratification of Existence

The most common error in discussions of tevah-type objects is confusion of levels. "Did the ark exist?" bears different meanings at different levels of existence. Answering at one level is not the same as answering at another. Failure to stratify leads to endless crossfire between apologetics, creationism, "ark-search" journalism, and sympathetically religious scholarship — each party speaking past the others.

The existence of tevah can be partitioned into four levels:

Level One: Physical existence. Whether a wooden or brick construction matching the description in Genesis 6:15 actually existed somewhere in third-millennium Mesopotamia. This question can only be answered by archaeological fieldwork. Methodological work cannot answer it directly.

Level Two: Archaeological-species existence. Whether the extant archaeological record contains structural remains compatible with a set of attribute constraints (dimensions, period, location, morphology). This question can be partially addressed by methodological work — candidates can be filtered, though the truth of any candidate cannot be adjudicated without fieldwork.

Level Three: Textual existence. Whether tevah as a textual object maintains attribute consistency across Hebrew Masoretic, Greek Septuagint, Syriac Peshitta, Latin Vulgate, and related textual traditions. This is a philological question with publicly verifiable evidence.

Level Four: Mythological existence. The function tevah performs as a mythic narrative object within cultural transmission, and how it responds to deep structural signals. This question is addressed by the foundational paper of SAE Mythology and lies outside the scope of this paper.

The filtering work of this paper is strictly conducted between Levels Two and Three. Given textual constraints from Level Three (the attribute set of tevah), it filters candidates in the archaeological record at Level Two. Level One exceeds methodological authority; Level Four is handled by the foundational paper.

This stratification directly shapes the structure of the paper. Section 3 addresses Level Three (the referent chain). Section 4 addresses a special subproblem of Level Three (numerical attributes). Sections 5 and 6 perform the Level Two filtering. Sections 7 and 8 present convergent candidates. Section 9 honestly addresses possible errors. All work is conducted between Levels Two and Three, never touching Level One.

---

Section 3 · Reverse Tracing of the Referent Chain

Tevah did not appear out of nowhere in Genesis. It has a cross-linguistic and cross-temporal transmission chain. Understanding what tevah refers to requires tracing this chain.

Downstream (latest, around the turn of the Common Era): Vulgate Latin uses arca, Peshitta Syriac uses qebutha, Septuagint Greek uses κιβωτός. All three terms correspond, in their respective languages, to "box, chest, container."

Midstream: Hebrew Masoretic Text uses tevah. The final redaction of Genesis is typically placed between the sixth and second centuries BCE, with the Priestly layer often dated to the sixth or fifth century BCE. Tevah appears only twice in the entire Hebrew Bible: once in the Noah narrative and once for the papyrus basket in which the infant Moses is placed (Exodus 2:3). Both instances denote a pitch-coated container. The word's narrow usage indicates a specialized term, not a common vocable.

Upstream: the Mesopotamian flood narrative family. The Sumerian Flood Story (Ur III period and possibly earlier, c. twenty-first century BCE); the Atrahasis Epic (Old Babylonian, c. seventeenth century BCE); the Ark Tablet described by Finkel (2014), which specifies a circular quffah (c. seventeenth century BCE); and the Gilgamesh Epic Tablet XI in its standard Akkadian version (redacted between the thirteenth and seventh centuries BCE). These narratives share the motif of "world-destroying flood, small number of survivors, a vessel that carries them through."

The Egyptian subbranch: the etymology of tevah. Brown-Driver-Briggs (BDB) traces tevah to Egyptian ṭbt (box, chest, container). This etymological hypothesis was first proposed by Brugsch and Erman in the late nineteenth century. HALOT (The Hebrew and Aramaic Lexicon of the Old Testament) continues to favor an Egyptian loanword origin, distinguishing tevah from the Mesopotamian eleppu (boat) lineage.

Here is a linguistic fact of consequence. The Hebrew text uses an Egyptian loanword (tevah) rather than a Mesopotamian loanword for Noah's vessel. Contemporary Mesopotamian flood narratives use eleppu or cognate terms, all etymologically Mesopotamian. The two word-systems do not share an origin.

This linguistic fact aligns with the geometric attributes of tevah. Genesis 6:15 describes tevah as rectangular (300 × 50 × 30 cubits, ratios 10 : 1.67 : 1). Gilgamesh XI describes eleppu as square (all sides equal, ešer nindanā meaning "ten rods"). The Ark Tablet describes the quffah as circular (diameter approximately 70 meters). Three distinct geometries.

Two independent lines of evidence — etymological (different word-system origins) and geometric (different shapes) — point to the same conclusion: tevah was not inherited directly from Gilgamesh eleppu or Atrahasis quffah. The prototype of tevah (if one exists) must be sought within the Hebrew tradition itself, or at the interface between the Hebrew tradition and the Egyptian tradition, but not within the Mesopotamian boat tradition.

This does not mean the overall framework of the flood narrative came from somewhere other than Mesopotamia. The Mesopotamian flood narrative family and the Hebrew narrative share the structural motif (annihilation, survival, renewal), and this shared motif may well originate in Mesopotamian tradition. But the specific geometry and name of the vessel are independent developments in the Hebrew version, not inheritances. This distinction matters for subsequent filtering.

Ontological candidates for tevah. Given the textual constraints (rectangular, 300 × 50 × 30 cubits, pitch-coated, with door and window, three levels), tevah admits at least three ontological types.

Candidate A: A vessel. The traditional understanding. But modern engineering experiments — including the 2014 attempt by Finkel's team to scale a quffah per the Ark Tablet — demonstrate that a wooden ship of this scale is not physically feasible. The largest documented commercial wooden ship (the schooner Wyoming, 1909, 99 meters) was smaller than tevah would be and nonetheless leaked continuously and eventually sank. A "300-cubit wooden ship that actually floats" is excluded on engineering grounds.

Candidate B: A fixed construction. A large rectangular-plan structure — possibly a temple, monument, granary, or other functional building. This candidate bypasses the engineering problem and preserves the dimensions and form. But Candidate B must answer a physical-mechanism question: how does a fixed building permit survival during a flood? One possible mechanism: a sealed large-scale construction (granary, refuge, or monumental building) built on a sufficiently high earthen platform or early ziggurat-form elevated base. When the flood submerges the surrounding plain (typical Mesopotamian river flooding reaching meters to tens of meters of depth), the building on the platform becomes an island in the waters. In the narrative imagination of distant observers, this island may be remembered as "a rectangular container floating on water." This mechanism allows Candidate B to be logically consistent with the core motif of flood narrative (a vessel that bears survivors through the flood), but it is itself speculative and requires archaeological verification (especially of platform existence and height).

Candidate C: Raft-plus-mudhif composite (Dickin 2018's hypothesis). A wooden raft base supporting reed-framed arched huts (mudhifs, still built today by the Marsh Arabs of southern Iraq). Dickin interprets tevah's "three levels" not as decks but as horizontal reed-bundle layers in the mudhif framework. This candidate lies between the vessel and the building options — it has floating properties and building morphology at once.

All three ontological candidates are geometrically compatible with a rectangular footprint of approximately 137 × 23 meters. Via Negativa filters location, not ontological type. The paper does not force a choice among the three. If the filtering produces a matching physical candidate, that candidate may be compatible with one or more of the three ontologies.

The furthest reliable anchor in reverse referent tracing is the Hebrew Masoretic tevah (the stable textual form from around the sixth century BCE). Further upstream is inference. The paper takes this anchor as its starting point and filters the physical world for possible correspondents.

Section 3 Appendix · The Boundary of Filtering under Referent Drift

The referent-chain analysis of this paper requires an explicit methodological boundary: Via Negativa operates directly only on the endpoint form of a drifting referent. This boundary is important enough that it must appear in the main text, not be buried in an appendix.

Applied to the tevah case: this paper uses the attributes of Genesis 6:15 (300 × 50 × 30 cubits, rectangular, pitch-coated, door, window, three levels) as filtering constraints and applies these constraints to the physical world. But the paper does not project these attributes directly onto the earlier vessel forms of Ziusudra, Atrahasis, Utnapishtim, or the Sumerian Flood Story.

The reason for this boundary: the Hebrew MT tevah is a stable form reached after long referent drift. Earlier forms (from the Ziusudra era, Atrahasis era, Gilgamesh era) may have had entirely different attributes. Projecting later textual attributes back onto earlier forms may incorrectly exclude the true early prototype — the true early prototype may lack attributes that were added later in the textual tradition yet still be the authentic starting point of the same chain.

Consequently, the filtering result of this paper should be understood as: filtering the physical referent of the stable MT-tevah textual form, and not filtering the earliest prototype of the entire flood mythological tradition. If Area D passes the filtering, it becomes a candidate anchor for the MT-tevah referent chain, not a proven position for "the earliest origin of flood mythology."

This boundary corresponds to a methodological tool formalized in Appendix E.1 for cross-object generality. The main-text treatment here ensures readers know, before entering the Section 5 filtering, what is being filtered and what is not.

---

Section 4 · A Criterion for the Reality-Status of Numbers

Not all numbers in mythic narrative belong to one kind. Treating them indiscriminately is a common error. Some numbers are narrative crystallizations ("forty days and forty nights"); others are empirical residues ("60 cm of flood deposit"). Their provenance and handling differ. This section offers a three-part criterion.

Category I: Symbolic fictive numbers. Round numbers that correspond to known cultural symbolic values. In the ancient Near East and Mediterranean, common symbolic numbers include 7, 12, 40, 70, one year, a thousand years, the twelve tribes. These numbers originate in narrative composition or cultural inheritance; they do not reflect physical measurement. Treatment: not to be used as physical evidence.

Examples: Noah living to 950 years; the flood lasting 40 days and 40 nights; the seven days of creation in Genesis 1; Abraham's age of 75 when leaving Haran. All Category I.

Category II: Measurement fossils. Non-round numbers, typically carrying decimals or scattered integers. These originate in direct physical measurement; their precision tracks measurement precision. Treatment: serve as hard evidence of physical events.

Examples: Schmidt's 1931 report of a flood deposit of approximately 60 cm at Shuruppak (with Raikes's 1966 comparison giving 40 cm or 16 inches at Kish). The 110-meter linear structure at Area D from Hahn et al. 2022 magnetometry. The 138 × 24 meter western rectangular platform from Einwag et al. 2025. All Category II.

Category III: Design-fossil candidates. Numbers that are round but non-symbolic, whose combinations exhibit engineering ratios. These numbers may originate in builders' design decisions, recording the design intent of some actual construction. Treatment: serve as strong diagnostic signals for construction events, with the acknowledgement that movement from "candidate" to "confirmed design record" requires additional evidence. Round, non-symbolic numbers in engineering ratios may also arise from later idealization, standardization, ritual formalization, or textual reconstruction — these possibilities must be excluded by other evidence before confirmation.

This third category is the key extension this paper's criterion makes over the usual binary (symbolic vs. measured). Round numbers do not automatically equal symbolic fictive numbers. Roundness can be symbolic (Category I) or functional (Category III). Three tests distinguish them: (one) whether the number corresponds to a known cultural symbolic value; (two) whether the combination exhibits an engineering ratio (as opposed to simple symbolic ratios like 1:1:1 or 2:1); (three) whether the number falls under some known standardized unit of measurement. Satisfying all three yields a Category III candidate designation — not a direct assertion of "this is a design fossil."

Examples: the Giza Great Pyramid's base of approximately 440 Egyptian royal cubits (440 is not a symbolic value; base area reflects geometric stability requirements). The various levels of the Etemenanki ziggurat (round but reflecting engineering stacking requirements). Tevah's 300 × 50 × 30 cubits (300 is not a symbolic value; the 10 : 1.67 : 1 ratio is an engineering proportion). All satisfy the Category III candidate test and can enter analysis as strong diagnostic signals.

Application order: first check roundness. Non-round goes directly to Category II. Round ones are checked for symbolic correspondence. If matched, Category I. If not, check proportional structure. Engineering proportion yields Category III candidate.

Time-floor constraint from units of measurement. This is a hard constraint accompanying the design-fossil criterion. Round numbers can only be produced when standardized units of measurement exist. Before c. 2700 BCE, no standardized cubit unit existed in the ancient Near East. The Sumerian cubit (kùš, approximately 49.6 cm) was standardized from Early Dynastic II onward. The Egyptian royal cubit (approximately 52.5 cm) was standardized from the Third Dynasty. Before c. 2700 BCE, different villages and different craftsmen used cubits of varying length; no cross-generationally stable figure of "300 cubits" could be produced.

This constraint decisively affects the possible provenance of the tevah numbers. The "300 × 50 × 30 cubits" of tevah cannot originate from physical construction memory prior to c. 2700 BCE. Any hypothesis placing the source event in the sixth or fifth millennium BCE (including Dickin 2018) must explain how this specific set of numbers could be generated and preserved across several thousand years in an era without standardized measurement. Such an explanation is difficult to sustain from a narrative-transmission standpoint — oral tradition can reliably preserve structural motifs across millennia (no question), but preserving specific numbers across millennia exceeds typical oral-transmission capacity.

Matching check. Tevah's 300 × 50 × 30 cubits, converted at approximately 45.7 cm per Hebrew cubit: length 137.16 meters, width 22.86 meters, height 13.72 meters. These metric values correspond closely, along two planar dimensions, to the western rectangular platform of Area D discussed in Section 7 (138 × 24 meters). The observation is a possible overlap between Category III (design-fossil candidate at 137 × 23 meters) and Category II (measurement fossil at 138 × 24 meters) on the same object — though whether it is truly the same object remains to be verified.

This matching merits further metrological analysis, for it involves an interesting reversal. If Area D was indeed built in Shuruppak's Early Dynastic period (c. 2700 BCE), the builders would have used the Sumerian cubit (c. 49.6 cm), not the Hebrew cubit. At 49.6 cm per cubit, 300 Sumerian cubits would be 148.8 meters — not 138 meters. This appears to undermine the hypothesis that Area D is a candidate physical referent for tevah.

But consider the reversal. Suppose narrative transmission preserves not the number "300" itself but the absolute physical length of the building (approximately 138 meters). When the Hebrew text authors, in their own era, express this remembered length using their own cubit unit, they compute 138 ÷ 0.457 ≈ 301.9, which naturally rounds to 300 cubits. On this reading, "300 Hebrew cubits" is not the builders' design number; it is the result of the recording tradition re-expressing an inherited absolute length in its own units. The original construction figure (perhaps approximately 280 Sumerian cubits) is lost; what was transmitted is the physical length. Such cross-cultural metrological conversion is not inconceivable in narrative transmission, but it is a speculative explanatory mechanism requiring other evidence.

To be honest about the epistemic status: the "absolute-length preservation plus cross-unit conversion" reading is one possible mechanism, not a demonstrated fact. Readers may adopt it or reject it. If rejected, the "300 Hebrew cubits ≈ 137 meters ≈ Area D's 138 meters" matching still holds but requires another explanation (perhaps coincidence, perhaps the Hebrew tradition used a variant cubit, perhaps Area D's actual construction period is not Early Dynastic). The matching is a fact; the mechanism behind it is open.

For detailed exposition of this criterion and cross-domain application examples, see Appendix E.2.

---

Section 5 · Systematic Derivation of the Fifteen Negations

Via Negativa is not arbitrary exclusion. It is item-by-item examination of each type of candidate. This section offers the systematic derivation. For reasons of length, the main text collapses the fifteen conditions into ten narrative paragraphs; the complete fifteen conditions are given in Appendix A.

Negation 1: Not sea. The tevah narrative scene is a river-plain flood, not an oceanic event. The water in the Noah narrative "rises" and "recedes," matching river-flood behavior rather than tsunami (single impact, immediate retreat) or marine transgression (rising without retreat). Excludes the Minoan Thera tsunami hypothesis, the Ryan–Pitman 1998 Black Sea Flood hypothesis, Persian Gulf marine transgression hypotheses, and all oceanic-disaster candidates.

Negation 2: Not peak. Genesis 8:4 records tevah coming to rest on the "mountains of Ararat." This is the narrative landing point, not the building site. A building site must be capable of hosting an annihilating flood, must accommodate large-scale construction, and must be near a navigable watercourse. Mountains fail these conditions. Excludes all searches treating "Mount Ararat," "Mount Judi" (the Muslim tradition's landing site), or any mountain as the physical prototype location of construction.

Negation 3: Not plain. The source event location must be one that permits the narrative perspective of "the entire known world is destroyed." Fully open plains, plateaus, or steppes make such a perspective narratively unsustainable — in open terrain, human vision extends far, and the sense of total closure is difficult to generate. Semi-enclosed low basins (surrounded by mountains or highlands) can generate this perspective. The Mesopotamian alluvial plain satisfies this condition (low-lying, ringed by mountains, subject to widespread river flooding).

Negation 4: Not distant. Must lie within the geographic range of the Hebrew transmission chain (Mesopotamia, the Levant, Egypt). This range is supported by converging linguistic, archaeological, and philological evidence. Excludes the Indus Valley civilization, the Yellow River basin, the Americas, sub-Saharan Africa, Southeast Asia, and Oceania. Independent flood narratives from these regions may have different origins or different transmission chains; they are not the source of Hebrew tevah narrative.

Negation 5: Not point. Tevah's textual dimensions (300 × 50 cubits = approximately 137 × 23 meters) indicate a structure on the hundred-meter scale. Candidate sites must have sufficient area to contain such a structure plus associated workspaces, channels, and supporting constructions. Excludes single small sites (areas under several thousand square meters).

Negation 6: Not dead. Candidate regions must show human activity during the source-event period (third millennium BCE and earlier), with material-culture remains. Fully uninhabited regions cannot produce a narrative perspective of "the entire known world." Excludes Arabian Peninsula interior deserts, the Empty Quarter, deep Syrian Desert, and other third-millennium uninhabited zones.

Negations 7 and 8: Not vessel (weakened) and not drift. Modern engineering tests (Finkel's team's 2014 quffah scaling experiment; broader research on wooden-ship limits including the 1909 Wyoming's performance) show that wooden ships at the 300-cubit scale are not feasible for actual water travel. Excludes the strong-form understanding of tevah as "a 137-meter wooden ship that actually sails." Methodological filtering also requires that the candidate has a traceable fixed anchor — pure drifting objects leave no geographic trace and cannot be located by filtering. Excludes pure drifting candidates. Negations 7 and 8 jointly require that the candidate be a fixed structure at a determinable location, though that structure may originally have been floatable (a raft, a raft-plus-mudhif composite).

Negation 9: Not low elevation. The Holocene marine transgression (c. 6000 to 5000 BCE, extending the Persian Gulf north to modern central Iraq) has buried pre-sixth-millennium surface elevations south of the transgression line under marine sediment. Physical evidence south of this line is inaccessible to direct filtering. Excludes locations south of Aqrawi 2001's described maximum transgression line (roughly modern Nasiriyah–Amarah latitude) for pre-sixth-millennium contexts.

Negation 10: Not exposed. If the tevah physical prototype were located in the core excavated zones of classical sites (the main mounds of Ur, Uruk, Nippur, Lagash, Eridu), a century of archaeology and textual documentation would already have produced a settled consensus on what it is. The absence of such consensus suggests that the prototype, if it exists, lies in a not-yet-fully-excavated location, or in a previously excavated but not-correctly-identified structure. Excludes the fully-excavated cores of the above sites; does not exclude their peripheral mounds, unexcavated areas, or structures within main mounds currently classified as "unidentified." This is a methodological heuristic, not a strict logical constraint.

Negation 11: Not far from water. Large-scale flood events require proximity to major river channels. Locations far from main channels are not subject to large-scale river flooding. The narrative's description (sustained, large-scale, covering "the entire known world") demands source events in channel-dense regions. Excludes locations more than five kilometers from the ancient Euphrates or Tigris main channels (as reconstructed in Cooke 1987). Desert oases, dried riverbeds, and settlements far from rivers are excluded.

Negation 12: Not without retreat trace. Flood events must leave traces in the geological record — deposits, avulsions, carbon layers, or other material evidence. "Pure narrative floods" (with no physical traces) have no physical anchor to be filtered. Forward constraint: candidate regions must have Holocene flood geological evidence. Excludes regions without such evidence.

Negation 13: Not without reference. Long-term oral narrative requires specific anchors to preserve detail. Abstract narrative without specific place names, king names, or event names cannot preserve precise attributes across thousands of years. The Sumerian King List anchors Ziusudra in Shuruppak — a traceable text-geography anchor pair. Excludes regions without pre-third-millennium textual memory. Most non-Mesopotamian Near East regions are excluded; within Mesopotamia, locations not mentioned in early texts are also excluded.

Negation 14: Not residential zone. Tevah's dimensions (137 × 23 meters) far exceed those of ordinary residences (Shuruppak Early Dynastic residences typically measure 10 × 15 meters). Structures at the hundred-meter scale must be anomalous buildings — temples, palaces, granaries, large public buildings, monumental structures, or large-scale constructions of uncertain function. Excludes ordinary residential quarters, artisan quarters, and burial areas, including zones of the Shuruppak main mound already identified as residential groups.

Negation 15: Not underground. Methodological filtering and future archaeological verification both require that candidates be detectable. Five thousand years of Mesopotamian sedimentation may bury third-millennium surfaces 2 to 5 meters deep (per Aqrawi 1995 sedimentation rates). Modern magnetometers can detect fired-brick anomalies to approximately 3 to 5 meters of depth. Structures deeper than this are neither filterable nor verifiable in the near term. Excludes structures buried deeper than 5 meters or fully eroded. Not a true existence constraint but an operational detectability constraint.

Conjunction. Candidates must simultaneously satisfy the negation of all fifteen conditions: river-plain low-lying location; within Mesopotamia–Levant range; sufficient scale (hundred-meter); adjacent to ancient channels; north of the transgression line; with Holocene flood evidence; within the memory range of textually-recorded ancient cities; not an ordinary residence; not in a fully-excavated core zone; surface-visible or at detectable depth; not oceanic, not mountain-peak, not an isolated point, not in an uninhabited zone, not a pure drifting object.

The conjunction dramatically compresses the candidate set. The first five negations narrow candidates from global scope to the Mesopotamian plain. The middle five narrow from the entire plain to the Shuruppak/Kish/Nippur core city circle. The final five narrow to "anomalous large-scale structures within the core city circle that have not been fully excavated." Within the publicly accessible material available to this author, the hardest-to-exclude candidate satisfying all fifteen conditions is the Area D structure at the southwestern Mound E of Tell Fara.

Each negation's supporting evidence and exclusion logic is given in Appendix A.

---

Section 6 · Geological Constraints

The Mesopotamian Holocene geological situation determines which ancient structures are visible today, which are buried under later sediment, which are eroded, and which are submerged. Geological constraints are hard constraints; they cannot be ignored. This section addresses specific evidence for the geological conditions referenced in Section 5's Negations.

Marine transgression line. The Holocene marine transgression (c. 6000 BCE onward) extended the Persian Gulf as far as modern central Iraq. Aqrawi 1995 (DOI: 10.1016/0264-8172(95)96903-4) and Aqrawi 2001 provide detailed stratigraphic data on the Mesopotamian plain's Holocene evolution. The approximate northern boundary of maximum transgression lies along the modern Nasiriyah–Amarah line (as described in Aqrawi 2001); pre-sixth-millennium locations south of this line are now deep under marine sediment. Tell Fara lies approximately 100 km north of this line, unaffected by transgression.

Sedimentation rate. Aqrawi 1995 gives an average Holocene sedimentation rate of approximately 1.3 to 2.2 mm per year on the Mesopotamian plain. Over five thousand years, this accumulates to approximately 6.5 to 11 meters of sediment. This implies that third-millennium surface structures may today lie at depths of 2 to 5 meters. Hahn et al. 2022's magnetometry can detect fired-brick anomalies at these depths — hence the visibility of Area D-type structures to magnetic survey.

Channel migration (avulsion). Morozova 2005 documents multiple Holocene avulsions of the Euphrates main channel. Avulsion differs from annual flood deposition (a few millimeters to centimeters per year): avulsion is a sudden, large-scale event (channel shifts of meters to kilometers). Cooke 1987's reconstruction of the ancient Euphrates routes the channel past Shuruppak. Area D's location on the southwestern side of Mound E lies adjacent to an ancient channel trace, consistent with "proximity to water."

Flood layer. Schmidt's 1931 excavation report at the Shuruppak main mound documents a flood deposit of approximately 60 cm at the base of the Early Dynastic stratum. Mallowan 1964 and Raikes 1966 confirm this figure in their respective Mesopotamian flood reviews. Raikes offers the precise comparison: approximately 40 cm (16 inches) at Kish, approximately 60 cm at Shuruppak. This deposit dates to approximately 2900 BCE and constitutes direct material evidence of the source-event era.

Tectonic subsidence. The Mesopotamian plain's neotectonic subsidence has long been a subject of geological discussion. The general position is that the plain has accumulated hundreds to thousands of meters of subsidence since the Upper Miocene. For Holocene-scale (last ten thousand years) specific subsidence rates, no precise regional figures are found in the publicly accessible literature. This paper does not use specific subsidence figures; it acknowledges tectonic subsidence only as a qualitatively present factor.

Combined geological constraints: Tell Fara (ancient Shuruppak) satisfies all relevant geological criteria — north of the transgression line (candidate visible); sedimentation rate within detectable depth (accessible to magnetometry); adjacent to ancient channels (water source and flood pathway available); direct flood-layer evidence (Schmidt 1931's 60 cm); textual memory (Ziusudra as "man of Shuruppak"). These conditions in conjunction further compress the candidate set.

---

Section 7 · Convergence to Area D

Integrating Section 5's fifteen negations with Section 6's geological constraints, the filtering process can be summarized as a progressive convergence funnel.

Layer 1 convergence: from global scope to the Mesopotamia–Levant range. Negations 1, 2, 4, and 9 directly exclude non-Mesopotamian candidates and the marine-transgression zone of southern Mesopotamia. Remaining candidates: the northern half of the Mesopotamian plain to the Levant.

Layer 2 convergence: from the entire Mesopotamian plain to the core ancient city circle. Negations 13, 11, and 12 require candidates within ancient-city memory ranges, adjacent to ancient channels, and with flood geological evidence. Locations meeting these conditions cluster around Shuruppak, Kish, Nippur, Eridu, Uruk, and Ur — the third-millennium core city circle.

Layer 3 convergence: from the core city circle to Shuruppak. Integrating textual evidence (the Sumerian King List anchors Ziusudra in Shuruppak, the final pre-flood dynasty), flood-layer evidence (Schmidt 1931's 60 cm layer at the base of the Shuruppak main mound), and etymology (the Egyptian origin of tevah, indicating the Hebrew tradition's vessel derives independently of the Mesopotamian boat tradition — though still within the overall transmission range of Mesopotamian flood narrative), Shuruppak emerges as the most consistent candidate city.

Layer 4 convergence: from the Shuruppak main mound to Mound E. Negation 10 requires that the candidate not be in a fully-excavated classical site core. The Shuruppak main mound has been studied since Andrae's 1902 excavation and is being re-surveyed by the LMU Munich team since 2022. Main mound contents are broadly classified. Anomalous structures lie in the mound periphery — Shuruppak's southwestern Mound E is an independent low rise separate from the main mound, briefly surveyed in 1902 and not subsequently excavated in depth until Hahn et al.'s 2022 magnetometry revealed the complexity of its internal structure.

Layer 5 convergence: Area D within Mound E. The 2022 magnetometry detected two anomalies in Mound E — an approximately 110-meter linear fired-brick structure, and a 20-to-26-meter-wide area of water or low-resistance (interpreted as a channel or ditch) with a T-shaped head. Einwag, Otto, and Fassbinder's 2025 follow-up analysis re-describes the structure as: linear component 115 × 15 meters, western adjacent rectangular platform 138 × 24 meters. Area D refers to the western 138 × 24 meter rectangular platform.

Posture of the filtering result. The output of this filtering is a candidate that Via Negativa, within the publicly accessible material available to this author, finds the hardest to exclude. This formulation is considerably more conservative than "we found it" or "this is the ark," and more precise.

It is necessary to clarify the hardness grade of Area D's candidate status. The term "physical-referent candidate" is too broad; it must be partitioned into three levels:

Tier I: Complete physical prototype candidate. All textual attributes (length 300 cubits, width 50 cubits, height 30 cubits, gopher wood material, pitch coating inside and out, side door, skylight, three internal levels, the function of carrying all living things through the flood) can be matched to the physical object. This is the strongest candidate status.

Tier II: Dimension-morphology anchor candidate. The geometric features and the location-period window of the textual attributes (length-width, rectangular plan, geographic range, possible period range, structural type) can be matched to the physical object. But material, complete function, three-dimensional height, internal details remain unverified. This is the medium-strength candidate status.

Tier III: Narrative-reference candidate. Some features of a physical object are reconstructed by later textual tradition into a narrative object. The physical object may not correspond to most details of the narrative description — it is invoked by narrative memory only through some core feature (for example, "there is an unusual large-scale anomalous structure at a certain place"). This is the weakest candidate status.

By the publicly accessible materials and this paper's filtering results, Area D is at most a Tier II (dimension-morphology anchor) candidate, not a Tier I (complete physical prototype) candidate. Specifically, Area D may account for the following subset of tevah textual attributes:

• Length approximately 137 meters (Hebrew cubit conversion close to Area D's measured 138 meters)
• Width approximately 23 meters (close to Area D's measured 24 meters)
• Rectangular plan (matches Area D's rectangular structure)
• Geographic location within the Mesopotamian ancient city circle (consistent with the Shuruppak textual-geographical memory)
• Possible period window between approximately 2700 and 2350 BCE (consistent with Andrae's 1903 classification of the wall to the Fara Z period)
• Structural type as a large anomalous construction (consistent with Andrae's "barely qualifies" evaluation and Einwag et al. 2025's "harbor installation" hypothesis)

Area D cannot account for the following textual attributes of tevah:

• Height of 30 cubits (Area D is a surface structure without a directly corresponding "height" dimension from current magnetometry)
• Gopher wood material (Area D is a fired-brick structure, not wood)
• Three internal levels (no interior has been excavated, no layering known)
• Side door, skylight (same as above)
• Pitch coating (same as above)
• The function of carrying all living things (this is a narrative function; the physical object can only support it indirectly)

Area D's candidate status is therefore strictly Tier II (dimension-morphology anchor candidate). Promoting it from Tier II to Tier I requires archaeological fieldwork (core drilling, radiocarbon dating, internal excavation, material analysis, function adjudication). These exceed the authority of this paper's methodological filtering.

Retention of Area D does not equal "Area D is the physical prototype of tevah." The nature of Via Negativa is exclusion. "Remaining candidate" means "a structure that has not yet been excluded given the fifteen constraints."

Conversely, retaining Area D as a Tier II candidate is a meaningful filtering result. It narrows the target range for future archaeological work. If one were to ask "what is the most probable location of tevah's physical referent within Mesopotamia," Area D should be among the first candidates for future core drilling.

---

Section 8 · The Empirical Data at Area D

This section presents all publicly available empirical data on Area D. These data span 120 years of observation, 1903 to 2025, three generations of researchers recording the same structure with the best tools of their time.

Andrae's 1903 original discovery. Walter Andrae co-led the 1902–1903 excavation of Fara for the Deutsche Orient-Gesellschaft. In the 1931 excavation report edited by Heinrich and Andrae, he describes a thick wall found at Mound E. The original text follows (book page 6, read directly from the Stony Brook AMAR open-access scan; archive number 585; entry at https://commons.library.stonybrook.edu/amar/585/):

> Hinsichtlich der Flächenausdehnung der Siedelung von Fara beschäftigte uns die Frage nach einer Befestigung. Da ist nur im Süden des Hügels ein kurzes Stück einer an beiden Enden abbrechenden dicken Mauer gefunden, die den Anspruch erheben könnte, Festungsmauer genannt zu werden. Ihre Lage und Richtung sind so, daß sie zur Not eine Südwestfront geschützt haben könnte. Da es Mauerwerk und nicht bloß Erdschüttung ist, möchte ich den Bau in die letzte Epoche, also in Fara Z setzen und die Mauer mit den Häusern zusammenbringen. Keiner unserer anderen Suchgräben hat die Umwallung sonst geschnitten, wir wissen also herzlich wenig davon…

English translation: Regarding the areal extent of the Fara settlement, the question of a fortification occupied us. Only in the southern part of the mound was found a short piece of a thick wall, broken off at both ends, which could perhaps claim the name of a fortress wall. Its position and orientation are such that it might, if need be, have protected a southwestern front. Since it is masonry and not merely earthen fill, I am inclined to place the construction in the last epoch, that is, in Fara Z, and to connect the wall with the houses. None of our other trial trenches cut through the enclosure, so we know very little about it…

Andrae's language deserves attention. Three subjunctive expressions — könnte (could), möchte ich (I am inclined), zur Not (if need be) — signal that he was already uncertain about this wall in 1903. He tentatively classified it as a fortification, tentatively assigned it to the Fara Z period (the latest layer in Andrae's own stratigraphic system, approximately 2700 BCE, based on cuneiform paleography), and acknowledged that the trial trenches may not have extended far enough.

Andrae's own stratigraphic system was: Fara H (Hüttenschichtung, reed-hut layer, earliest), Rundkeller (intermediate circular underground structures), Fara Z (Ziegel, brick-house layer, latest). This system is Andrae's 1903 construction-material-based division, not the Early Dynastic I/II/III system used in modern archaeology. Fara Z's absolute date, estimated by Andrae from the paleographic form of cuneiform tablets found within the houses, is approximately 2700 BCE.

On the following book page 7, Andrae continues on the same wall:

> Der Ruinenzustand an der Peripherie macht es wahrscheinlich, daß die Umwallung gänzlich vernichtet ist. In den älteren Schichtungen sind ebenfalls keine Spuren eines Erdwalls festgestellt, die Siedelung war damals möglicherweise offen und nur durch ihre Lage im Wasser geschützt. Siedelungen zu umwallen ist eine besondere Erfindung, die bekanntlich in Uruk dem Gilgamesch zugeschrieben wird.

English translation: The state of ruin at the periphery makes it probable that the enclosure is entirely destroyed. In the older strata no traces of an earthwork are discovered either; the settlement was perhaps open then, protected only by its position in water. Enclosing settlements is a particular invention, known to be attributed in Uruk to Gilgamesh.

Here Andrae invokes the Uruk legend of Gilgamesh building city walls as an analogy. He uses mythic narrative to explain the wall — but he does not connect it to Ziusudra or the flood narrative. This is one of his methodological blindspots: he uses a neighboring contemporary mythic figure (Gilgamesh) as analogy but does not think of the earlier mythic figure from the same city (Ziusudra). This blindspot is revisited in Section 10.

Hahn et al. 2022 magnetometry. Hahn, Fassbinder, Otto, Einwag, and Al-Hussainy 2022's paper in Archaeological Prospection (DOI: 10.1002/arp.1878, open access) reports the LMU Munich team's 2022 magnetic survey at Shuruppak's southwestern Mound E. Primary findings:

• A linear fired-brick structure approximately 110 meters long
• A 20-to-26-meter-wide area of low-resistance or water (interpreted as channel or moat)
• A T-shaped head structure (interpreted possibly as a water gate or extension)
• Coordinates 31.777222°N, 45.510833°E

The magnetic survey extends what Andrae saw in 1903 — 71 meters traced — to 110 meters, and reveals the adjacent water area and head structure that Andrae could not reach.

Einwag, Otto, Fassbinder 2025 further analysis. In Einwag, Otto, and Fassbinder 2025's "Harbours, Water Gates, and Canals in Šuruppak and Ur: New Results from Old Sites," published in Licia Romano (ed.), Sumer and the Sea: Deltas, Shoreline, and Urban Water Management in 3rd Millennium Mesopotamia (ARATTA 3, Brepols, DOI: 10.1484/M.ARATTA-EB.5.143624), pp. 91–100, the description is updated. The precise wording on p. 96:

Linear structure: 115 × 15 meters

Western rectangular platform: 138 × 24 meters

The 2025 version refines the 2022 110-meter linear structure to 115 × 15 meters and identifies an adjacent western rectangular platform of 138 × 24 meters. Area D refers to this western rectangular platform. The chapter hypothesizes a harbor installation but preserves uncertainty regarding function.

Conversion of tevah's textual dimensions. The tevah dimensions from Genesis 6:15 are 300 × 50 × 30 cubits. Converting at approximately 45.7 cm per Hebrew cubit:

Length = 300 × 0.457 = 137.16 meters

Width = 50 × 0.457 = 22.86 meters

Height = 30 × 0.457 = 13.72 meters

Degree of matching.

Area D's western rectangular platform: 138 × 24 meters

Tevah conversion (Hebrew cubit): 137.16 × 22.86 meters

Length difference: 0.84 meters (error 0.6%)

Width difference: 1.14 meters (error 5%)

It must be clearly stated: this is a two-dimensional length-width matching, not a complete three-dimensional matching. Area D is a magnetometrically visible surface structure without an observable height dimension directly corresponding to tevah's textual 30-cubit height. The three internal levels, side door, skylight, gopher wood material, and pitch coating mentioned in the tevah text have no current correspondence in the Area D magnetic structure; verification would require archaeological excavation.

Within reasonable bounds for cubit-unit choice and magnetometric boundary judgment, Area D and tevah's textual dimensions form a strikingly close match along two planar dimensions. The two datasets come from entirely independent systems: Area D's 138 × 24 meters from 2022–2025 magnetic measurement (a Category II measurement fossil), tevah's 137 × 23 meters from Hebrew-cubit conversion of a 300 × 50 cubit text (a Category III design-fossil candidate).

Constraint on the "same object" reading. If Area D and the tevah text belong to the same referent chain — that is, if Area D's physical-object design or measurement data entered the Hebrew tevah text through some narrative-transmission mechanism between the third millennium BCE and the sixth century BCE — then the meter-level difference can be understood as the normal deviation between design numbers and physical measurements (weathering erosion, construction error, measurement boundary judgment). But this "same referent chain" hypothesis itself requires other supporting evidence; this paper does not directly assert it. If future archaeological work demonstrates Area D's construction period is much later than Shuruppak's Early Dynastic era (for example, Old Babylonian harbor installation), then the dimensional match requires reinterpretation as coincidence or as shared standardized design patterns (138 × 24 meters is one engineering-optimized value for large rectangular structures).

On the role of dimensional matching in the argument. It must be emphasized: dimensional matching is not the starting point of this paper's filtering, nor an independently valid conclusion. Area D is retained as a candidate through the conjunction of Section 5's fifteen negations, Section 6's geological constraints, and Section 7's five-layer convergence. The length-width match is merely the most striking posterior support. Readers who interpret the entire paper as "observing a 138 × 24 meter structure and then matching Genesis dimensions to it" misread the argumentative flow. The argument is: independent structural filtering first, dimensional match observed afterward as additional evidence.

Metrological reversal observation. Section 4 already discussed: if Area D was indeed built in the Early Dynastic period, the original construction unit would have been the Sumerian cubit (approximately 49.6 cm), and 300 Sumerian cubits would be approximately 149 meters — differing from Area D's measured 138 meters by nearly 11 meters. The precise match between Hebrew cubit 45.7 cm and 138 meters may indicate that narrative transmission preserves not the number "300" but the physical length (approximately 138 meters), which Hebrew scribes, using their own cubit units, compute as 301.9 ≈ 300 cubits. If this "absolute-length preservation plus cross-unit conversion" reading holds, the match's explanation becomes more striking — the textual numbers are not fabricated but the precise result of cross-cultural metrological conversion during transmission. But this reading is speculative, not a demonstrated mechanism. Readers may adopt it or reject it. The match itself is a fact; the mechanism is open.

Discussion of period attribution. Andrae 1903 assigned the wall to Fara Z (approximately 2700 BCE). Hahn et al. 2022 and Einwag et al. 2025 offer no final dating. The modern magnetic survey reveals a structure larger and more complex than what Andrae saw, but magnetometry does not directly yield dates. Precise dating requires field core drilling and radiocarbon measurement. The possible attribution range is Early Dynastic II to III (approximately 2700 to 2350 BCE), covering Shuruppak's heyday and also the reconstruction period following Schmidt 1931's flood layer.

This paper does not force a specific construction date. It merely notes: if the construction date falls within this range, it sits precisely in a window where three conditions simultaneously hold:

(i) The Sumerian cubit unit is already standardized (supporting the possible production of a "300-cubit" design-fossil candidate)

(ii) Shuruppak functions as an active political center (supporting large-scale construction activity)

(iii) Post-Schmidt-1931-flood-layer (potentially triggering monumental reconstruction)

All three are conditional statements, not conclusions. Area D's precise construction date is the key pending item for future archaeological work.

---

Section 9 · Analysis of False Positives and False Negatives

Honest discussion of possible filtering errors is a necessary part of methodological work. This section addresses two error types.

False positives: Area D is erroneously retained, actually not the tevah prototype.

Scenario 1: Area D is an unrelated later-period structure (for example, Ur III harbor installation or Old Babylonian defensive installation), and the dimensional match is pure coincidence. This scenario cannot be excluded probabilistically. A 138 × 24 meter footprint is a common large-building dimension, and tevah's 300 × 50 cubits may be some standardized design ratio; the two matching may be "two independent standardized design decisions happening to coincide." Refuting this scenario requires precise dating of Area D — if construction is dated between 2700 and 2350 BCE, coincidence becomes harder to defend.

Scenario 2: The 138 × 24 meter measurement itself has boundary-judgment error. Magnetic survey results depend on human judgment of anomaly boundaries. Einwag et al. 2025's figures are best estimates; actual dimensions may lie within 135 to 141 meters long, 22 to 26 meters wide. This error range does not undermine the "meter-level match" conclusion but must be honestly acknowledged.

Scenario 3: The Hebrew cubit of 45.7 cm is a modern reconstruction; the actual cubit used in pre-first-millennium Hebrew tradition may differ. Ancient cubit units had multiple variants (common cubit, royal cubit, sacred cubit); different periods and regions of Hebrew tradition may have used cubits of varying length. If the actual unit was 44 cm, conversion yields 132 meters; if 47 cm, 141 meters. This uncertainty means "precise match" must be qualified as "match within reasonable cubit-range."

False negatives: the true tevah prototype is erroneously excluded, or no such prototype exists yet the filtering appears to have found a candidate.

Scenario 1: The true prototype may lie in a region the paper has excluded — an unexcavated small mound near Kish, a position in Nippur's periphery, Eridu's surroundings. The paper's filtering is based on publicly accessible archaeological data; for unexcavated regions it has no power.

Scenario 2: The true prototype may have been entirely covered or eroded by later construction. Five thousand years of human activity in Mesopotamia has high coverage rates; many third-millennium structures are fully superimposed by later buildings and invisible today.

Scenario 3: Tevah may have no single physical prototype; it may be a synthesis in narrative transmission of multiple real buildings (or multiple construction events). This is common in mythology — mythic objects are often crystallizations of historical events rather than records of them. If so, any attempt to "find a single physical prototype" commits a methodological category error; Area D would be at most one fragment of the composite.

Scenario 4: Tevah may be entirely narrative construction, with no physical prototype. This extreme hypothesis cannot be a priori excluded. If true, Area D's dimensional match is sheer coincidence. The paper's filtering and argument rest on the working hypothesis that there is a physical prototype; this hypothesis is itself refutable.

Response to Dickin 2018's hypothesis. Alan Dickin's 2018 article in Perspectives on Science and Christian Faith argues that the actual date of Noah's flood is approximately 5700 BCE, not the traditional 2900 BCE. His geological argument (Soreq Cave isotopes, Lake Van sediments, Sapropel 1 layer) supports 6500 to 5000 BCE as the wettest Holocene period, when Mesopotamia could have suffered unprecedented flooding. His archaeological argument (no settlement sites in southern Mesopotamia before 5500 BCE) suggests the 5700 BCE flood may have erased all prior settlements. He proposes tevah as a "raft-plus-mudhif" ontological candidate.

This paper's response to Dickin proceeds in two layers.

First, on physical events. Dickin's geological evidence is valid. A major flood may well have occurred in Mesopotamia around 5700 BCE. This does not conflict with the present argument. This paper addresses the physical-referent question of tevah as a mythic narrative object, not the question of "when did the physical flood event actually occur." The two questions can be treated independently.

Second, on narrative transmission. Dickin's hypothesis faces a difficulty. If the source event is around 5700 BCE, yet specific narrative details (the name Ziusudra, the place name Shuruppak, the 300 × 50 × 30 cubit dimensions) appear in texts dated no later than the second millennium BCE, then these details must have traversed more than three thousand years of oral transmission from source to textual record. Oral transmission can preserve structural motifs (annihilation, survival, renewal) across thousands of years — but preserving specific numbers, place names, and person names exceeds typical oral-transmission capacity. The unit-of-measurement constraint (Section 4) intensifies this difficulty: no standardized cubit existed before 2700 BCE, so "300 cubits" could not be inherited directly from building memory of that era.

Two-layer source model. Combining the above, this paper proposes a two-layer source model.

Distal layer (c. sixth millennium BCE): a real large flood or series of related flood events in Mesopotamia. Provides the later narrative with the structural motif of "world-annihilating flood." Most specific details are lost in oral transmission.

Proximal layer (c. 2900 BCE or thereafter): the Shuruppak flood layer (Schmidt 1931's 60 cm) and its subsequent reconstruction. Provides the existing structural motif with concretized physical anchors — specific place name (Shuruppak), specific king name (Ziusudra), specific vessel dimensions (300 × 50 × 30 cubits, if Area D is a post-flood monumental reconstruction). Specific details form on the eve of a literate society (Ur III from 2100 BCE onward permits written record), more readily preserved to the time of textual fixation.

This two-layer model is compatible with Dickin's single-layer model (5700 BCE) at the level of physical events but more consistent at the level of narrative transmission. This paper's Area D filtering addresses the proximal layer; it does not touch the distal layer. Dickin's 5700 BCE hypothesis addresses the distal layer; that layer need not provide specific details.

Structural explanation of the Egyptian etymological displacement. The two-layer source model leaves one subproblem: if the physical prototype (proximal layer) lies in the Mesopotamian heartland, why does the Hebrew text use an Egyptian loanword tevah (from Egyptian ṭbt, box/chest) rather than some Mesopotamian loanword (say, a derivative of eleppu)? This poses a geographical-linguistic tension — physical anchor to the east, etymological source to the west.

One possible structural explanation: the Hebrew tradition (situated in the Near Eastern corridor, long in contact with both Mesopotamian and Egyptian civilizations) faced a choice when settling on linguistic designation for a Mesopotamian-origin "non-vessel large rectangular construction." Mesopotamia's own word eleppu unambiguously denotes "ship" (a floatable paddled vessel), morphologically incompatible with Area D-type structures. Hebrew scribes using eleppu would introduce formal misdirection (leading readers to imagine a ship). Hebrew scribes not borrowing from the Mesopotamian word-system but turning instead to the morphologically correct Egyptian ṭbt ("box, chest, container") may represent a deliberate linguistic correction — to maintain formal consistency with the physical prototype, Hebrew tradition actively crosses the Mesopotamian word-system to find a morphologically matching word in the Egyptian word-system.

If this explanation holds, the Egyptian etymology is not evidence that tevah is unrelated to Mesopotamian tradition; rather, it is precisely the Hebrew tradition's linguistic insistence on physical-prototype form-accuracy in naming — a building is not a ship, so the word for "ship" cannot be used. This aligns directionally with the ontological observation in Section 3 (tevah is a rectangular building/composite, not in the Mesopotamian ship tradition).

A necessary qualification: this "linguistic correction" account is speculative, not a demonstrated fact. It is a hypothesis that happens to explain, requiring additional linguistic and historical evidence. Alternative accounts also exist — for instance, tevah's Egyptian etymology simply reflects Hebrew cultural contact at some historical period (perhaps around the Exodus) and has nothing to do with a Mesopotamian physical prototype, being merely a coincidence of borrowing. This paper lists the "linguistic correction" hypothesis as one possible structural explanation, letting readers see that the tension within the two-layer source model has a workable path, without asserting any particular explanation as the answer.

---

Section 10 · Methodological Reflection

This work yields four reflections on methodology itself.

First: Via Negativa's robustness under information incompleteness.

All work in this paper was conducted from publicly accessible materials; the author did no fieldwork, accessed no undigitized archives, contacted no excavation teams for unreleased data. Under information constraints, the paper nonetheless produces meaningful filtering output. This is not accidental but a property of Via Negativa.

Via Negativa's core operation is exclusion. Exclusion requires only enough counterexample material, not exhaustive confirmatory material. To claim "candidate X is not at this location" requires only showing that some known attribute of this location is incompatible with some textual attribute of X. Public archaeological data suffice for most such exclusions. The truly data-complete work is confirmation — claiming "candidate Y is tevah" requires showing that all attributes of Y match, which needs field drilling, radiocarbon dating, internal excavation, and so on.

Via Negativa partitions these two kinds of work. Exclusion uses public data and can be done at a desk. Confirmation requires fieldwork; this paper does not undertake it. This division allows Methodology VII to remain useful under a "working-from-a-computer" work mode.

Second: Applicability to long-standing puzzles.

Area D is not a new discovery. Andrae saw the wall in 1903 and frankly acknowledged uncertainty. The 1931 excavation report made this record public. Martin reassessed the material in 1988 (Fara: A Reconstruction of the Ancient Mesopotamian City of Shuruppak) and drew no firm conclusion. The LMU team re-observed the structure via magnetometry in 2022, extending it to 110 meters, but function remained uncertain. Einwag, Otto, and Fassbinder analyzed again in 2025, still preserving uncertainty.

For 120 years, three generations of researchers have handled the same structure with the best tools of their time; each generation advanced one step; each stopped at "we don't quite know what this is yet." Our Methodology VII intervention is not "solving" the puzzle. Its contribution is explicitation of filtering logic.

Earlier generations' filtering was implicit. Andrae said "barely qualifies as a fortress wall" — implicitly excluding other candidates (residence, temple, warehouse) without articulating why. The 2022 team said "possibly a harbor" — implicitly filtering without listing the fifteen constraints. Our Via Negativa writes out this filtering process — fifteen constraints, exclusion logic for each, convergence path for the conjunction.

Explicitation is itself a contribution. It does not guarantee that the filtering result is more correct than that of predecessors; but it makes filtering checkable, reproducible, and criticizable. The next generation of researchers can raise specific objections to each of our fifteen negations, rather than only saying "we have always felt this wall is some unidentified structure."

Third: Implicit precedents for Via Negativa at the archaeology-philology intersection.

Dickin 2018 in practice performs Via Negativa work. He excludes 2900 BCE (argument: too late, too small, Uruk World System already formed), excludes 2900 BCE's flood as the source event (too weak for annihilation), excludes the Black Sea Flood hypothesis (transgression, no ebb), and retains 5700 BCE as the most consistent source-event candidate. The workflow is structurally identical to this paper's, but Dickin does not name the methodology.

Some chapters of Mallowan 1964 have similar structure — he excludes several candidate flood layers as Noah's flood prototype (the upper Kish layer too late, the lower Ur layer too early, etc.) and retains the lower Shuruppak layer (around 2900 BCE) as the most consistent candidate. Again, implicit Via Negativa.

The position of Methodology VII at the archaeology-philology intersection is not inauguration but naming and regularizing an already-existing implicit practice. This positioning matters. It avoids the posture "we have invented a new method" — which invites challenge — and avoids "we are taking credit for others' work" — which is dishonest. We acknowledge that this method has always existed; our contribution is to convert it from implicit practice to a teachable and learnable explicit tool.

Fourth: Cross-domain applicability of the three-category numerical criterion.

The three-category numerical criterion from Section 4 (symbolic fictive / measurement fossil / design-fossil candidate) applies not only to tevah. It can be applied to any dimensional numbers in ancient texts. The Giza pyramid's 440-cubit base; the Etemenanki's layer dimensions; bay numbers in Chinese ritual architecture; column spacings in Greek temples; segment lengths in Roman aqueducts — wherever specific dimensional numbers appear in ancient texts, the three-category criterion is a potentially useful diagnostic.

The criterion's generality shows it is not an ad hoc tool designed for tevah. It is an independent methodological tool with applicability far exceeding the present application. Appendix E.2 offers detailed exposition.

---

Section 11 · Positioning in Mythological Studies

This section clarifies the position of this paper in the discipline.

Relation to the foundational paper of SAE Mythology.

SAE Mythology is founded in Myth as Remainder Intuition. That paper addresses the subject-function dimension of myth — myth's response to the 5DD–6DD structural remainder (the cannot-not preservation of individual boundary), a dimension that cannot be replaced by science. The present paper addresses a different dimension: physical reference — whether the specific objects mentioned in narratives have physical anchors, and how to identify them.

The two dimensions are complementary, not competing. A single mythic object can bear both functions simultaneously. Tevah as a subject-function vehicle carries in the Genesis flood narrative the remainder expression of "how individuals are preserved when the boundary of human totality is threatened by external forces." Simultaneously, tevah as physical reference may correspond to some real third-millennium Mesopotamian large-scale building. The two interpretations do not conflict ontologically; they enrich one another.

A mythic object that has only subject-function without physical reference (such as some purely metaphorical mythic figures) has that structural choice. A mythic object that has both functions has the complete structure. Tevah likely belongs to the latter category. This paper's work makes the physical-reference dimension of tevah explicit within mythological discussion.

Relation to traditional comparative mythology.

Eliade in The Sacred and the Profane observed that the sacred cannot be fully absorbed by the profane. Jung in collective-unconscious theory observed the cross-civilizational commonality of mythic structures. Tylor in Primitive Culture placed myth at an early rung of the evolutionary ladder. These traditions have their contributions and their limitations.

The difference between SAE Mythology (to which this paper belongs) and traditional comparative mythology: SAE Mythology makes the physical-reference dimension into an explicit subfield of mythology research. Traditional comparative mythology attends to motif, archetype, and structural commonality, all narrative-internal properties. SAE Mythology adds a narrative-external dimension — possible anchors of narrative objects in the physical world.

This is not replacement but supplement. Traditional tools (motif analysis, archetype analysis, structural analysis) remain in use in SAE Mythology application papers. The added dimension is the operational layer of "physical reference." The two together handle more complete mythic objects — understanding both why a myth has its narrative structure (traditional tools) and whether it corresponds to something in the physical world (the SAE-added dimension).

Relation to archaeology-philology intersection research.

The work of Mallowan, Dickin, and others lies at the intersection of archaeology and philology. They treat physical objects mentioned in specific texts (Genesis, Gilgamesh) and their possible correspondences in the archaeological record. This is an established research area.

This paper's contribution is not to open a new domain but to offer this established domain a set of explicit methodological tools. Predecessors implicitly employed Via Negativa, referent-chain analysis, and numerical reality-status diagnosis, without naming or regularizing these tools. This paper (and the cited Methodology VII, Methodology 00) makes them explicit, so future researchers can directly invoke them.

Relation to religious studies.

This paper explicitly does not address religious or theological questions. Whether tevah has a physical prototype and whether Genesis is divinely inspired are entirely distinct questions. This paper addresses only the former. The latter exceeds methodological authority and lies outside the scope of SAE Mythology.

Readers with religious commitments may read the filtering result as "supporting Genesis historical truth" or "contradicting Genesis historical truth." Both readings are extensions by the reader, not claims of this paper. This paper's conclusion is "Area D is, under currently accessible materials, the candidate hardest to exclude as physical referent of tevah," a conclusion that does not directly conflict with or directly support any religious position.

Relation to future SAE Mythology application papers.

As Application 1 of SAE Mythology, this paper establishes a working paradigm for subsequent application papers. Future applications (Application 2, 3, 4, and so on) may treat other mythic objects' physical reference (the Tower of Babel, Kunlun, the Xia dynasty); other mythic objects' referent drift (the Queen Mother of the West, Pangu, Prometheus); or other mythic-object groups' structural comparison (the flood-narrative family, the cosmogonic myth family, the hero's-descent-to-the-underworld family). Each application selects a methodology toolkit suitable for its object; Via Negativa is not mandatory. One contribution of this paper is to build a reusable skeleton.

---

Section 12 · Conclusion

This paper uses Via Negativa to conduct a complete methodological filtering of the physical-reference question for the mythic narrative object tevah. The filtering converges from global scope to the Area D structure at the southwestern Mound E of Tell Fara. Area D is, within publicly accessible material available to this author, the candidate hardest to exclude.

What we did: demonstrated the complete workflow of Via Negativa at the physical-reference dimension of mythology — reverse referent-chain tracing, three-category numerical criterion, systematic fifteen-negation exclusion, geological-constraint review, convergent candidate analysis, honest false-positive/false-negative discussion. Contributed two candidate methodological tools (Appendix E) as input to the SAE methodology series.

What we did not do: prove that Area D is the physical prototype of tevah. That adjudication exceeds the authority of methodological filtering and requires field archaeology — core drilling, radiocarbon dating, internal structural excavation, construction-period determination, functional attribution. This paper does not cross that authority boundary.

The value of this work lies not in "discovery" but in "regularization." Area D itself was already seen by Andrae in 1903, and his uncertainty is repeatedly cited in this paper. The 2022 and 2025 LMU team magnetometry updated our understanding, function still uncertain. This paper does not claim to have solved the 120-year puzzle. This paper merely writes out how to think about the puzzle, so that the thinking becomes checkable, reproducible, criticizable, and improvable. This is the typical posture of methodological work.

Via Negativa's commitment is finite. It commits to the best filtering result given current information, not to truth. It commits to making the filtering process explicit, not to making the filtering result unchallengeable. It commits to honest annotation of all material gaps and uncertainties, not to closing all open questions. This paper does its best to keep these commitments.

If future researchers, equipped with more material (more drilling data, more texts, more geological evidence), re-do this filtering and reach conclusions different from the present paper, that is not the paper's failure but the methodology's success — once the working process is made explicit, successors can continue on a clear basis.

Area D's truth requires archaeologists' work. Methodology's truth requires more applications to verify. SAE Mythology's truth requires the continuation of the entire series. This paper is one node.

---

Appendix A · Complete Derivation of the Fifteen Negations

The main-text Section 5 collapsed the fifteen conditions into ten narrative paragraphs. This appendix provides the complete fifteen conditions; for each, the premise, scope of exclusion, and logical relationship are listed.

Negation 1: Not sea

Premise: The flood narrative in Genesis 6–9 describes water that "rises" (עָלָה), "returns" (שׁוּב), and "recedes" (חָסֵר). Water movement follows the pattern of river flood and retreat, not tsunami (single impact then retreat) or marine transgression (rising without retreat).

Scope of exclusion: Minoan Thera's tsunami-driven hypothesis, Ryan–Pitman 1998's Black Sea Flood hypothesis, Persian Gulf marine-transgression hypothesis, all other candidates based on "oceanic disaster."

Logical relationship: Independent. This is the outermost filter, executed before other negations.

Negation 2: Not peak

Premise: Genesis 8:4 records that tevah comes to rest on the "mountains of Ararat" (הָרֵי אֲרָרָט). This is the narrative landing point at the close. The text does not suggest tevah was built on a mountain. Building site and landing site are two different locations.

Scope of exclusion: All searches treating "Mount Ararat," "Mount Judi" (the Muslim tradition's landing site), or other mountains as the physical prototype location of tevah's construction. This negation does not exclude the landing narrative itself.

Logical relationship: Independent. Parallel to Negation 1, both are direct textual constraints.

Negation 3: Not plain

Premise: The narrative requires a source event location that allows survivors to "believe this is the destruction of the entire known world." Fully open plains, plateaus, or steppes make this perspective narratively unsustainable — in open terrain, human vision extends far, producing no sense of total enclosure. Semi-enclosed low basins (ringed by mountains or highlands) can produce this perspective.

Scope of exclusion: Central Asian steppe, North American Great Plains, Sahara, Mongolian plateau, Siberian plain, Tibetan plateau, other open terrain.

Logical relationship: Weak dependence on Negation 1. Since Negation 1 excludes oceanic disasters, the source event is land flood, requiring basin-type water-accumulating terrain. Negation 3 is Negation 1's terrain-geomorphological corollary.

Negation 4: Not distant

Premise: The Hebrew narrative transmission chain's geographic range lies in the Near East (Mesopotamia, the Levant, Egypt). This is supported by converging linguistic, archaeological, and philological evidence.

Scope of exclusion: Indus Valley civilization, Chinese Yellow River basin, Americas, sub-Saharan Africa, Southeast Asia, Oceania's independent flood narratives. These civilizations' flood narratives may be independent in origin or have other transmission chains, not Hebrew tevah's source.

Logical relationship: Independent. A transmission-chain-based geographic constraint.

Negation 5: Not point

Premise: Tevah's textual dimensions (300 × 50 cubits = approximately 137 × 23 meters) point to a structure on the hundred-meter scale. Candidate sites must have sufficient space to contain such a structure plus associated workspace, channels, and supporting constructions.

Scope of exclusion: Single small sites with areas under 10,000 square meters. This excludes many village-scale settlements but does not exclude medium-sized cities or large public-building complexes' specific zones.

Logical relationship: Independent. A geometric corollary of the dimensional constraint.

Negation 6: Not dead

Premise: Candidate regions must show human activity during the source-event period (third millennium BCE and earlier), with material-culture remains. Fully uninhabited regions can produce neither an "annihilation" narrative (no one to annihilate) nor checkable material remains.

Scope of exclusion: Arabian Peninsula interior deserts, Empty Quarter, deep Syrian Desert, and other third-millennium uninhabited zones.

Logical relationship: Parallel to Negation 4. Negation 4 constrains geography by transmission chain; Negation 6 constrains geography by archaeological evidence.

Negation 7: Not vessel (weakened form)

Premise: Modern engineering experiments (Finkel's team's 2014 scaled quffah build; broader wooden-ship-limit research including the 1909 Wyoming's actual performance) show that wooden ships at the 300-cubit scale are not engineering-feasible for actual floating.

Scope of exclusion: The strong-form ontological interpretation of tevah as "a 137-meter wooden ship that can actually be launched and sailed."

Logical relationship: Weakened negation. This negation does not fully exclude the vessel ontology — it excludes only the strong form of "actually floatable and sailable"; it retains the weak form of "having vessel-like form but static." Works cooperatively with Negation 8.

Negation 8: Not drift

Premise: Methodological filtering requires that candidates have a fixed anchor that can be traced. Pure drifting objects leave no geographic trace and cannot be located by filtering. If tevah's prototype were indeed some drifting object, it either came to rest in some fixed location (filterable) or disappeared entirely (unfilterable). This paper addresses only the former case.

Scope of exclusion: Pure drifting-object candidates (hypothetical objects with no fixed resting point).

Logical relationship: Works cooperatively with Negation 7. Negation 7 excludes the "actually-sailable" vessel ontology; Negation 8 excludes "pure drift with no fixed anchor." Together they require the candidate to be a fixed-location structure (possibly a building, possibly a fixed raft, possibly a raft-plus-mudhif composite, but not purely drifting).

Negation 9: Not low elevation

Premise: The Holocene marine transgression (c. 6000 to 5000 BCE) brought the Persian Gulf north as far as modern central Iraq. The pre-sixth-millennium surface south of the maximum transgression line is now deep under marine sediment. If the source event occurred south of the transgression line, physical evidence is inaccessible to direct filtering.

Scope of exclusion: Pre-sixth-millennium locations south of Aqrawi 2001's described maximum transgression line (roughly the modern Nasiriyah–Amarah line).

Logical relationship: Weak overlap with Negation 3. Negation 3 requires semi-enclosed basin terrain; Negation 9 excludes the submerged southern basin zone. The intersection — the northern half of the Mesopotamian southern alluvial plain (north of the transgression line) — is the joint constraint.

Negation 10: Not exposed

Premise: If tevah's physical prototype were located in the core excavated zones of classical sites (Ur main mound, Uruk main mound, Nippur main mound, Lagash main mound, Eridu main mound, and so on), the past century of excavation, classification, and documentation would already have produced a settled consensus in the literature. The absence of such consensus suggests that the prototype, if it exists, lies in a not-yet-fully-excavated location, or in previously excavated but currently-misidentified structures.

Scope of exclusion: The fully-excavated core zones of the above sites. Not excluding their peripheral mounds, unexcavated zones, or parts of main mounds currently classified as "unidentified structures."

Logical relationship: This is a methodological heuristic, not a strict logical constraint. It reasons from "absence of evidence" — if the prototype were in a core zone, the literature would have records. This reasoning is not airtight (the prototype may be in a core zone but misclassified), but it is a reasonable prioritization for filtering.

Negation 11: Not far from water

Premise: Large-scale flood events require proximity to major river channels. Locations far from main channels are not subject to large-scale river flooding. The narrative's flood description (sustained, large-scale, covering "the entire known world" from the narrative perspective) demands source events in channel-dense regions.

Scope of exclusion: Locations more than five kilometers from the ancient Euphrates or Tigris main channels (per Cooke 1987's reconstruction). Desert oases, dried riverbeds, settlements far from rivers — all excluded.

Logical relationship: Cooperative with Negations 3 and 9. Negation 3 requires basin terrain; Negation 9 requires being north of the transgression line; Negation 11 requires adjacency to ancient channels. Their conjunction restricts candidates to the ancient-channel corridor of the central Mesopotamian alluvial plain.

Negation 12: Not without retreat trace

Premise: A flood event must leave traces in the geological record — deposits, avulsion, carbon layers, or other material evidence. A "pure narrative flood" (with no physical traces) has no physical anchor for filtering.

Scope of exclusion: Regions without Holocene flood geological evidence.

Logical relationship: Forward constraint. Negation 12 does not exclude positions; rather, it requires candidate regions to have forward-facing geological evidence. A region that satisfies Negations 1 through 11's reverse side but lacks Negation 12's forward evidence (no Holocene flood layer) is still excluded.

Negation 13: Not without reference

Premise: Long-term oral narrative needs specific anchors to preserve detail. Pure abstract narrative without specific place names, king names, and event names cannot preserve precise attributes across thousands of years. The Sumerian King List anchors Ziusudra in Shuruppak — a traceable text-geography anchor pair.

Scope of exclusion: Regions without third-millennium-or-earlier textual memory. Excludes most non-Mesopotamian Near East regions and locations within Mesopotamia unmentioned in early texts.

Logical relationship: Distinct from Negation 4. Negation 4 constrains the geographic range of the transmission chain (where Hebrew tradition could have reached); Negation 13 constrains the philological range (where textual memory can anchor specific details). The conjunction is stricter than either alone.

Negation 14: Not residential zone

Premise: Tevah's dimensions (137 × 23 meters) far exceed those of ordinary residences (Shuruppak Early Dynastic residences' typical dimensions approximately 10 × 15 meters). Hundred-meter-scale structures must be anomalous buildings — temples, palaces, granaries, large public buildings, monumental structures, or large-scale constructions of uncertain function.

Scope of exclusion: Ordinary residential zones, artisan zones, burial zones. Excludes regions of the Shuruppak main mound already identified as residential groups.

Logical relationship: Cooperative with Negation 5. Negation 5 requires sufficient space for a 137-meter structure; Negation 14 requires that the candidate structure itself be an anomalous large building. Their conjunction requires not just space but structural type.

Negation 15: Not underground

Premise: Both methodological filtering and future archaeological verification require that candidates be detectable. Five thousand years of Mesopotamian sedimentation may bury third-millennium surfaces 2 to 5 meters deep (per Aqrawi 1995 sedimentation rates). Modern magnetometers can detect fired-brick anomalies at approximately 3 to 5 meters depth. Structures deeper than this are neither filterable nor verifiable in the near term.

Scope of exclusion: Structures buried deeper than 5 meters or fully eroded.

Logical relationship: Operational constraint. Negation 15 is not a true existence constraint (the prototype may indeed be deeply buried); it is an operational detectability constraint. It defines the physical accessibility range of methodological filtering.

Logical dependency structure of the fifteen negations

Independent negations (not depending on other negations): 1, 2, 4, 5, 6, 11, 12, 13

Dependent negations (based on the standing of other negations):

• Negation 3 depends on Negation 1 (land-flood corollary)
• Negations 7 and 8 cooperate (joint ontological constraint)
• Negation 9 weakly overlaps with Negation 3 (basins north of the transgression line)
• Negation 10 is a heuristic constraint ("absence of evidence" reasoning)
• Negation 14 depends on Negation 5 (structural size-function joint constraint)
• Negation 15 is an operational constraint (detectability boundary)

Convergence path by conjunction:

Layer 1 (Negations 1, 2, 4, 6) filters from global scope to Near East land

Layer 2 (Negations 3, 9, 11) filters from Near East to the ancient-channel zone of the central Mesopotamian alluvial plain

Layer 3 (Negations 12, 13) filters from the ancient-channel zone to third-millennium ancient city circles with flood-layer evidence

Layer 4 (Negations 10, 5, 14) filters from ancient city circles to not-yet-fully-researched anomalous large non-residential structures

Layer 5 (Negations 15, 7, 8) restricts candidates to magnetometrically detectable, fixed-location, non-strict-ship-ontology structures

Tell Fara (ancient Shuruppak) southwestern Mound E's Area D simultaneously satisfies all the above constraints. This does not imply Area D is tevah's prototype; it implies only that Area D is the hardest-to-exclude candidate under the fifteen constraints used in this paper.

Independent reader verification

One purpose of this appendix is to allow readers to independently verify the filtering logic. Readers who disagree with the paper's conclusion have three specific operations:

One: Add a new negation. If a reader believes the paper has omitted some constraint, they can propose "Negation 16," "Negation 17." If the added constraint further excludes Area D, the conclusion changes.

Two: Object to a particular negation. If a reader believes some negation does not hold (for example, believing tevah can be a strict floating ship, objecting to Negation 7; or believing the source event can be oceanic, objecting to Negation 1), they can offer counter-reasoning. If successful, the filtering space expands; Area D may no longer be the unique remaining candidate.

Three: Propose missed candidates. If a reader finds candidates beyond Area D under the fifteen constraints (positions this paper has not noted), they can raise them. This does not negate Area D but means the candidate set is no longer a single point.

All three operations are natural consequences of making the filtering logic explicit. This appendix exists to support them.

---

Appendix B · Genealogy of the Tevah Referent Chain

Node list

Downstream nodes (first through fourth centuries CE):

• Latin Vulgate arca (approximately 400 CE)
• Syriac Peshitta qebutha (approximately 200 CE)
• Greek Septuagint κιβωτός (approximately third century BCE)

Midstream node:

• Hebrew Masoretic tevah (text fixation between sixth and second centuries BCE)

Upstream nodes:

• Berossos version (via Alexander Polyhistor; transmitted via Greek secondary sources)
• Gilgamesh Epic Tablet XI eleppu (approximately thirteenth to seventh centuries BCE)
• Atrahasis Epic vessel (approximately seventeenth century BCE, Old Babylonian)
• Ark Tablet circular quffah (approximately seventeenth century BCE, c. Finkel 2014's interpretation)
• Sumerian Flood Story vessel (approximately twenty-first century BCE, Ur III and earlier)

Egyptian subbranch: Hebrew tevah ← Egyptian ṭbt (box, chest, container). The etymology proposed by Brugsch and Erman in the late nineteenth century, adopted by BDB, continued by HALOT.

Key breakpoint: Hebrew tevah (rectangular) vs. Mesopotamian eleppu (square)

Hebrew tevah: rectangular, 300 × 50 × 30 cubits, ratio 10 : 1.67 : 1. Egyptian-origin etymology, connoting "box, chest."

Mesopotamian eleppu: square, all sides equal, ešer nindanā ("ten rods"). Etymology from the Mesopotamian word-system, connoting "ship" (a sailable paddle vessel).

The Hebrew Bible does not use eleppu or a cognate for tevah. This represents a breakpoint in the referent chain — the Hebrew tradition does not continue the Mesopotamian boat-word lineage but draws across the word-system to take an Egyptian container-word. (Section 3's main text and Section 9's Egyptian etymological-displacement explanation address this in detail.)

---

Appendix C · Key Magnetometric Data of Area D

C.1 Referenced figures

This appendix references the key magnetometric figures from Hahn et al. 2022 and Einwag et al. 2025. Figures are not reproduced here due to copyright considerations; readers should consult the original papers directly.

Hahn et al. 2022 magnetometric map: see Figure 7 in Archaeological Prospection 29, pp. 623–635 (DOI: 10.1002/arp.1878, open access). Shows the anomaly distribution of Mound E, including the approximately 110-meter linear structure and the 20-to-26-meter-wide low-resistance zone with T-shaped head.

Einwag et al. 2025 Area D structural description: see pp. 91–100 of Sumer and the Sea: Deltas, Shoreline, and Urban Water Management in 3rd Millennium Mesopotamia (ARATTA 3, Brepols 2025; DOI: 10.1484/M.ARATTA-EB.5.143624). Page 96 contains the precise dimensional description (115 × 15 meters linear structure; 138 × 24 meters western rectangular platform).

Coordinates: 31.777222°N, 45.510833°E. Open satellite imagery (Google Earth, Sentinel-2) at these coordinates shows the general Mound E topography.

C.2 Data summary

Source	Year	Linear structure	Rectangular platform	Water/channel
Heinrich & Andrae 1931 (original report)	1903	~71 m traced	not identified	not identified
Hahn et al. 2022 (magnetometry)	2022	~110 m, fired brick	not explicitly identified	20–26 m wide + T-head
Einwag et al. 2025 (revised)	2025	115 × 15 m	138 × 24 m	acknowledged

The progression from 1903 to 2025 represents iterative refinement by three generations of researchers, each extending the understanding but none reaching definitive function attribution.

---

Appendix D · Genesis 6:15 in Four Languages

Hebrew (Masoretic Text):

וְזֶ֕ה אֲשֶׁ֥ר תַּעֲשֶׂ֖ה אֹתָ֑הּ שְׁלֹ֧שׁ מֵא֣וֹת אַמָּ֗ה אֹ֚רֶךְ הַתֵּבָ֔ה חֲמִשִּׁ֤ים אַמָּה֙ רָחְבָּ֔הּ וּשְׁלֹשִׁ֥ים אַמָּ֖ה קוֹמָתָֽהּ׃

(English: "And this is how you shall make it: the length of the ark shall be three hundred cubits, its breadth fifty cubits, and its height thirty cubits.")

Greek (Septuagint):

καὶ οὕτως ποιήσεις τὴν κιβωτόν· τριακοσίων πήχεων τὸ μῆκος τῆς κιβωτοῦ καὶ πεντήκοντα πήχεων τὸ πλάτος καὶ τριάκοντα πήχεων τὸ ὕψος αὐτῆς.

Syriac (Peshitta):

ܘܗܟܢܐ ܥܒܕ ܠܗ ܠܩܒܘܬܐ܂ ܬܠܬܡܐܐ ܐܡܝܢ ܢܗܘܐ ܐܘܪܟܗ܂ ܘܚܡܫܝܢ ܐܡܝܢ ܦܬܝܗ܂ ܘܬܠܬܝܢ ܐܡܝܢ ܪܘܡܗ܂

Latin (Vulgate):

et sic facies eam: trecentorum cubitorum erit longitudo arcae, quinquaginta cubitorum latitudo, et triginta cubitorum altitudo illius.

Key observations

The four versions maintain perfect agreement on the 300 × 50 × 30 cubit figures. Across the Hebrew text (fixed approximately sixth century BCE), Greek (third century BCE), Syriac (second century CE), and Latin (fourth century CE) — spanning approximately one thousand years of textual transmission — the dimensional numbers are preserved precisely.

The four-language terms for the ark:

• Hebrew: תֵּבָה (tevah)
• Greek: κιβωτός (kibōtos)
• Syriac: ܩܒܘܬܐ (qebutha)
• Latin: arca

All four words correspond in their respective languages to "box, chest, container," not to "ship" (Greek ναῦς, Latin navis, and so on). This aligns with the ontological candidates for tevah (vessel / fixed building / raft-plus-mudhif composite) — the traditional understanding treats tevah as "a special kind of container," not "a ship."

---

Appendix E · Two Candidate Methodological Tools Surfacing in the Work

Positioning statement

This appendix records two candidate methodological tools that surfaced during the present application work. They are submitted as input to the SAE methodology series (Via Negativa, Via Rho, and any future methodology papers); whether they enter the formal methodology toolkit is determined by the methodology series' own development. This appendix is a working note, not a methodological proposal.

E.1 · The Boundary of Via Negativa under Referent Drift

Problem statement

When research objects undergo long-term referent drift, at what stages can Via Negativa effectively operate?

Observation

Apophatic theology (the origin of traditional Via Negativa) treats postulated-existence objects — "what is God?" — where "God" as an object has not undergone fundamental object-transformation across historical stages. Referent drift (the nature of this paper's object) is different — "what is tevah in the sixth century BCE?" and "what is Ziusudra in the twenty-first century BCE?" may point to fundamentally different physical referents, with only a narrative-transmission chain connecting them.

Candidate boundary

Via Negativa, for referent-drift objects, can operate directly only on the endpoint form of drift. "Endpoint form" means the stable form after textual fixation — for tevah, this is the Hebrew Masoretic form; for other objects, it is each tradition's final stable version.

On the endpoint form, Via Negativa can operate effectively: given textual attributes (rectangular, 300 × 50 × 30 cubits, and so on), exclude incompatible physical candidates.

On intermediate or origin forms, directly applying Via Negativa fails. The attributes of early forms may have been lost or distorted during transmission; the attributes preserved in the text may not be the early real attributes. Directly applying textual attribute constraints to early forms may exclude the true early prototype.

Operating guide

For filtering referent-drifting objects with Via Negativa, two steps:

One, first conduct referent-chain analysis (the tool of Appendix B), identifying each stage's form and its transmission relationships.

Two, apply Via Negativa only on the endpoint form of drift (or the stable textual form).

Three, for early forms, if filtering is needed, first use referent-chain analysis to reverse-engineer each early form's possibly independent attributes, then filter on each independent attribute set.

Specific application to tevah

This paper's Via Negativa filtering takes Hebrew MT tevah (the drift endpoint) as starting point, using Genesis 6:15's attributes as constraints. It does not directly filter on Ziusudra or Atrahasis forms. This accords with the tool's usage guidance.

Scope of applicability

This tool may apply to all mythic objects undergoing long-term referent drift: the Queen Mother of the West, Prometheus, Odin, Pangu, Houyi, Cthulhu (as an artificial-mythology control), and many others. Detailed applications are reserved for future SAE Mythology application papers.

E.2 · Reality-Status Criterion for Numbers V2 (Three-Category Diagnostic)

Background

Methodology VII's existing numerical treatment is a binary classification (fictive vs. factual). This application work shows that the binary is insufficient for treating round-but-non-symbolic engineering numbers. The third category, "design-fossil candidates," is an extension of the existing tool.

Criterion structure

Category I: Symbolic fictive numbers

• Features: round and corresponding to known cultural symbolic values
• Origin: narrative composition or cultural inheritance
• Treatment: not used as physical evidence
• Examples: seven days, forty days, twelve tribes, Noah's 950 years, Abraham's 75 years at departure from Haran, Jesus's twelve apostles

Category II: Measurement fossils

• Features: non-round, typically carrying decimals or scattered integers
• Origin: direct physical measurement
• Treatment: hard evidence of physical events
• Examples: Shuruppak's 60 cm flood layer, Kish's 40 cm flood layer, Area D's magnetometric 138 × 24 m, Schliemann's Troy VIIa wall thickness, Pompeii's 79 CE volcanic-ash layer thickness

Category III: Design-fossil candidates

• Features: round but non-symbolic, combinations exhibiting engineering ratios
• Origin: builders' design decisions
• Treatment: strong diagnostic signal for construction events, requires other evidence for confirmation
• Examples: tevah's 300 × 50 × 30 cubits (ratio 10 : 1.67 : 1), Giza's 440 Egyptian royal cubits, Etemenanki ziggurat's layer dimensions, Forbidden City's bay numbers

Application order

One, first check roundness. Non-round → Category II directly.

Two, round ones are checked for symbolic correspondence. If matched → Category I.

Three, round non-symbolic ones are checked for proportional structure. Simple symmetric ratios (1:1, 1:2) → possibly Category I. Engineering ratios (10:1.67, 1:1.618) → Category III candidate.

Accompanying constraint: time-floor for units of measurement

Design-fossil production requires standardized units of measurement. The ancient Near East's earliest standardized cubit unit appeared around 2700 BCE. No pre-2700-BCE source could generate round design numbers like "300 cubits." Different civilizations' standardization times vary and must be checked per culture:

• Sumerian cubit (kùš): approximately 2700 BCE, Early Dynastic II onward
• Egyptian royal cubit (meh): approximately 2700 BCE, Third Dynasty onward
• Hebrew cubit (amah): derived from earlier Near Eastern systems, second millennium BCE
• Chinese Zhou-period chi: Western Zhou onward

For design-fossil numbers in any ancient text, the source-civilization's standardization period must be checked. Numbers cannot come from before standardization.

Application to tevah

Tevah's 300 × 50 × 30 cubits satisfies the Category III candidate test: 300 is not a symbolic value; the 10 : 1.67 : 1 ratio is an engineering proportion. The unit-of-measurement constraint indicates the physical origin of this number set is not earlier than approximately 2700 BCE. This directly constrains any hypothesis placing the source event before 2700 BCE.

Scope of applicability

This tool may apply to dimensional numbers in any ancient text. Gilgamesh XI's eleppu dimensions in ešer nindanā ("ten rods"), Homer's description of Achilles's shield, the distances in various locations of the Shanhaijing, the yojana distances in Indian Vedic texts, the construction dimensions in the Chinese Kaogong Ji — the three-category criterion offers a general diagnostic framework.

---

Joint posture of the two candidate tools

These two tools are byproducts of the present tevah application work.

On generality. Both tools' applicability far exceeds this paper's tevah problem. Appendix E.1 (Via Negativa's operating boundary under referent drift) may apply to any long-drifting mythic object — Queen Mother of the West, Prometheus, Odin, Pangu, Houyi, Cthulhu (as artificial-mythology control), Shanhaijing place names, Kunlun, second-millennium Erlitou–Xia referent correspondence, and so on. Appendix E.2 (the three-category numerical diagnostic) may apply to dimensional numbers in any ancient text — pyramids, ziggurats, Greek temple modules, Roman aqueducts, Chinese palace construction dimensions, Indian yojana distances, dimensions in Homer — the two tools are independent of tevah; the present paper is merely their first application occasion.

On positioning. This appendix records; it does not promote. Whether the two tools merit inclusion in SAE's formal methodology toolkit (Via Negativa, Via Rho, or future methodology papers) is determined by the methodology series' own development. They are submitted as input, not with a presumption of adoption. If the methodology series discovers better tools in its development, these two tools may be replaced; if they prove useful, they may be absorbed. This appendix makes no presuppositions about future directions.

Offered to future versions of the SAE methodology series.

---

References

Honest declaration regarding this reference list: entries are ordered alphabetically by author surname. Each entry includes an indication of the author's access level: "full text read," "partial pages read," "abstract only read," or "via secondary citation" (the present author has not directly accessed the original; the citation is mediated through a named secondary source). This transparency is basic hygiene for a methodological application paper; readers may use it to judge the evidentiary grade of each citation.

---

Aqrawi, Adnan A. M. 1995. "Correction of Holocene Sedimentation Rates for Mechanical Compaction: The Tigris-Euphrates Delta, Lower Mesopotamia." Marine and Petroleum Geology 12 (4): 409–416. DOI: 10.1016/0264-8172(95)96903-4. [Partial pages read: abstract and key sedimentation-rate figures (1.3–2.2 mm/yr)]

Aqrawi, Adnan A. M. 2001. "Stratigraphic signatures of climatic change during the Holocene evolution of the Tigris-Euphrates delta, lower Mesopotamia." Global and Planetary Change 28 (1–4): 267–283. DOI: 10.1016/S0921-8181(00)00078-3. [Full text read, ResearchGate open access]

Bar-Matthews, Miryam, Avner Ayalon, and Aaron Kaufman. 1997. "Late Quaternary Paleoclimate in the Eastern Mediterranean Region from Stable Isotope Analysis of Speleothems at Soreq Cave, Israel." Quaternary Research 47 (2): 155–168. [Via secondary citation through Dickin 2018; present author has not read the original]

Bar-Matthews, Miryam, Avner Ayalon, and Aaron Kaufman. 2000. "Timing and Hydrological Conditions of Sapropel Events in the Eastern Mediterranean, as Evident from Speleothems, Soreq Cave, Israel." Chemical Geology 169 (1–2): 145–156. [Via secondary citation through Dickin 2018]

Brown, Francis, Samuel Rolles Driver, and Charles A. Briggs. 1906. A Hebrew and English Lexicon of the Old Testament [BDB]. Oxford: Clarendon Press. Commonly cited as BDB. The תֵּבָה entry appears in the relevant etymological section. Multiple open digital editions are available (for example, blueletterbible.org, biblehub.com). [Partial pages read: תֵּבָה entry and Egyptian-etymology discussion]

Brugsch, Heinrich. 1892 (or related year). On the Egyptian ṭbt to Hebrew tevah etymology, see Zeitschrift der Deutschen Morgenländischen Gesellschaft 46 (1892) or contemporaneous articles. Erman's related discussion appears in contemporaneous articles. Honest annotation: the present paper's references to Brugsch-Erman etymological arguments rest on BDB and HALOT secondary citations; the 1892 original German article has not been directly read. Journal digitization is available through the MENAdoc library (https://menadoc.bibliothek.uni-halle.de/).

Cooke, G. A. 1987. "Reconstruction of the Holocene coastline of Mesopotamia." Geoarchaeology 2 (1): 15–28. [Abstract only read]

Dalley, Stephanie. 1989 (revised 2000). Myths from Mesopotamia: Creation, the Flood, Gilgamesh, and Others. Oxford: Oxford University Press. [Partial pages read: Atrahasis and Gilgamesh XI relevant chapters, mainly cross-checked through Dickin 2018's citations]

Dickin, Alan. 2018. "New Historical and Geological Constraints on the Date of Noah's Flood." Perspectives on Science and Christian Faith 70 (3): 176–193. [Full text read, open access at https://www.asa3.org/ASA/PSCF/2018/PSCF9-18Dickin.pdf]

Einwag, Berthold, Adelheid Otto, and Jörg W. E. Fassbinder. 2025. "Harbours, Water Gates, and Canals in Šuruppak and Ur: New Results from Old Sites." In Sumer and the Sea: Deltas, Shoreline, and Urban Water Management in 3rd Millennium Mesopotamia, edited by Licia Romano, ARATTA 3, 91–100. Turnhout: Brepols. DOI: 10.1484/M.ARATTA-EB.5.143624. [Partial pages read: page 96 precise quotation ("115 × 15 m linear structure, western 138 × 24 m rectangular platform"). The full chapter is under Brepols paywall; the present paper's page-96 citation is based on an LMU Munich academic repository excerpt; other content on the same page has not been fully read]

Finkel, Irving. 2014. The Ark Before Noah: Decoding the Story of the Flood. London: Hodder & Stoughton. [Partial pages read: core chapters on the Ark Tablet (BM 78941) cross-checked through Internet Archive Lending and Dickin 2018's citations. A popular work; the present paper cites only the data point of the circular vessel's approximate 70-meter diameter]

George, Andrew R. 2003. The Babylonian Gilgamesh Epic: Introduction, Critical Edition and Cuneiform Texts. 2 vols. Oxford: Oxford University Press. ISBN 978-0198149224. [Via secondary citation: the present paper has not directly accessed George 2003 (copyright-restricted, requires specialized library access). References to Gilgamesh XI lines 24–31 and 57–63 on vessel description are cross-checked through SOAS BAPLAR's open translation (https://baplar.soas.ac.uk/) and George's 2000 Penguin popular edition, The Epic of Gilgamesh: A New Translation]

George, Andrew R. 2000. The Epic of Gilgamesh: A New Translation. London: Penguin Classics. [Partial pages read: Tablet XI chapter for cross-checking the ešer nindanā translation]

Hahn, Stefanie E., Jörg W. E. Fassbinder, Adelheid Otto, Berthold Einwag, and Alaa A. Al-Hussainy. 2022. "Revisiting Fara: Comparison of merged prospection results of diverse magnetometers with the earliest excavations in ancient Šuruppak from 120 years ago." Archaeological Prospection 29: 623–635. DOI: 10.1002/arp.1878. [Full text read, Wiley open access]

Heinrich, Ernst, and Walter Andrae, editors. 1931. Fara: Ergebnisse der Ausgrabungen der Deutschen Orient-Gesellschaft in Fara und Abu Hatab 1902/03. Berlin: Deutsche Orient-Gesellschaft. An open-access digital scan is provided by the Stony Brook University Archive of Mesopotamian Archaeological Reports (AMAR), record 585, entry at https://commons.library.stonybrook.edu/amar/585/, PDF direct link https://commons.library.stonybrook.edu/cgi/viewcontent.cgi?article=1584&context=amar. [Partial pages read: book pages 6–7 (scan PDF pages 18–19) containing the Andrae original text on the Mound E thick wall, directly read from this scan. Other portions of the book (table of contents, plates, artifact descriptions) have not been fully read]

Hill, Alan E. 2006. "Quantitative Hydrology of Noah's Flood." Perspectives on Science and Christian Faith 58 (2): 130–141. [Abstract only read, as background to Dickin 2018]

Hill, Carol A. 2002. "The Noachian Flood: Universal or Local?" Perspectives on Science and Christian Faith 54 (3): 170–183. [Abstract only read]

Hill, Carol A. 2006. "Qualitative Hydrology of Noah's Flood." Perspectives on Science and Christian Faith 58 (2): 120–129. [Abstract only read]

Jensen, Peter. 1889. Early etymological discussion of Akkadian terminology, Zeitschrift für Assyriologie 4: 272ff. Honest annotation: the present paper's reference to Jensen's argument rests on BDB secondary citation; the 1889 original German article has not been directly read. Early ZA volumes may be accessible through Archive.org or DigiZeitschriften.

Köhler, Ludwig, and Walter Baumgartner. 1994–2000. The Hebrew and Aramaic Lexicon of the Old Testament [HALOT]. Translated by M. E. J. Richardson. 5 vols. Leiden: Brill. [Via secondary citation: the present paper's reference to HALOT's etymological judgment on tevah (favoring Egyptian loanword origin) rests on Google Deep Research synthesis; HALOT original (Brill subscription database, requires university library access) has not been directly accessed]

Landmann, Gunter, Andreas Reimer, and Stephan Kempe. 1996. "Climatically Induced Lake Level Changes at Lake Van, Turkey, during the Pleistocene-Holocene Transition." Global Biogeochemical Cycles 10 (4): 797–808. [Via secondary citation through Dickin 2018's discussion of Lake Van sediment records]

Mallowan, M. E. L. 1964. "Noah's Flood Reconsidered." Iraq 26 (2): 62–82. DOI: 10.2307/4199766. [Full text read, Cambridge Core open access]

Martin, Harriet P. 1988. Fara: A Reconstruction of the Ancient Mesopotamian City of Shuruppak. Birmingham: Chris Martin Associates. [Via secondary citation: the present paper's reference to Martin 1988's Fara I/II/III periodization system rests on Hahn et al. 2022 and other secondary sources; the original book has not been directly accessed]

McCann, Jason M. 2013. "'Woven-of-Reeds': Genesis 6:14b as Evidence for the Preservation of the Reed-Hut Urheiligtum in the Biblical Flood Narrative." In Opening Heaven's Floodgates: The Genesis Flood Narrative, Its Context and Reception, edited by Jason M. Silverman, 131–158. Piscataway, NJ: Gorgias Press. [Via secondary citation through Dickin 2018's discussion of tevah's reed-hut interpretation]

Morozova, Galina S. 2005. "A review of Holocene avulsions of the Tigris and Euphrates Rivers and possible effects on the evolution of civilizations in lower Mesopotamia." Geoarchaeology 20 (4): 401–423. [Partial pages read: key passages distinguishing avulsion from annual flood deposition]

Postgate, J. Nicholas. 1992. Early Mesopotamia: Society and Economy at the Dawn of History. London: Routledge. [Via secondary citation: the present paper's reference to Postgate's discussion of Sumerian measurement standardization periods rests on scholarly synthesis; the original book has not been directly accessed]

Qin, Han. 2026a. Myth as Remainder Intuition: A Mythology Application of SAE. Zenodo. DOI: 10.5281/zenodo.18917796. Available in Chinese and English at https://self-as-an-end.net/papers/myth.html.

Qin, Han. 2026b. Methodology VII: Via Negativa. Zenodo. DOI: 10.5281/zenodo.19481304.

Qin, Han. 2026c. Methodology 00: Via Rho. Zenodo. DOI: 10.5281/zenodo.19657439.

Qin, Han. 2026d. Methodology 0: Via Negativa Foundations. Zenodo. DOI: 10.5281/zenodo.19544619.

Raikes, Robert L. 1966. "The Physical Evidence for Noah's Flood." Iraq 28 (1): 52–63. DOI: 10.2307/4199796. [Abstract only read; the author cross-checked Raikes's precise Shuruppak 60 cm / Kish 40 cm comparison through Dickin 2018 and other syntheses]

Rossi, Corinna. 2020. "On Measuring Ancient Egyptian Architecture." Journal of Egyptian Archaeology 106 (1–2): 229–238. DOI: 10.1177/0307513320975782. [Abstract only read; the present paper cites only Rossi's basic argument on the symbolic/functional distinction in ancient architectural measurement]

Ryan, William B. F., and Walter C. Pitman. 1998. Noah's Flood: The New Scientific Discoveries about the Event That Changed History. New York: Simon and Schuster. [Via secondary citation through Dickin 2018's discussion of the Black Sea Flood hypothesis]

Schmidt, Erich F. 1931. "Excavations at Fara, 1931." The Museum Journal (University of Pennsylvania) 22: 193–245. [Partial pages read: passages on the Shuruppak flood deposit (approximately 60 cm) in the University of Pennsylvania Museum's open-access archive]

Seely, Paul H. 2004. "Noah's Flood: Its Date, Extent, and Divine Accommodation." Westminster Theological Journal 66: 291–311. [Via secondary citation through Dickin 2018's critique of the 2900 BCE flood hypothesis]

Steinkeller, Piotr. 2003. "An Ur III Manuscript of the Sumerian King List." In Literatur, Politik und Recht in Mesopotamien: Festschrift für Claus Wilcke, edited by Walther Sallaberger, Konrad Volk, and Annette Zgoll, Orientalia Biblica et Christiana 14, 267–292. Wiesbaden: Harrassowitz. [Via secondary citation through Dickin 2018's discussion of the Ur III version of the Sumerian King List]

Textual sources for Genesis 6:15 in four languages:

The Hebrew Masoretic Text is based on Biblia Hebraica Stuttgartensia (BHS) or standard MT digital editions (for example, sefaria.org, mechon-mamre.org open versions). [Full text read]

The Greek Septuagint (LXX) is based on the Rahlfs-Hanhart critical edition or open digital editions (for example, biblehub.com/interlinear). [Full text read]

The Syriac Peshitta is based on the British and Foreign Bible Society edition or other open digital editions (for example, dukhrana.com). [Full text read]

The Latin Vulgate is based on the Nova Vulgata or Clementine Vulgate open digital editions (for example, vatican.va or thelatinlibrary.com). [Full text read]

Honest declaration on missing references: this paper depends on secondary citation at several points. The following authors and works have not been directly accessed by the present author; readers should assess the evidentiary grade of corresponding citations accordingly:

• Brugsch 1892 original German article (through BDB)
• Jensen 1889 original German article (through BDB)
• Halévy 1888 related discussion (mentioned but not cited)
• HALOT original (through Google Deep Research synthesis)
• Martin 1988 original (through Hahn et al. 2022)
• Postgate 1992 original (through scholarly synthesis)
• Steinkeller 2003 Ur III King List version (through Dickin 2018)
• George 2003 Critical Edition original (through SOAS BAPLAR and George 2000 Penguin edition cross-checking)
• Original primary literature for Lake Van, Sapropel, and Soreq Cave data (all through Dickin 2018)

All core arguments of this paper (Via Negativa filtering, fifteen-negation derivation, numerical reality-status criterion, Area D dimensional matching) rest on primary materials directly read by the author (Hahn et al. 2022; Einwag et al. 2025 page 96; Heinrich & Andrae 1931 pages 6–7; Aqrawi 2001; Morozova 2005; Schmidt 1931; Mallowan 1964; Dickin 2018; Genesis 6:15 four-language texts). Secondary citations serve mainly background support and linguistic annotation; they do not bear core argumentative weight.

---

Paper complete.

This paper is Application 1 of SAE Mythology. Work is based on publicly accessible materials. Open publication at Zenodo. CC BY 4.0.

引论

Tevah 是希伯来语词，在 Genesis 6:15 出现，通常译为"方舟"。文本给出了具体属性：尺寸 300 × 50 × 30 肘，材料 gopher 木，内外涂 pitch，有一扇侧门，分三层，设天窗。这组属性在物理世界中是否有对应的原型，如何识别这个原型，是一个可被方法论处理的问题。

本文属于 SAE 神话学系列的应用篇 1。SAE 神话学的理论奠基由《Myth as Remainder Intuition》（DOI: 10.5281/zenodo.18917796）完成，该论文处理神话的主体功能维度，即神话如何回应 5DD-6DD 层的余项直觉。本文处理神话叙事的另一维度：物理指称维度。两个维度互补，不冲突。同一神话对象（tevah）可以同时承担主体功能（在洪水灭世的叙事中承载"个体边界不可关闭"的余项直觉）和物理指称（可能对应某个真实存在过的大型建造物）。

方法论上，本文使用 Via Negativa（方法论七，DOI: 10.5281/zenodo.19481304）作为主要筛选工具。工作过程中浮现的两个候选方法论工具记录在附录 E，作为 input 提交给 SAE 方法论系列。

本文不承诺找到 tevah 的物理原型。方法论工作的产出是一个筛选结果，不是一个证明。筛选到的候选（Tell Fara 遗址 Area D）是否真的是 tevah 的物理原型，这个判断超出本文方法论的权限，需要未来现场考古工作判定。本文演示方法论本身，并呈现筛选结果供未来工作参考。

---

第一节 · 研究承诺的范围

方法论工作的第一条纪律是明确承诺范围。在进入具体分析之前，本文先声明本文做什么、不做什么。

本文承诺：用 Via Negativa 系统排除不可能的物理指称候选，给出剩余候选的结构描述，演示完整的筛选过程使读者可以独立复核，诚实标注工作中的不确定性和材料缺口。

本文不承诺：证明某一具体候选就是 tevah 的物理原型，证明 Genesis 洪水叙事的某一版本比其他版本更接近"历史真相"，给出关于 tevah 真实存在与否的神学或宗教判断，跨越方法论权限给出需要实地考古支持的结论。

这个承诺分界重要。Via Negativa 的性质是排除。排除性工具的产出是"最难被排除的候选"，不是"真相"。承诺分界不清会让读者误读结果，也会让工具被无理苛责。

本文承诺范围的另一重表达是，本文的工作全部基于网络可达的公开材料完成。作者没有进行实地调研，没有访问未数字化的档案，没有联系考古团队索取未发表数据。这不是方法论的遗憾，是方法论的边界声明。Via Negativa 的一个优点是它对信息完备性要求较低，可以在信息受限条件下给出有意义的产出。本文利用这个特性，同时诚实标注由此带来的所有局限。

---

第二节 · 存在性分层

讨论 tevah 这类对象最容易犯的错误是层次混淆。"方舟是否存在"这个问题在不同层次上有不同含义。回答一个层次的问题不等于回答另一个层次的问题。不做层次区分的讨论会在历史护教、创世论辩护、"方舟搜索"新闻、以及同情宗教的学术研究之间无穷循环，彼此打架而不自知。

Tevah 的存在性可以分为四个层次。

层次一：物理存在性。在前第三千纪的 Mesopotamia 某处，是否真实存在过一个对应 Genesis 6:15 描述的木造或砖造大型结构。这个问题的答案只能由现场考古证据给出。方法论工作无法直接回答这个问题。

层次二：考古物种存在性。在现有的考古记录中，是否存在符合某些属性约束（尺寸、时代、位置、形态）的结构遗存。这个问题可以由方法论工作部分回答——方法论可以筛选候选，但不能判定候选的真伪。

层次三：文本存在性。Tevah 作为文本对象在 Hebrew Masoretic、Greek Septuagint、Syriac Peshitta、Latin Vulgate 等版本中是否保持属性一致。这个问题由文本学工作回答，证据开放可查。

层次四：神话学存在性。Tevah 作为神话叙事对象在文化传承中扮演什么功能，如何回应深层的结构性信号。这个问题由 SAE 神话学奠基论文处理，不在本文范围。

本文的筛选工作严格在层次二进行。给定层次三的文本约束（tevah 的属性组合），在考古记录中筛选符合约束的候选。层次一超出方法论权限，层次四由奠基论文处理。

这个分层对后文的论证结构有直接影响。第三节处理层次三（指称链），第四节处理层次三的一个特殊问题（数字）。第五和第六节是层次二的筛选工作。第七和第八节给出筛选的收敛候选。第九节诚实处理筛选的可能错误。所有工作都在层次二和层次三之间，不触及层次一。

---

第三节 · 指称链反向追溯

Tevah 不是从 Genesis 凭空出现的词。它有跨语言跨时代的传承链。理解 tevah 的指称对象需要追溯这个链。

下游（最晚，公元纪年前后）：Latin Vulgate 用 arca，Syriac Peshitta 用 qebutha，Greek Septuagint 用 κιβωτός。这三个词在各自语言中都对应"箱、柜、容器"类概念。

中游：Hebrew Masoretic 用 tevah。Genesis 文本的定型期跨越较长，通常认为 Priestly 层约在公元前第六到第五世纪定型。Tevah 这个词在整个 Hebrew Bible 中只出现两次，一次在 Noah 故事，一次在 Moses 出生的蒲草篮子故事（Exodus 2:3）。两处都表示一个涂了 pitch 的容器。这个词本身的使用极有限，说明它是一个专门术语，不是通用词。

上游：Mesopotamia 洪水叙事族。Sumerian Flood Story（约公元前 21 世纪 Ur III 及更早）、Atrahasis 史诗（约公元前 17 世纪 Old Babylonian）、Ark Tablet（Finkel 2014 描述的圆形 quffah 设计，约公元前 17 世纪）、Gilgamesh XI（标准 Akkadian 版本约公元前 13 到前 7 世纪定型）。这些叙事共享"灭世洪水 + 一个或少数幸存者 + 某种载具承载度过"的母题。

埃及支线：tevah 的词源。BDB（Brown-Driver-Briggs 希伯来词典）追溯 tevah 到埃及语 ṭbt（盒、箱、柜）。这个词源观点最早由 Brugsch 和 Erman 在 1892 年的 ZDMG 文章中提出，被 BDB 采纳。HALOT（The Hebrew and Aramaic Lexicon of the Old Testament）的现代判断仍倾向于埃及借词，区别于美索不达米亚的 eleppu（船）系统。

这里有一个关键语言学事实。希伯来文本用一个埃及借词（tevah）而不是一个美索不达米亚借词命名 Noah 的载具。而同时代 Mesopotamia 洪水叙事中的载具词是 eleppu 或类似词，词源于美索不达米亚自身。两个词系不同源。

这一事实和 tevah 的几何属性方向一致。Genesis 6:15 的 tevah 是长方形（300 × 50 × 30 肘，比例 10 : 1.67 : 1），Gilgamesh XI 的 eleppu 是方形（各边相等，ešer nindanā 意为"十根"），Ark Tablet 的 quffah 是圆形（直径约 70 米）。三种不同的几何。

语言学（词源不同系）和几何（形状不同）两份独立证据都指向同一结论：tevah 不是从 Gilgamesh eleppu 或 Atrahasis quffah 直接继承的。Tevah 的原型（如果有）需要在希伯来传统自身中寻找，或者在希伯来传统和埃及传统的接触面中寻找，不在 Mesopotamia 船形传统里。

这不等于说洪水叙事的整体框架不来自 Mesopotamia。Mesopotamia 洪水叙事族和希伯来洪水叙事在结构层面（灭世、幸存、重生）高度一致，这个共同母题的来源可以在 Mesopotamia 传统里找到。但具体载具的几何和名称在希伯来版本中是独立发展的，不是继承来的。这个区分对后文的筛选逻辑重要。

关于 tevah 的本体论候选。给定文本约束（长方形、300 × 50 × 30 肘、内外 pitch、有门有窗、分三层），tevah 至少有三种可能的本体论类型。

候选甲：船。传统理解。但现代实验（包括 2014 年 Finkel 团队根据 Ark Tablet 按比例建造的 quffah）表明，这种尺寸的木制船在工程上不可行——现代最大木船（1909 年 Wyoming 号 99 米）比 tevah 小很多且仍因船体过度弯曲而漏水沉没。Tevah 作为"漂浮于水的 300 肘木船"这一候选在工程学层面被否定。

候选乙：固定建筑。长方形矩形平面的大型建造物，可能是神庙、纪念性建筑、或某种功能性大型结构（储粮、祭祀、行政）。这个候选避开了船的工程困难，保留了尺寸和形态。但候选乙需要解释一个物理机制问题：固定建筑在洪水中如何让人幸存？一个可能的机制是建在足够高的夯土台基或早期 ziggurat 形态高台之上的封闭式巨型建筑（储粮舱、避难舱、纪念性建筑均可）。当洪水淹没平原（典型 Mesopotamia 河流泛滥水深米级到十米级），建在高台上的长方形封闭建筑成为水中孤岛。在远观叙事视野中，这个孤岛可能被记忆为"在水中漂浮的方形容器"。这一物理机制可以让候选乙和洪水叙事的核心母题（载具承载幸存者度过洪水）逻辑闭合，但本身是推测性的，需要现场考古验证（特别是台基存在性和高度）。

候选丙：raft + mudhif 复合体（Dickin 2018 假说）。一个木筏基底上搭建芦苇拱形屋顶建筑。Mesopotamia 南部至今还有 Marsh Arab 搭建类似的 mudhif 建筑。Dickin 认为 tevah 的"三层"不是甲板，而是 mudhif 框架的水平芦苇捆。这个候选介于船和建筑之间，既有漂浮物的特征，又有建筑的形态。

三种本体论候选在几何层面都兼容一个约 137 × 23 米的矩形基底。Via Negativa 筛选的是位置，不是本体论类型。后文不强制在三者中选一个。如果筛选到一个匹配的物理候选，该候选可能兼容三种本体论中的某一种或某几种。

指称链反向追溯的最远可靠锚点是 Hebrew Masoretic tevah（文本定型期约公元前第六世纪）。再往上是推论。本文的工作以这个锚点为起点，向物理世界筛选可能的对应物。

第三节附 · 指称漂移下的筛选边界

本文的指称链分析需要明确一条方法论边界：Via Negativa 在指称漂移对象上只能直接工作于漂移终点形态。这条边界很重要，必须显式说明，不能埋在附录里。

具体到 tevah 的情况：本文使用 Genesis 6:15 的属性（300 × 50 × 30 肘、长方形、内外 pitch、有门、有窗、分三层）作为筛选约束，把这些约束应用到物理世界中筛选候选。但本文不把这些属性直接投射到 Ziusudra、Atrahasis、Utnapishtim 或 Sumerian Flood Story 中的早期载具形态上。

这条边界的理由：Hebrew MT tevah 是经过长期指称漂移之后的稳定形态。早期形态（Ziusudra 时代、Atrahasis 时代、Gilgamesh 时代）可能有完全不同的属性。把后期文本属性直接投射回早期形态可能会错误排除真正的早期原型——真正的早期原型可能没有 tevah 文本中后来添加的某些属性，但仍然是同一指称链的真实起点。

因此本文的筛选结果应理解为：对 MT-tevah 这一稳定文本形态的物理指称候选筛选，而不是对整个洪水神话最早原型的筛选。如果 Area D 通过筛选成为候选，它成为的是 MT-tevah 指称链中某个位置的候选锚点，不是"最早洪水神话原型"的已证位置。

这一边界对附录 E.1 的方法论工具有直接对应：附录 E.1 形式化讨论这条边界作为一个跨对象通用工具。本节正文的论述确保读者在进入第五节筛选之前已经清楚本文筛的是什么、不筛的是什么。

---

第四节 · 数字虚实判据

神话叙事中的数字不是一类。不加区分地对待所有数字是常见错误。有些数字是叙事凝结物（比如"四十天四十夜"），有些数字是实证残余（比如"60 厘米沉积层"），两者的来源和处理方式不同。本节给出一个三分判据。

第一类：象征虚指。特征是数字圆整且对应文化象征数。近东和地中海地区常见的象征数包括 7、12、40、70、一年、千岁、十二支派等。这类数字的来源是叙事编造或文化继承，不反映任何物理测量。处理方式是不作为物理证据使用。

示例：Noah 活到 950 岁，洪水持续 40 天 40 夜，Genesis 一章的七日创造，Abraham 离开 Haran 时 75 岁。这些数字都属于第一类。

第二类：测量化石。特征是数字非圆整，通常带小数或零散整数。这类数字的来源是直接物理测量，精度对应测量精度。处理方式是作为物理事件的硬证据使用。

示例：Schmidt 1931 在 Shuruppak 发掘报告给出的洪水沉积层厚度约 60 厘米（Raikes 1966 给出的 Kish 对照数字是 40 厘米或 16 英寸）。Hahn et al. 2022 磁测给出的 Area D 线性结构长度 110 米。Einwag et al. 2025 给出的 Area D 西侧矩形平面 138 × 24 米。这些数字都属于第二类。

第三类：设计化石候选。特征是数字圆整但非象征，组合呈现工程比例。这类数字的可能来源是建造者的设计决策，记录的是某次真实建造的设计意图。处理方式是作为建造事件的强诊断信号，但要承认仍需考古验证才能从"候选"提升为"已证设计记录"。圆整且非象征且呈工程比例的数字也可能来自后期理想化、规范化、礼制化或文本化重构，需要其他证据排除这些可能后才能确认。

这一类是本判据相对于通常的"圆整 vs 非圆整"二分判据的关键扩展。圆整数不自动等于虚指。圆整可以是象征性的（第一类），也可以是功能性的（第三类）。区分两者的标准有三条：一，数字是否对应已知的文化象征数；二，数字组合是否呈现工程比例（非 1:1:1、非 2:1 这种简单象征比例）；三，数字是否处于某个已知的标准化度量单位下。

示例：Giza 大金字塔基底约 440 埃及王室肘（440 不是象征数，基底面积反映几何稳定性要求）。Etemenanki ziggurat 各层尺寸（圆整但对应工程堆叠要求）。Tevah 的 300 × 50 × 30 肘（300 不是象征数，比例 10 : 1.67 : 1 是工程比例）。这些数字都满足第三类候选判定，作为强诊断信号可进入分析，但是否最终确认为"已证设计化石"需要其他证据支持。

三类判据的应用顺序：先看圆整性，再看象征性，再看比例结构。非圆整直接归入第二类。圆整再检查是否对应文化象征数，对应就归第一类，不对应看比例结构，工程比例就归第三类候选。

度量单位的时间下限。这是一个伴随设计化石判据的硬约束。圆整数只能在标准化度量单位存在时被产生。前 2700 年之前，近东没有标准化肘单位。苏美尔肘约 49.6 厘米（早王朝 II 期开始标准化），埃及王室肘约 52.5 厘米（第三王朝开始标准化）。前 2700 年之前，不同村庄、不同工匠的"肘"长度不一致，产生不了跨代稳定的"300 肘"这种精确数字。

这条约束对 tevah 数字的来源有决定性影响。Tevah 的"300 × 50 × 30 肘"这组数字不可能来自前 2700 年之前的物理建造记忆。任何把源事件置于前第六或第五千纪的假说（包括 Dickin 2018）都必须解释这组具体数字如何在没有标准化测量的时代被产生并跨越数千年保留下来。这一解释在叙事学上难以成立——口传可以保留结构性母题几千年（无疑问），但难以保留具体数字几千年（超出一般口传保留能力）。

匹配度检查。Tevah 300 × 50 × 30 肘按希伯来肘 45.7 厘米换算：长 137.16 米，宽 22.86 米，高 13.72 米。这组米数和第七节要讨论的 Area D 西侧矩形平面（138 × 24 米）在长宽两个维度上接近。这个观察是第三类判据（设计化石候选 = 137 × 23 米）和第二类判据（测量化石 = 138 × 24 米）在同一物件上的可能重合，但是否真的"同一物件"还有待验证。

这个匹配在度量学层面值得进一步分析，因为它涉及一个有趣的反转。如果 Area D 真是 Shuruppak 早王朝时期建造（约前 2700 年），建造者使用的标准化单位应该是苏美尔肘（约 49.6 厘米），不是希伯来肘。按苏美尔肘计算，300 苏美尔肘应该是 148.8 米，不是 138 米。这看起来对 Area D = tevah 物理候选的假说不利。

但反过来思考：如果叙事传承保留的不是"300"这个数字本身，而是建筑的绝对物理长度（约 138 米），那么希伯来文本作者在用自己时代的希伯来肘记录这个长度时，做的换算是 138 ÷ 0.457 ≈ 301.9，自然圆整化为 300 肘。在这种解读下，"300 希伯来肘"不是建造者的设计数字，而是记录者用自己单位重新表达祖传绝对长度的结果。原始建造数字（约 280 苏美尔肘？）丢失了，传承下来的是物理长度。这种跨文化度量衡转换在叙事传承中并非不可能，但本身是一个推测性的解释机制，需要其他证据支持。

需要诚实标注：上述"绝对长度保留 + 跨单位转换"的解读是一种可能机制，不是已证事实。读者可以采纳它，也可以拒绝它。如果拒绝，"300 希伯来肘 = 137 米 ≈ Area D 138 米"的匹配仍然成立但需要其他解释（也许是巧合、也许是希伯来传统使用了某个变体肘单位、也许 Area D 实际建造时代不在早王朝）。匹配本身是可观察的事实，匹配的解释机制是开放的。

关于这个判据扩展的详细说明和跨领域应用示例，见附录 E.2。

---

第五节 · 十五非的系统推导

Via Negativa 不是任意的排除。它是对每一种可能的候选类型进行逐项检验。本节给出系统推导。为节省篇幅，正文把 15 条合并为 10 条叙述，完整 15 条见附录 A。

非海。Tevah 叙事场景是河流平原洪水，不是海洋灾难。Noah 叙事中洪水"上涨"、"退落"，符合河流泛滥后退的特征，不符合海啸或海进的特征。这一非排除 Minoan Thera 火山假说（海啸驱动）、Black Sea Flood 假说（Bosporus 海进，Ryan-Pitman 1998）、以及其他海洋类候选。

非峰。Tevah 叙事结尾船停在 Ararat "mountains"（山群），这是叙事尾声的登陆点，不是建造地。建造地必须能发生灭世洪水，必须能容纳大型建造物的生产，必须临近可航行水道。高山不满足这些条件。这一非排除所有把"Ararat 山"作为物理原型位置的搜索活动。

非平。源事件发生地必须是能让叙事视角认为"这是整个已知世界"的地理——某种半封闭低地盆地。完全开放的平原或高原让"灭世"难以在叙事逻辑上成立。Mesopotamia 冲积平原符合这个条件（低地、周围有山脉、河流泛滥可覆盖大面积）。

非远。必须在叙事传承链的地理范围内（Mesopotamia 到 Levant）。这一非排除印度河文明、黄河流域、美洲各地的同类传说作为 tevah 原型。虽然这些文明也有独立的洪水叙事，但它们不在希伯来传统的传承链上。

非点。Tevah 是大型建造物（137 米长）。需要一片足够大的区域容纳这样的结构。单一小遗址（面积几百平方米）不足以容纳。候选必须是有相当规模的考古位点。

非死。候选区域在叙事源事件时代（前第三千纪）必须有人类活动，有物质文化遗存。完全无人区无法产生"这是整个已知世界"的叙事视角。这一非排除沙漠腹地和完全未定居的荒原。

非舟（弱化）。上文第三节已说明，tevah 作为"真正漂浮的 300 肘木船"在工程学层面不可行。本文采纳三种本体论候选（船/建筑/raft+mudhif）。这一非不是排除船这个本体论，而是弱化对"漂浮可行性"的约束。候选的物理形态不必是一艘可下水的船。

非漂。Tevah 作为原型对象必须有某个固定锚点。纯粹漂流物不留下可查考的地理痕迹，方法论筛选无法处理。候选必须是某个固定位置的结构，哪怕这个结构原本是可漂浮的（如 raft）。

非低海拔。源事件地必须高于 Holocene 海进浸没线。约前 6000 年至前 5000 年海进使波斯湾延伸到今伊拉克中部。海进线以南的古代遗址今日在海相沉积之下，不可直接筛选。候选必须在海进线以北。

非暴露。候选必须不是早已被考古完全揭露并归类的经典遗址核心区。Ur 主丘、Uruk 主丘、Nippur 主丘这些位置已经被发掘一百多年，如果 tevah 原型就在这些位置，相关讨论早已在学术文献中形成定论。真正的候选（如果有）应该在某个尚未完全揭露的位置，或者一个已经被发掘但当时没有被正确识别功能的结构。

非水源远。候选必须邻近古代 Euphrates 或 Tigris 主河道，能解释洪水叙事的物理来源。远离主河道的位置不受大规模洪水影响，不适合作为灭世洪水的发生地。

非单向退路。候选区域必须有洪水退落后的地理痕迹——沉积层、avulsion（河道改道）证据、或其他洪水事件的物质记录。这一非是正向检验：候选区域必须有实际洪水事件的地质证据。

非无参照。候选必须在某个有文本记忆的古城圈里，让口传叙事可以锚定具体名字。前第三千纪 Mesopotamia 有多个文本记忆载体（Shuruppak、Kish、Nippur、Ur 等）。候选位置应在这些名字之一的地理辐射范围内，使叙事有具体锚点（如 Ziusudra 被称为"Shuruppak 之人"）。

非居住带。候选结构必须不是普通居住建筑。Tevah 的尺寸（137 米长）远超普通住宅。结构必须是某种异常建筑——神庙、纪念性建筑、大型公共建筑、或不明功能的大型结构。

非地下。候选必须是地表可见或浅层可及的结构，不是深埋到完全不可观察的深度。方法论筛选和未来考古验证都需要该结构可被探测到。2018 年后 Mesopotamia 磁测技术能探测到约 3 到 5 米深的烤砖结构。这个深度约束就是本非的操作化形式。

合取结果。候选必须同时满足所有 15 条的反面：河流平原低地位置，在 Mesopotamia 到 Levant 范围内，尺寸足够大（百米级），邻近古河道，海进线以北，有洪水事件地质证据，在文本记忆古城辐射圈内，不是普通住宅，不在已知经典遗址核心区，地表可见或浅层可及，非高峰非海滨非孤立点非无人区非漂流物。

15 非合取大幅压缩候选集合。经过前 5 非的地理筛选，候选从全球缩到 Mesopotamia 平原。经过中间 5 非的功能筛选，候选从整个平原缩到 Shuruppak/Kish/Nippur 这一核心古城圈。经过后 5 非的精细筛选，候选从整个核心圈缩到"古城圈中尚未完全揭露的异常大型结构"。在作者掌握的公开材料范围内，符合全部条件的最难排除候选是 Tell Fara 遗址西南 Mound E 的 Area D 结构。

每一非的反向证据和具体排除逻辑见附录 A。

---

第六节 · 地质约束

Mesopotamia Holocene 地质状况决定哪些古代结构今日可见、哪些被后世沉积覆盖、哪些被侵蚀或淹没。地质约束是硬约束，不可跨越。本节处理第五节 15 非中涉及地质的几条约束的具体证据。

海进线。约前 6000 年开始的 Holocene 海进使波斯湾最大延伸约达今伊拉克中部。Aqrawi 1995（DOI 10.1016/0264-8172(95)96903-4）和 Aqrawi 2001 给出 Mesopotamia 平原 Holocene 演化的详细地层数据。海进最大范围的北界大致在今 Nasiriyah-Amarah 一线（Aqrawi 2001 描述），此线以南的前第六千纪位置今日深埋于海相沉积之下。Tell Fara 在此线以北约 100 公里，不受海进影响。

沉积速率。Aqrawi 1995 给出 Mesopotamia 平原 Holocene 平均沉积速率约 1.3 到 2.2 毫米每年。五千年累积约 6.5 到 11 米沉积。这个量级意味着前第三千纪的地表结构今天可能在 2 到 5 米深处。Hahn et al. 2022 的磁力计能探测到这个深度的烤砖异常，所以 Area D 这样的结构对磁测可见。

河道改道。Morozova 2005 详细记录 Euphrates 主河道 Holocene 多次改道（avulsion）。改道与年度洪泛沉积（年度几毫米到厘米级）不同——改道是突发性的大规模事件（河道横移几米到几公里）。Cooke 1987 重建的古 Euphrates 河道通过 Shuruppak 附近。Area D 所在的 Mound E 西南侧紧邻一条古河道痕迹，符合"临水"约束。

洪水层。Schmidt 1931 在 Shuruppak 主丘发掘报告中记录了一层约 60 厘米厚的洪水沉积层，位于 Early Dynastic 地层底部。Mallowan 1964 和 Raikes 1966 在各自的 Mesopotamia 洪水综述中确认这个数字。Raikes 给出精确对照：Kish 约 40 厘米（16 英寸），Shuruppak 约 60 厘米。这层沉积的时代约前 2900 年，是源事件候选时代的直接物质证据。

构造沉降。Mesopotamia 平原新构造下沉是一个长期讨论的问题。地质学界一般认为平原自上中新世以来有累计达数百米到数千米的长期沉降。但对 Holocene 尺度（最近一万年）的具体沉降速率，公开文献中没有找到精确的区域性数字。本文不使用具体构造沉降数字，只承认构造沉降作为定性存在的因素。

综合地质约束，Tell Fara（古 Shuruppak）位置满足：海进线以北（候选可见）、沉积速率在可探测深度（磁测可及）、邻近古河道（有水源和洪水通道）、有直接洪水层证据（Schmidt 1931 的 60 厘米）、有文本记忆（Ziusudra = "Shuruppak 之人"）。这些条件的合取把候选集合进一步压缩。

---

第七节 · 筛选收敛到 Area D

综合第五节的 15 非和第六节的地质约束，筛选过程可以总结为一个渐进收敛的漏斗。

第一层收敛：从全球范围到 Mesopotamia 到 Levant 范围。非海、非峰、非远、非低海拔这几条直接排除非 Mesopotamia 的候选和 Mesopotamia 南部海进区。剩余候选集合：Mesopotamia 平原北半部到 Levant。

第二层收敛：从整个 Mesopotamia 到核心古城圈。非无参照、非水源远、非单向退路这几条要求候选在古城记忆范围内、邻近古河道、有洪水地质证据。符合这些条件的位置集中在 Shuruppak、Kish、Nippur、Eridu、Uruk、Ur 这一前第三千纪古城圈。

第三层收敛：从整个核心古城圈到 Shuruppak。综合文本证据（Sumerian King List 把 Ziusudra 锚定在 Shuruppak，最后一位洪水前王朝）、洪水层证据（Schmidt 1931 的 60 厘米层在 Shuruppak 主丘底部）、以及 tevah 词源（埃及借词，暗示希伯来传统有独立于 Mesopotamia 船传统的物体来源——但仍然在 Mesopotamia 洪水叙事的整体传承范围内），Shuruppak 是最一致的候选城。

第四层收敛：从 Shuruppak 主丘到 Mound E。非暴露要求候选不在已经完全研究的经典遗址核心区。Shuruppak 主丘自 1902 年 Andrae 发掘以来已经多次研究，2022 年起 LMU Munich 团队重新进行磁力测量和钻探。主丘内容基本归类。异常结构在主丘周边——Shuruppak 西南 Mound E 是主丘之外的独立小丘，1902 年简单勘察，此后一直没有深入发掘，到 2022 年 Hahn et al. 的磁测才重新发现其内部结构复杂性。

第五层收敛：Mound E 中的 Area D。2022 磁测在 Mound E 发现两个异常：一条 110 米长的烤砖线性结构，和一块约 20 到 26 米宽的水域或低电阻区，带一个 T 形头部。2025 年 Einwag、Otto、Fassbinder 的后续分析把结构重新描述为：线性部分 115 × 15 米，西侧伴生一个矩形平面 138 × 24 米。Area D 指的就是这个西侧 138 × 24 米矩形平面。

筛选结果的姿态。这个筛选过程的产出是 Via Negativa 在可及材料范围内给出的最难被排除的候选之一。这个表述比"我们找到了"或"这是方舟"保守得多，也准确得多。

需要进一步明确 Area D 作为候选的硬度等级。"物理指称候选"这个词太宽，需要拆分为三层：

层一：完整物理原型候选。所有文本属性（长 300 肘、宽 50 肘、高 30 肘、gopher 木材料、内外 pitch 涂层、侧门、天窗、内分三层、承载洪水中所有生命）都能与某物理对象对应。这是最强的候选地位。

层二：尺寸-形态锚点候选。文本属性中的几何特征和位置-时代窗口（长宽、矩形平面、所在地理范围、可能时代区间、结构类型）能与某物理对象对应。但材料、完整功能、三维高度、内部细节未证。这是中等强度的候选地位。

层三：叙事指称候选。某物理对象的部分特征被后期文本传统重构为叙事对象。物理对象本身可能与叙事描述的多数细节不直接对应，只在某个核心特征（比如"在某地有一个不寻常的大型异常结构"）上被叙事记忆调用。这是最弱的候选地位。

按当前公开材料和本文筛选结果，Area D 最多是层二（尺寸-形态锚点候选），不是层一（完整物理原型候选）。具体而言，Area D 可能解释 tevah 文本属性中的以下一组特征：

• 长 137 米（按希伯来肘换算与 Area D 实测 138 米接近）
• 宽 23 米（与 Area D 实测 24 米接近）
• 长方形平面（与 Area D 矩形结构相符）
• 地理位置在 Mesopotamia 古城圈内（与 Shuruppak 文本-地理记忆一致）
• 可能时代窗口在前 2700 至前 2350 年之间（与 Andrae 1903 归类的 Fara Z 时期一致）
• 结构类型为大型异常建造物（与 Andrae 的 "勉强可算" 评价和 Einwag 2025 的 "港口设施假设" 一致）

但 Area D 不能解释以下 tevah 文本属性：

• 高 30 肘（Area D 是地表结构，没有可观察的"高度"维度直接对应）
• gopher 木材料（Area D 是烤砖结构，不是木结构）
• 内分三层（Area D 没有内部结构发掘，无从知道是否分层）
• 侧门、天窗（同上）
• 涂 pitch（同上）
• 承载所有生命的功能（这是叙事功能，物理对象只能间接支持）

因此 Area D 的候选地位严格定位为层二（尺寸-形态锚点候选）。要把它从层二提升到层一，需要现场考古（钻探采样、碳十四定年、内部结构发掘、材料分析、功能归属）。这些工作超出本文方法论筛选的权限。

Area D 被保留不等于 Area D 就是 tevah 的物理原型。Via Negativa 的性质是排除。剩余候选的意思是"在 15 条约束下还没有被排除的结构"。

反过来，Area D 作为层二候选被保留也是一个有意义的筛选结果。它缩小了未来考古工作的目标范围。如果未来需要检验"tevah 物理指称在 Mesopotamia 范围内最可能的位置是什么"这个问题，Area D 应该是首选钻探目标之一。

---

第八节 · Area D 实证数据

本节呈现 Area D 的全部已知公开实证数据。这些数据跨越 1903 年到 2025 年的 120 年观察历史，三代研究者用各自时代最好的工具记录同一片结构。

Andrae 1903 年的原始发现。Walter Andrae 是 1902 到 1903 年德国东方学会 Fara 发掘的负责人之一。他在 Heinrich 和 Andrae 合编的 1931 年发掘报告中描述了 Mound E 发现的一堵厚墙。以下是原文（书页 6，直接从 Stony Brook AMAR 开放扫描版读取，档案号 585，入口 https://commons.library.stonybrook.edu/amar/585/）：

> Hinsichtlich der Flächenausdehnung der Siedelung von Fara beschäftigte uns die Frage nach einer Befestigung. Da ist nur im Süden des Hügels ein kurzes Stück einer an beiden Enden abbrechenden dicken Mauer gefunden, die den Anspruch erheben könnte, Festungsmauer genannt zu werden. Ihre Lage und Richtung sind so, daß sie zur Not eine Südwestfront geschützt haben könnte. Da es Mauerwerk und nicht bloß Erdschüttung ist, möchte ich den Bau in die letzte Epoche, also in Fara Z setzen und die Mauer mit den Häusern zusammenbringen. Keiner unserer anderen Suchgräben hat die Umwallung sonst geschnitten, wir wissen also herzlich wenig davon...

中文译文：关于 Fara 聚落的面积范围，我们当时正在考虑城防问题。仅在小丘南面发现了一段两端断开的厚墙，它勉强可以被称为城墙。它的位置和朝向表明它或许保护过一个西南前面。由于它是砌墙而不仅仅是土堆，我倾向于把这个建筑归到最晚期，即 Fara Z 时期，并把这堵墙和房屋建筑联系起来。我们的其他探沟都没有切到城墙，所以我们对它知之甚少……

Andrae 的语气值得注意。他三处用虚拟语气：könnte（可能）、möchte ich（我倾向于）、zur Not（勉强）。这表明他 1903 年面对这堵墙时就不确定它是什么。他勉强把它归类为城墙，勉强归到 Fara Z 时期（Andrae 自己的分期系统中最晚的一层，约公元前 2700 年，基于文字形态学判定），承认探沟可能挖得不够远。

Andrae 的分期系统是：Fara H（Hüttenschichtung，茅屋层，最早）、Rundkeller（中间圆形地下室）、Fara Z（Ziegel，砖房层，最晚）。这个系统是 Andrae 1903 年基于建筑材料观察的划分，不是现代考古学常用的 Early Dynastic I/II/III 系统。Fara Z 对应的绝对年代，Andrae 根据房屋内发现的楔形文字泥板形态估计约在公元前 2700 年。

在接下来的书页 7，Andrae 继续讨论这堵墙：

> Der Ruinenzustand an der Peripherie macht es wahrscheinlich, daß die Umwallung gänzlich vernichtet ist. In den älteren Schichtungen sind ebenfalls keine Spuren eines Erdwalls festgestellt, die Siedelung war damals möglicherweise offen und nur durch ihre Lage im Wasser geschützt. Siedelungen zu umwallen ist eine besondere Erfindung, die bekanntlich in Uruk dem Gilgamesch zugeschrieben wird.

中文译文：外围的废墟状态使城墙很可能已经完全毁坏。在更早的地层中也没有发现土墙的痕迹，当时的聚落可能是开放的，仅靠其水中位置得到保护。城墙建造是一项特殊发明，众所周知，在 Uruk 这被归于 Gilgamesh。

这里 Andrae 提到 Gilgamesh 建 Uruk 城墙的传说作为类比。他用神话叙事解释这堵墙，但没有把它和 Ziusudra 或洪水叙事联系起来。这是他的方法论盲点之一——他用了同时代相邻的神话对象（Gilgamesh）作为类比，但没有想到同城的更早神话对象（Ziusudra）。这个盲点在本文第十节再讨论。

Hahn et al. 2022 磁测数据。Hahn、Fassbinder、Otto、Einwag、Al-Hussainy 2022 年发表于 Archaeological Prospection（DOI: 10.1002/arp.1878，开放获取）的文章报告 LMU Munich 团队 2022 年在 Shuruppak 西南 Mound E 的磁力测量结果。主要发现：

• 一条约 110 米长的烤砖（fired brick）线性结构
• 一块约 20 到 26 米宽的水域或低电阻区（解释为水道或壕沟）
• 一个 T 形头部结构（解释可能为水闸或扩展结构）
• 位置坐标 31.777222°N, 45.510833°E

这次磁测把 Andrae 1903 年看到的 71 米可追墙的真实范围扩展到 110 米，并揭示了 Andrae 挖不到的伴生水域和头部结构。

Einwag, Otto, Fassbinder 2025 进一步分析。Einwag、Otto、Fassbinder 2025 年在 Romano 主编的论文集 Sumer and the Sea: Deltas, Shoreline, and Urban Water Management in 3rd Millennium Mesopotamia（ARATTA 3, Brepols，DOI: 10.1484/M.ARATTA-EB.5.143624）的"Harbours, Water Gates, and Canals in Šuruppak and Ur"一章（pp. 91-100）中更新描述。该章第 96 页的精确措辞：

线性结构 115 × 15 米

西侧矩形平面 138 × 24 米

2025 年版描述把 2022 年的 110 米线性结构精确化为 115 × 15 米，并识别出西侧伴生一个矩形平面 138 × 24 米。这个矩形平面即本文所称的 Area D。该章把结构假设为港口设施（harbour installation），但作者保留了对功能归属的不确定性。

Tevah 文本尺寸的换算。Genesis 6:15 的 tevah 尺寸是 300 × 50 × 30 肘。按希伯来肘约 45.7 厘米换算：

长 = 300 × 0.457 = 137.16 米

宽 = 50 × 0.457 = 22.86 米

高 = 30 × 0.457 = 13.72 米

匹配度。

Area D 西侧矩形平面：138 × 24 米

Tevah 换算（按希伯来肘）：137.16 × 22.86 米

长度差 0.84 米（误差 0.6%）

宽度差 1.14 米（误差 5%）

需要明确的是，这是二维长宽匹配，不是完整三维匹配。Area D 是磁测可见的地表结构，没有可观察的高度维度直接对应 tevah 文本中的 30 肘高度。Tevah 文本中提到的内部三层、侧门、天窗、gopher 木材料、内外 pitch 涂层等属性，在 Area D 这一磁测结构中目前无对应物，需要现场考古发掘才能检验。

在合理的肘单位选择和磁测边界判定误差范围内，Area D 与 tevah 文本尺寸在长宽两个平面维度上形成高度醒目的匹配。两份数据来自两种完全独立的体系：Area D 的 138 × 24 米来自 2022-2025 年磁力测量（第二类测量化石），tevah 的 137 × 23 米来自文本 300 × 50 肘按希伯来肘换算（第三类设计化石候选）。

关于"同一物件"的解读约束。如果 Area D 和 tevah 文本属于同一指称链——也就是说，如果 Area D 这个物理对象的设计或测量数据，在前第三千纪到前第六世纪之间，通过某种叙事传承机制，进入了 Hebrew tevah 文本——那么一米级的差异可以被理解为设计数字与实物测量之间的正常偏差（风化侵蚀、建造误差、测量边缘判定）。但这一"同一指称链"假说本身需要其他证据支持，本文不直接断言。如果未来现场考古证明 Area D 的建造时代远晚于 Shuruppak 早王朝时期（例如属于 Old Babylonian 港口设施），那么尺寸匹配就需要被重新解释为巧合或某种共享的标准化设计模式（138 × 24 米是大型矩形结构的工程优化值之一）。

关于尺寸匹配在论证中的位置。需要强调：尺寸匹配不是本文筛选的起点，也不是单独成立的结论。Area D 是在第五节 15 非合取筛选 + 第六节地质约束 + 第七节五层收敛之后仍未被排除的候选。长宽匹配只是其中最醒目的后验支持。如果有读者把全文读成"先看到一个 138 × 24 米的结构，然后套 Genesis 尺寸"，那是对论证流程的误读。本文的论证是先做完独立于尺寸的结构性筛选，再观察到尺寸匹配作为额外印证。

度量学层面的反转观察。第四节末已经讨论：如果 Area D 真是早王朝时期建造，原始建造单位应是苏美尔肘（约 49.6 厘米），300 苏美尔肘应是约 149 米，与 Area D 实测 138 米差近 11 米。希伯来肘 45.7 厘米和 138 米的精确匹配，可能意味着叙事传承保留的不是数字"300"，而是物理长度（约 138 米），希伯来记录者用自己的肘单位重新换算时得到 301.9 ≈ 300 肘。如果这一"绝对长度保留 + 跨单位转换"的解读成立，匹配的解释会更精彩——文本数字不是凭空编造，是经过一次跨文化度量衡转换的精确传承结果。但这一解读是推测性的，不是已证机制。读者可以采纳，也可以拒绝。匹配本身是事实，匹配的解释机制是开放的。

时代归属讨论。Andrae 1903 把墙归到 Fara Z（约前 2700 年）。Hahn et al. 2022 和 Einwag et al. 2025 没有最终定年。现代磁测揭示的结构比 Andrae 看到的更大更复杂，但磁测本身不能直接给出年代。精确定年需要现场钻探和碳十四测定。可能的归属区间是 Early Dynastic II 到 III（约公元前 2700 到前 2350 年），覆盖 Shuruppak 鼎盛期，也覆盖 Schmidt 1931 报告的洪水层之后的重建期。

本文不强行主张一个具体的建造时代，只指出：如果建造时代在这个区间，就正好落在以下三个条件同时满足的窗口：

一，苏美尔肘单位已经标准化（支持 300 肘设计化石的可能产生）

二，Shuruppak 作为政治中心处于活跃期（支持大型建造活动）

三，Schmidt 1931 的 60 厘米洪水层之后（可能触发纪念性建造）

以上三条都是条件陈述，不是定论。Area D 的精确建造时代是未来考古工作的关键待定项。

---

第九节 · 假阳性假阴性分析

诚实讨论筛选可能的错误是方法论工作的必要部分。本节处理两类可能错误。

假阳性：Area D 被错误保留，实际不是 tevah 原型。

可能情境一：Area D 是完全无关的晚期建筑（如 Ur III 时期港口设施或 Old Babylonian 时期防御工事），尺寸匹配 tevah 纯属巧合。这个情境在概率上不能排除。138 × 24 米是一个常见的大型建筑尺寸，tevah 的 300 × 50 肘也可能是某种标准化设计比例，两者匹配可能是"两个独立的标准化设计决策偶然相遇"。反驳这个情境需要 Area D 的精确定年——如果建造时代在前 2700 到前 2350 年之间，匹配就更难解释为纯巧合。

可能情境二：138 × 24 米的测量本身有边缘判定误差。磁测结果依赖对磁异常边界的人工判定。Einwag et al. 2025 给出的尺寸是一个最佳估计，实际尺寸可能在 135 到 141 米长、22 到 26 米宽的范围内。这个误差本身不大，不足以颠覆"一米级匹配"的结论，但需要诚实承认。

可能情境三：希伯来肘 45.7 厘米的换算值是现代重建，前第一千纪希伯来实际使用的肘单位可能略有不同。古代肘单位有多个变体（普通肘、王室肘、圣肘），不同时代不同地区的希伯来人使用的肘长度可能略有差异。如果实际使用的肘单位是 44 厘米，换算结果是 132 米；如果是 47 厘米，换算结果是 141 米。这个不确定性使"精确匹配"的说法需要限定为"在合理的肘单位范围内匹配"。

假阴性：真正的 tevah 原型被错误排除，或者本不存在但筛选假装找到了候选。

可能情境一：真正的原型可能位于我们筛选时排除的区域。比如某个未发掘的 Kish 周边小丘，或 Nippur 外围某个位置，或 Eridu 周围。本文的筛选基于公开考古数据，对未发掘区域无能为力。

可能情境二：真正的原型可能早已被后世建筑完全覆盖或侵蚀。Mesopotamia 5000 年的人类活动覆盖率很高，许多前第三千纪结构被后世建筑叠加到完全看不见。

可能情境三：tevah 根本没有单一物理原型，它是多个真实建筑（或多次建造事件）在叙事传承中的合成物。这个情境在神话学上常见——神话对象往往是历史事件的凝结而不是记录。如果真是这样，任何"找到单一物理原型"的尝试都是方法论范畴错误，Area D 最多只是合成物中的一个片段。

可能情境四：Tevah 完全是纯叙事构造，没有物理原型。这是极端假设但不能先验排除。如果真是这样，Area D 的尺寸匹配就是彻底的巧合。本文的筛选和论证建立在"有物理原型"这个工作假设上，这个假设本身是可被否定的。

对 Dickin 2018 假说的回应。Alan Dickin 在 Perspectives on Science and Christian Faith 2018 年的文章中主张 Noah 洪水的真实时代是公元前 5700 年，而不是传统的前 2900 年。他的地质学论证（Soreq Cave 同位素、Lake Van 沉积、Sapropel 1 层）支持前 6500 到前 5000 年是 Holocene 最湿润期，这个时期 Mesopotamia 确实可能发生过前所未有的大洪水。他的考古学论证（南 Mesopotamia 前 5500 年之前无定居遗址）暗示前 5700 年的洪水可能抹去了此前所有定居。他提出 tevah 作为"raft + mudhif"的本体论候选。

本文对 Dickin 的回应分两层。

第一层，物理事件方面。Dickin 的地质学证据是有效的。前 5700 年 Mesopotamia 可能确实发生过大洪水。这不和本文的论证冲突。本文处理的是 tevah 这一神话叙事对象的物理指称问题，不是"洪水物理事件何时真正发生"的问题。两个问题可以独立处理。

第二层，叙事传承方面。Dickin 的假说有一个困难。如果源事件在前 5700 年，而具体叙事细节（Ziusudra 名字、Shuruppak 地名、300 × 50 × 30 肘尺寸）在文本中出现最早不晚于前第二千纪，那么这些具体细节必须经过 3000 年以上的口传才能从源事件到达文本记录。3000 年口传可以保留结构性母题（灭世、幸存、重生），但难以保留具体数字、地名、人名。度量单位约束（第四节）进一步加强这一困难：前 5700 年没有标准化肘单位，"300 肘"这组数字不可能从那个时代的建造记忆直接继承。

两层源头模型。综合以上，本文建议一个两层源头模型。

远层（前第六千纪前后）：真实的 Mesopotamia 大洪水或一系列相关洪水事件。为后世叙事提供"灭世洪水"这一结构性母题。具体细节在口传中大部分丢失。

近层（约前 2900 年或其后）：Shuruppak 洪水层（Schmidt 1931 的 60 厘米）及其后续重建。为已经存在的结构性母题提供具体化的物理锚点——具体的地名（Shuruppak）、具体的王名（Ziusudra）、具体的载具尺寸（300 × 50 × 30 肘，如果 Area D 是洪水后重建的纪念性结构）。具体细节在有文字社会（前 2100 年 Ur III 起可书写记录）前夜形成，更容易保留到文本定型期。

这个两层模型和 Dickin 的单层模型（前 5700 年）在物理事件层面兼容，但在叙事传承层面更一致。本文的 Area D 筛选处理的是近层，不涉及远层。Dickin 的前 5700 年假说处理远层，不需要要求远层也提供具体细节。

埃及词源置换的结构性解释。两层源头模型留下一个待解释的子问题：如果物理原型（近层）在 Mesopotamia 腹地，为什么希伯来文本用埃及借词 tevah（源自埃及 ṭbt，盒、箱）而不是用某个美索不达米亚借词（如从 eleppu 派生的词）来指代它？这在地理和语言学上构成一种张力——物理锚点在东，命名词源在西。

一个可能的结构性解释：希伯来传统（处于近东走廊，长期处于 Mesopotamia 和埃及两大文明接触带）在用语言固化这个源自 Mesopotamia 的"非船型巨型方形建造物"时面临一个选择——Mesopotamia 自己的 eleppu 词义明确是"船"（可漂浮航行的带桨载具），和 Area D 类型结构的形态不符。希伯来记录者如果借用 eleppu，会引入一个形态误导（让读者误以为是船）。希伯来记录者不借用 Mesopotamia 词系，而转向形态正确的埃及词 ṭbt（意为"箱、盒、容器"），可能是一次有意识的语言学矫正——为了维持与物理原型的形态一致性，而主动跨越 Mesopotamia 词系，在埃及词系中找到一个形态匹配的词。

如果这一解释成立，埃及词源就不是 tevah 与 Mesopotamia 传统无关的证据，而恰恰是希伯来传统在选择命名时对物理原型形态精确性的语言学坚持——建筑不是船，所以不能用"船"词。这和第三节的本体论观察（tevah 是长方形建筑/复合体，不是 Mesopotamia 船形传统）方向一致。

必须明确：这一"语言学矫正"的解释是推测性的，不是已证事实。它是一个"恰好能解释"的假说，需要其他语言学和历史学证据支持。替代解释也存在——比如 tevah 的埃及词源纯粹反映希伯来传统在某个历史时期（如 Exodus 前后）的文化接触，和 Mesopotamia 物理原型无关，只是借词的偶然巧合。本文列出"语言学矫正"假说作为一种可能的结构性解释，让读者知道这个张力在两层源头模型中有可处理的路径，但不强主张某一解释就是答案。

---

第十节 · 方法论反思

本次工作对方法论本身有四点启示。

启示一：Via Negativa 对信息不完备条件的稳健性。

本文全部工作基于网络可达的公开材料完成，作者没有实地调研，没有访问未数字化档案，没有联系考古团队索取未发表数据。在信息受限条件下，本文仍然能给出有意义的筛选产出。这不是意外，而是 Via Negativa 的性质使然。

Via Negativa 的核心操作是排除。排除只需要足够多的反例材料，不需要穷尽的肯定材料。要说"候选 X 不在此位置"只需要说明此位置的某个已知属性和候选 X 的某个文本属性不兼容。公开考古数据已经足够支持大部分这样的排除。真正需要完备数据的工作是肯定——要说"候选 Y 就是 tevah"需要证明 Y 的所有属性都匹配，这需要现场钻探、碳十四、内部发掘等等。

Via Negativa 把这两种工作分开处理。排除用公开数据做，在网络和电脑前就能完成。肯定需要现场工作，本文不承担。这个分工让方法论七在"坐在电脑前"的工作模式下仍然有用武之地。

启示二：处理"长期悬而未决的谜题"的适用性。

Area D 不是新发现。Andrae 1903 年就看到那堵墙，坦率承认不确定。1931 年发掘报告公开这一记录。1988 年 Martin 在 Fara: A Reconstruction of the Ancient Mesopotamian City of Shuruppak 重新整理资料，仍然没有定论。2022 年 LMU 团队用磁测重新观察，把结构扩大到 110 米，但功能仍然存疑。2025 年 Einwag、Otto、Fassbinder 再次分析，仍然保留不确定性。

120 多年的三代研究者用当时最好的工具处理同一片结构，每一代都推进了一步，每一代都停在"这是什么还没搞清楚"。我们的方法论七介入不是"解决"谜题。方法论七的贡献是显性化筛选逻辑。

前三代研究者的筛选工作是隐性的。Andrae 说"勉强可算城墙"——这是在隐性地排除其他候选（住宅、神庙、仓库）但没有明说为什么。2022 团队说"可能是港口"——这是在隐性筛选但没有列出 15 条约束。我们的 Via Negativa 把这个筛选过程写出来——15 条约束、每条的排除逻辑、合取的收敛路径。

显性化本身就是贡献。它不保证筛选结果比前人更正确，但它让筛选可检查、可复核、可批评。下一代研究者可以针对我们的 15 非逐条提出异议，而不是只能说"我们总觉得这堵墙是某种不明结构"。

启示三：Via Negativa 在考古-文本学交叉的隐性先例。

Dickin 2018 文章实际上在做一种 Via Negativa 工作。他排除前 2900 年（论证太晚、太小、Uruk World System 已形成），排除 2900 年洪水作为源事件（太弱不能灭世），排除 Black Sea Flood 假说（海进无退潮），最后保留前 5700 年作为最一致的源事件候选。这个工作流在结构上和本文一致，但 Dickin 没有给这个方法论命名。

Mallowan 1964 文章的某些章节也有类似结构——他排除若干候选洪水层作为 Noah 洪水原型（Kish 上层太晚、Ur 下层太早等），保留 Shuruppak 下层（2900 BC 附近）作为最一致的候选。再次是隐性 Via Negativa。

方法论七在考古-文本学交叉的定位不是开创，而是给已有的隐性实践命名并规范它。这个定位重要。它避免了"我们发明了新方法"这种容易招致质疑的姿态，也避免了"我们把别人的工作据为己有"这种不诚实。我们承认这种方法一直存在，我们的贡献是把它从隐性做法变成可教可学的显性工具。

启示四：三类数字判据的跨领域适用性。

本文第四节的三类数字判据（象征虚指/测量化石/设计化石）不只对 tevah 适用。它可以应用到任何古代文本中的尺寸数字。Giza 金字塔的 440 王室肘基底，Etemenanki 的各层尺寸，中国礼制建筑的开间数，希腊神庙的柱距模数，罗马引水道的段落长度——凡是古代文本中出现具体尺寸数字的场景，三类判据都可能是有用的诊断工具。

这个判据的通用性说明它不是专为 tevah 设计的 ad hoc 工具。它是一个可独立使用的方法论工具，其适用范围远超本文的应用场景。附录 E.2 给出详细说明。

---

第十一节 · 在神话学研究中的定位

本节说清楚本文在神话学这个学科中的位置。

和 SAE 神话学奠基论文的关系。

SAE 神话学由《Myth as Remainder Intuition》奠基。该论文处理神话的主体功能维度——神话回应 5DD-6DD 层的结构性余项（个体边界 cannot not 保留的直觉），这一维度不可被科学替代。本文处理神话叙事的另一个维度：物理指称维度——叙事中提到的具体对象是否有物理锚点，如何识别。

两个维度互补不冲突。同一个神话对象可以同时承担两种功能。Tevah 作为主体功能载体，在 Genesis 洪水叙事中承担"人类整体边界被外部力量威胁时如何保存个体"的余项表达。同时，tevah 作为物理指称，可能对应某个真实的前第三千纪 Mesopotamia 大型建筑。两种解读在本体论上不冲突，在解释上互相丰富。

一个神话对象如果只有主体功能而没有物理指称（比如某些纯隐喻的神话人物），那是它的结构选择。如果同时有主体功能和物理指称，那是它的完整结构。tevah 很可能属于后者。本文的工作是把 tevah 的物理指称这一维度从神话学讨论中显性化并加以处理。

和传统比较神话学的关系。

Eliade 在 The Sacred and the Profane 中观察神圣经验不可被世俗完全吸收。Jung 在集体无意识理论中观察跨文明神话结构的共通性。Tylor 在 Primitive Culture 中把神话作为进化阶梯的初级阶段。这些传统各有其贡献，也各有其局限。

本文所在的 SAE 神话学和传统比较神话学的区别：SAE 神话学把物理指称维度显性化为神话研究的独立子领域。传统比较神话学关注母题、原型、结构共通性，这些都是叙事内部的性质。SAE 神话学增加一个叙事外部维度——叙事对象在物理世界中的可能锚点。

这不是对传统比较神话学的替代，是补充。传统工具（母题分析、原型分析、结构分析）在 SAE 神话学应用篇中仍然使用。SAE 神话学增加的是"物理指称"这一操作层。两者合力可以处理更完整的神话对象——既理解它为什么有这样的叙事结构（传统工具），又理解它是否对应物理世界中的某个东西（SAE 增加的维度）。

和考古-文本学交叉研究的关系。

Mallowan、Dickin 等学者的工作是考古学和文本学的交叉。他们处理具体文本（Genesis、Gilgamesh）中提到的物理对象（洪水层、方舟、塔）在考古记录中的可能对应。这是一个已有的研究领域。

本文的贡献不是开辟新领域，而是给这个已有领域提供一套显性化的方法论工具。前人的工作隐性地使用了 Via Negativa、指称链分析、数字虚实判断等工具，但没有把这些工具命名和规范。本文（和其引用的方法论七、方法论 00）把这些工具显性化，让后续研究者可以直接调用它们。

和宗教研究的关系。

本文明确不处理宗教或神学问题。Tevah 是否有物理原型和"Genesis 是否是神启"是两个完全不同的问题。本文只处理前者。后者超出方法论权限，也不在 SAE 神话学的研究范围内。

读者如果是宗教立场的人，本文的筛选结果可能被他们解读为"支持 Genesis 历史真实性"或"反对 Genesis 历史真实性"。这两种解读都是读者自己的延伸，不是本文的主张。本文的结论是"Area D 是当前可及材料下最难被排除的 tevah 物理指称候选"，这个结论和任何宗教立场都不直接冲突，也不直接支持。

和未来 SAE 神话学应用篇的关系。

本文作为 SAE 神话学应用 1，也为后续应用篇提供一个工作范式。后续应用篇（应用 2、3、4 等）可以处理其他神话对象的物理指称（巴别塔、昆仑、夏朝等）、其他神话对象的指称漂移（西王母、盘古、Prometheus 等）、或其他神话族群的结构比较（洪水神话族、造物神话族、下冥界神话族等）。每一篇按对象需要选择合适的方法论工具栈，不强制使用 Via Negativa。本文的贡献之一是建立一个可重复使用的骨架。

---

第十二节 · 结论

本文用 Via Negativa 对神话叙事对象 tevah 的物理指称问题做了一次完整的方法论筛选。筛选从全球范围收敛到 Tell Fara 遗址西南 Mound E 的 Area D 结构。Area D 是作者可及公开材料范围内最难被排除的候选。

做了什么：演示了 Via Negativa 在神话学物理指称维度的完整工作流程，包括指称链反向追溯、数字虚实判据三分类、15 非系统排除、地质约束审查、候选收敛分析、诚实的假阳性假阴性讨论。贡献了两个候选方法论工具（附录 E），作为 input 提交给 SAE 方法论系列。

没做什么：没有证明 Area D 是 tevah 的物理原型。这个判断超出方法论筛选的权限，需要现场考古工作——钻探采样、碳十四定年、内部结构发掘、建造时代确定、功能归属判定。本文没有跨越这个权限界限。

本工作的价值不在"发现"，而在"规范"。Area D 本身早在 1903 年就被 Andrae 看到，他当时的不确定感被本文反复引用。2022 和 2025 的 LMU 团队用磁测更新了我们对它的了解，功能仍然存疑。本文不声称解决了这个 120 年谜题。本文只是把如何思考这个谜题的方法写出来，让思考过程变得可检查、可复核、可批评、可改进。这是方法论工作的典型姿态。

Via Negativa 的承诺是有限的。它承诺给出当前信息下最好的筛选结果，不承诺给出真相。它承诺把筛选过程显性化，不承诺筛选结果不可质疑。它承诺诚实标注所有材料缺口和不确定性，不承诺闭合所有开放问题。本文尽力遵守这些承诺。

如果未来有人携带更多材料（更多钻探数据、更多文本、更多地质证据）重做这个筛选，得出和本文不同的结论，那不是本文的失败，是方法论的成功——工作过程被显性化之后，后来者可以在清楚的基础上继续。

Area D 的真相需要考古学家的工作。方法论的真相需要更多应用来检验。SAE 神话学的真相需要整个系列的持续展开。本文只是其中一个节点。

---

附录 A · 15 非的完整推导

正文第五节合并为十条叙述。本附录给出完整 15 条，每条列前提、排除范围、逻辑关系。

非 1 非海

前提：Genesis 6 到 9 的洪水叙事描述洪水"上涨"（עָלָה）、"回流"（שׁוּב）、"退落"（חָסֵר）。水的运动符合河流泛滥后退模式，不符合海啸（单次冲击即退）或海进（水位上升后不退）模式。

排除范围：Minoan Thera 火山驱动的海啸假说、Ryan-Pitman 1998 的 Black Sea Flood 假说、Persian Gulf 海进假说、其他所有以"海洋灾难"为物理基础的候选。

逻辑关系：独立。这一非是最外层的筛选，先于其他非执行。

非 2 非峰

前提：Genesis 8:4 记载 tevah 最终停在"Ararat mountains"（הָרֵי אֲרָרָט，山群）。这是叙事尾声的登陆点。文本没有暗示 tevah 在高山上建造。建造地和登陆地是两个不同地点。

排除范围：所有把"Ararat 山"、"Judi 山"（穆斯林传统登陆点）、或其他高山作为 tevah 物理原型位置的搜索。这一非不排除登陆叙事本身。

逻辑关系：独立。和非 1 并列，都是对叙事的直接读解约束。

非 3 非平

前提：叙事要求源事件发生地让幸存者"认为这是整个已知世界的毁灭"。完全开放的平原、高原或草原让这种叙事视角难以成立——在开阔地貌中人类视野能看到很远，难以产生"一切尽毁"的封闭感。半封闭低地盆地（周围有山脉或高地围合）能产生这种视角。

排除范围：中亚草原、北美大平原、撒哈拉、蒙古高原、西伯利亚平原、青藏高原等开放地貌。

逻辑关系：和非 1 有弱依赖——既然非 1 排除海洋，那源事件是陆地洪水，陆地洪水需要水能聚集的盆地地形。非 3 是非 1 的地形学推论。

非 4 非远

前提：希伯来叙事传承链的地理范围在近东（Mesopotamia 到 Levant 到埃及）。这是语言学、考古学、文本学多重证据支持的范围。源事件必须在这个传承链的地理可达性内，才能被希伯来传统继承。

排除范围：印度河文明、中国黄河流域、美洲各文明、撒哈拉以南非洲、东南亚、大洋洲的各自独立洪水叙事。这些文明的洪水叙事可能是独立起源或有其他传承链，不是希伯来 tevah 叙事的源头。

逻辑关系：独立。基于传承链的地理学约束。

非 5 非点

前提：Tevah 的文本尺寸（300 × 50 肘 = 约 137 × 23 米）指向一个百米量级的大型结构。候选遗址必须有足够的空间容纳这样的结构，以及附属的工作区、水道、支撑结构。

排除范围：面积小于数公顷的单一小遗址（小于 10000 平方米）。这排除了许多独立村落遗址，但不排除中型城市或大型公共建筑群的部分区域。

逻辑关系：独立。基于尺寸约束的几何学推论。

非 6 非死

前提：源事件时代（前第三千纪及之前）候选区域必须有人类活动迹象。完全无人区既不可能产生"灭世"叙事（无人可灭），也不太可能留下可查考的物质文化遗存。

排除范围：阿拉伯半岛腹地沙漠、Empty Quarter、Syrian Desert 深处等前第三千纪无定居区。

逻辑关系：和非 4 并行。非 4 从传承链约束地理，非 6 从考古证据约束地理。

非 7 非舟（弱化版）

前提：现代工程实验（Finkel 2014 团队按 Ark Tablet 尺寸缩比建造的 quffah 船在 2014 年的实验；更广义的木造船极限研究包括 Wyoming 号 1909 年 99 米木船实际表现）表明，300 肘量级的木造可漂浮船在现实工程上不可行。

排除范围：把 tevah 严格理解为"一艘可下水航行的 137 米木船"的本体论候选。

逻辑关系：弱化非。这一非不完全排除船本体论——它只排除"可实际漂浮航行"的强版本，保留"具有船形态的静态结构"的弱版本。和非 8 协同工作。

非 8 非漂

前提：方法论筛选需要候选有可查考的固定锚点。纯粹漂流物不留下地理痕迹，方法论筛选无法定位。如果 tevah 原型确实是某个漂浮物，它要么最终停留在某个固定位置（可筛选），要么已经彻底消失（不可筛选）。本文只处理前者。

排除范围：纯漂流物候选（没有固定停留点的假想物）。

逻辑关系：和非 7 协同。非 7 排除"可实际航行"的船本体论，非 8 排除"纯漂流无固定锚点"的物体论。两者合起来要求候选是固定位置的结构（可能是建筑、可能是固定木筏、可能是 raft + mudhif 复合体，但不能是纯粹的漂浮物）。

非 9 非低海拔

前提：Holocene 海进（约前 6000 到前 5000 年波斯湾向北延伸到今伊拉克中部）使海进最大线以南的前第六千纪地表今日深埋海相沉积之下。如果源事件发生在海进线以南，物证今日不可直接筛选。

排除范围：Aqrawi 2001 描述的海进最大线（大致今 Nasiriyah-Amarah 一线）以南的前第六千纪位置。

逻辑关系：和非 3 有弱重叠——非 3 要求半封闭盆地地形，非 9 排除盆地南部被海淹没区。Mesopotamia 南部冲积平原北半部（海进线以北）是两者的交集。

非 10 非暴露

前提：如果 tevah 物理原型就在已经被一百多年考古完全发掘、归类、文献化的经典遗址核心区（Ur 主丘、Uruk 主丘、Nippur 主丘、Lagash 主丘、Eridu 主丘等），相关讨论在现有学术文献中早已有定论。既然现有文献中没有这样的定论，原型如果存在应该在尚未完全揭露的位置，或者已经被发掘但当时未被正确识别功能的结构。

排除范围：上述各主丘的核心已发掘区。不排除这些遗址的周边小丘、未发掘区、或主丘内被认定为"不明结构"的部分。

逻辑关系：这一非是方法论上的启发式，不是严格逻辑约束。它依据"缺失证据"推理——如果原型在经典遗址核心区，文献会有记录。这个推理不是严密的（可能原型在核心区但被错误归类），但作为筛选优先级排序是合理的。

非 11 非水源远

前提：大规模洪水事件需要邻近主河道。远离主河道的位置不受大规模河流泛滥影响。文本中洪水的描述（持续、大规模、覆盖叙事视角中的"整个世界"）要求源事件发生在河道密集区。

排除范围：距离古 Euphrates 或古 Tigris 主河道（Cooke 1987 重建）5 公里以上的位置。沙漠绿洲、干涸河床、远河居住点都被排除。

逻辑关系：和非 3、非 9 协同。非 3 要求盆地地形、非 9 要求海进线以北、非 11 要求临古河道。三者合取把候选限制到 Mesopotamia 中部冲积平原的古河道沿线。

非 12 非单向退路

前提：洪水事件必须在地质记录中留下痕迹——沉积层、avulsion（河道改道）、carbon 层（淹没有机物）、或其他物质证据。"纯粹的叙事洪水"（没有任何物理痕迹）没有对应物理锚点可筛选。

排除范围：没有 Holocene 洪水地质证据的区域。

逻辑关系：正向约束。非 12 不排除位置，而是要求候选区域必须有正向地质证据。如果一个区域满足非 1 到非 11 的所有反向但没有非 12 的正向证据（即没有任何 Holocene 洪水层），那它仍然被排除。

非 13 非无参照

前提：长期口传叙事需要具体锚点才能保留细节。没有具体地名、王名、事件名的纯抽象叙事难以跨越数千年保留精确属性。Sumerian King List 把 Ziusudra 锚定在 Shuruppak，这是一个可追溯的文本-地理锚点对。

排除范围：没有前第三千纪文本记忆的区域。这排除了 Mesopotamia 之外的大部分近东区域，以及 Mesopotamia 内部没有早期文本提及的位置。

逻辑关系：和非 4 不同。非 4 约束传承链的地理学范围（希伯来传统能接触到哪里），非 13 约束文学学范围（哪里有文本记忆可以锚定具体细节）。两者合取比单独使用更严格。

非 14 非居住带

前提：Tevah 的尺寸（137 × 23 米）远超普通住宅（Shuruppak Early Dynastic 住宅典型尺寸约 10 × 15 米）。百米级结构必然是异常建筑——神庙、宫殿、仓库、大型公共建筑、纪念性建筑、或不明功能的大型结构。

排除范围：普通住宅区、工匠区、墓葬区。这排除了 Shuruppak 主丘中已经被识别为住宅群的区域。

逻辑关系：和非 5 协同。非 5 要求候选有足够空间容纳 137 米结构，非 14 要求候选结构本身必须是异常大建筑。两者合起来要求的不仅是空间，还是结构类型。

非 15 非地下

前提：方法论筛选和未来考古验证都需要候选可被探测到。Mesopotamia 5000 年沉积使前第三千纪地表可能深埋 2 到 5 米（Aqrawi 1995 沉积速率）。现代磁力计能探测到约 3 到 5 米深的烤砖异常。超出这个深度的结构不可筛选也不可近期验证。

排除范围：深埋超过 5 米的结构，或完全被侵蚀的结构。

逻辑关系：操作化约束。非 15 不是真实存在性约束（原型可能确实深埋），而是可探测性约束。它定义了方法论筛选的物理可及范围。

15 非的逻辑依赖结构

独立非（不依赖其他非）：非 1、非 2、非 4、非 5、非 6、非 11、非 12、非 13

依赖非（基于其他非的成立）：

• 非 3 依赖非 1（陆地洪水推论）
• 非 7 和非 8 互为协同（本体论联合约束）
• 非 9 和非 3 弱重叠（海进线以北的盆地）
• 非 10 是启发式约束（"缺失证据"推理）
• 非 14 依赖非 5（结构尺寸-功能类型的联合约束）
• 非 15 是操作化约束（可探测性边界）

合取收敛路径：

第一层（非 1, 2, 4, 6）从全球范围筛到近东陆地

第二层（非 3, 9, 11）从近东筛到 Mesopotamia 中部冲积平原古河道带

第三层（非 12, 13）从古河道带筛到有洪水层证据的前第三千纪古城圈

第四层（非 10, 5, 14）从古城圈筛到尚未完全研究的异常大型非住宅结构

第五层（非 15, 非 7, 非 8）把候选限制到可磁测探测的、固定位置的、非船形严格理解的结构

Tell Fara（古 Shuruppak）西南 Mound E 的 Area D 同时满足以上所有约束。这不等于 Area D 就是 tevah 原型，这只等于 Area D 是在本文使用的 15 条约束下最难被排除的候选。

读者的独立复核

本附录的目的之一是让读者可以独立复核筛选逻辑。读者如果反对本文结论，具体操作有三种：

一，增加新的非。如果读者认为有其他约束本文遗漏，可以提出"非 16"、"非 17"。如果新增约束进一步排除 Area D，结论改变。

二，反对某条非。如果读者认为本文某条非不成立（比如认为 tevah 可以严格是漂浮船，反对非 7；或认为源事件可以在海洋，反对非 1），可以给出反驳理由。如果反驳成立，筛选空间扩大，Area D 可能不再是唯一剩余候选。

三，提出漏掉的候选。如果读者在 15 条约束下找到 Area D 之外的候选（本文没有注意到的位置），可以提出来。这不否定 Area D，但让候选集合不再是单点。

这三种操作都是显性化筛选逻辑的自然后果。本附录的存在就是为了支持它们。

---

附录 B · Tevah 指称链系谱

（此附录展开正文第三节的指称链分析。）

结构：

节点列表：

• Latin Vulgate arca（约 400 CE）
• Syriac Peshitta qebutha（约 200 CE）
• Greek LXX κιβωτός（约前 3 世纪）
• Hebrew MT tevah（文本定型前 6 - 前 2 世纪）
• Berossos 版本（通过 Alexander Polyhistor）
• Gilgamesh XI eleppu（约前 13 - 前 7 世纪）
• Atrahasis 版本（约前 17 世纪）
• Ark Tablet 版本（约前 17 世纪）
• Sumerian Flood Story（约前 21 世纪 Ur III）

传承关系图。

埃及词源支线：tevah ← 埃及 ṭbt（盒、箱）。

关键断点：tevah（埃及词源、长方形）vs eleppu（美索不达米亚、方形）的断点。

---

附录 C · Area D 磁测数据关键图

（此附录引用 Hahn et al. 2022 和 Einwag et al. 2025 的关键磁测图，以合理引用方式使用。）

结构：

图 C.1：Hahn et al. 2022 磁测图，显示 Mound E 中 Area D 结构。

图 C.2：Einwag et al. 2025 对 Area D 的解释图（如公开版本中有）。

图 C.3：Area D 位置的卫星图（Google Earth 开放数据）。

表 C.1：Area D 已知尺寸数据汇总。

---

附录 D · Genesis 6:15 四语对照

Hebrew（MT）:

וְזֶ֕ה אֲשֶׁ֥ר תַּעֲשֶׂ֖ה אֹתָ֑הּ שְׁלֹ֧שׁ מֵא֣וֹת אַמָּ֗ה אֹ֚רֶךְ הַתֵּבָ֔ה חֲמִשִּׁ֤ים אַמָּה֙ רָחְבָּ֔הּ וּשְׁלֹשִׁ֥ים אַמָּ֖ה קוֹמָתָֽהּ׃

（"这是你造它的方法：方舟长三百肘，宽五十肘，高三十肘。"）

Greek（LXX）:

καὶ οὕτως ποιήσεις τὴν κιβωτόν· τριακοσίων πήχεων τὸ μῆκος τῆς κιβωτοῦ καὶ πεντήκοντα πήχεων τὸ πλάτος καὶ τριάκοντα πήχεων τὸ ὕψος αὐτῆς.

Syriac（Peshitta）:

ܘܗܟܢܐ ܥܒܕ ܠܗ ܠܩܒܘܬܐ܂ ܬܠܬܡܐܐ ܐܡܝܢ ܢܗܘܐ ܐܘܪܟܗ܂ ܘܚܡܫܝܢ ܐܡܝܢ ܦܬܝܗ܂ ܘܬܠܬܝܢ ܐܡܝܢ ܪܘܡܗ܂

Latin（Vulgate）:

et sic facies eam: trecentorum cubitorum erit longitudo arcae, quinquaginta cubitorum latitudo, et triginta cubitorum altitudo illius.

关键观察：

四版本在 300 × 50 × 30 肘 这组数字上完全一致。跨越 Hebrew（约前第六世纪定型）、Greek（前第三世纪）、Syriac（公元第二世纪）、Latin（公元第四世纪）——跨越约千年的文本传承，尺寸数字精确保留。

Ark 词的四语：

• Hebrew: תֵּבָה (tevah)
• Greek: κιβωτός (kibōtos)
• Syriac: ܩܒܘܬܐ (qebutha)
• Latin: arca

四个词在各自语言中都对应"箱、柜、容器"类概念，不对应"船"（希腊 ναῦς、拉丁 navis 等）。这和 tevah 的本体论候选（船/建筑/raft+mudhif）方向一致——传统把 tevah 理解为"一种特殊的容器"，不是"一艘船"。

---

附录 E · 工作过程中浮现的两个候选方法论工具

定位说明

本附录记录本次应用工作中浮现的两个候选方法论工具。它们作为 input 提交给 SAE 方法论系列（非之道、余之道、以及未来其他可能的方法论论文），是否纳入正式方法论工具库由方法论系列自身的发展决定。本附录不是方法论提议，是工作笔记。两个候选工具的通用性远超本文的 tevah 应用，这里只给出简要说明，更深入的讨论留给方法论系列的未来版本。

E.1 · 指称链漂移时 Via Negativa 的适用边界

问题陈述

当研究对象经历长期指称漂移时，Via Negativa 能在哪些阶段有效工作？

观察

否定神学（传统 Via Negativa 的源起）处理的对象是假设存在性对象——"神是什么"，其中"神"作为对象在历史各阶段没有根本性的对象变形。指称漂移（本文处理的对象性质）则不同——"tevah 在前 6 世纪是什么"和"Ziusudra 在前 21 世纪是什么"可能指向根本不同的物理所指，只是叙事传承链把它们连在一起。

候选边界

Via Negativa 在指称漂移对象上只能工作于漂移终点形态。漂移终点指的是文本定型后的稳定形态——对 tevah 来说是 Hebrew Masoretic 形态，对其他对象来说是各自文本传统中的最终稳定版本。

在漂移终点形态上，Via Negativa 可以有效工作：给定文本属性（长方形、300 × 50 × 30 肘等），排除不兼容的物理候选。

在漂移中间或起点形态上，Via Negativa 直接应用会失败。因为早期形态的属性在传承中可能已经丢失或变形，文本中保留的属性可能不是早期真实的属性。对早期形态直接应用文本属性约束可能排除掉真正的早期原型。

操作指南

对有长期指称漂移的对象做 Via Negativa 筛选，分两步：

一，先做指称链分析（附录 B 的工具），识别各阶段形态及其传承关系。

二，只在漂移终点形态（或文本稳定形态）上应用 Via Negativa。

三，对早期形态，如果需要筛选，必须先用指称链分析反推出各早期形态可能的独立属性，再在各自独立属性集上做筛选。

对 tevah 的具体应用

本文的 Via Negativa 筛选以 Hebrew MT tevah（漂移终点）为起点，用 Genesis 6:15 的属性作为约束。没有直接对 Ziusudra 或 Atrahasis 形态做筛选。这符合本工具的使用建议。

适用范围

本工具对所有经历长期指称漂移的神话对象都可能适用：西王母、Prometheus、Odin、盘古、后羿、Cthulhu（对照组）、以及许多其他长期演化的神话对象。详细应用留给未来 SAE 神话学应用篇。

E.2 · 数字虚实判据 V2（三类诊断）

背景

方法论七现有的数字处理是二元分类（虚指 vs 实指）。本次应用工作发现这个二分不足以处理圆整但非象征的工程数字。第三类"设计化石"是对现有工具的扩展。

判据结构

第一类：象征虚指

• 特征：圆整且对应已知文化象征数
• 来源：叙事编造或文化继承
• 处理：不作为物理证据使用
• 示例：七日、四十日、十二支派、Noah 950 岁、Abraham 75 岁离开 Haran、耶稣 12 使徒

第二类：测量化石

• 特征：非圆整，通常带小数或零散整数
• 来源：直接物理测量
• 处理：作为物理事件的硬证据
• 示例：Shuruppak 洪水层 60 厘米、Kish 洪水层 40 厘米、Area D 磁测 138 × 24 米、Schliemann 发现的 Troy VIIa 城墙厚度、庞贝 A.D. 79 火山灰层厚度

第三类：设计化石

• 特征：圆整但非象征，组合呈现工程比例
• 来源：建造者的设计决策
• 处理：作为建造事件的硬证据
• 示例：tevah 300 × 50 × 30 肘（比例 10:1.67:1）、Giza 金字塔基底 440 埃及王室肘、Etemenanki ziggurat 各层尺寸、紫禁城开间数

应用顺序

一，先检查圆整性。非圆整 → 直接归第二类。

二，圆整则检查是否对应文化象征数。对应 → 归第一类。

三，圆整不对应象征则检查比例结构。简单对称比例（1:1、1:2）→ 可能归第一类。工程比例（10:1.67、1:1.618）→ 归第三类。

伴随约束：度量单位时间下限

设计化石的产生需要标准化度量单位。近东最早标准化肘单位出现在约前 2700 年。此前不可能产生"300 肘"这类圆整设计数字。不同文化的标准化单位出现时间不同，需要按文化核对：

• 苏美尔肘（kùš）：约前 2700 年早王朝 II 期
• 埃及王室肘（meh）：约前 2700 年第三王朝
• 希伯来肘（amah）：派生自更早近东系统，前第二千纪
• 中国周代尺：西周起

对任何古代文本中的设计化石数字，必须核对其来源文明的度量标准化时代。数字不可能来自标准化之前。

应用到 tevah

Tevah 的 300 × 50 × 30 肘属于第三类（设计化石）：300 不是象征数，比例 10 : 1.67 : 1 是工程比例。度量单位约束表明这组数字的物理源头不早于约前 2700 年。这一点直接对任何把源事件置于前 2700 年之前的假说构成约束。

适用范围

此工具对所有古代文本中的尺寸数字都可能适用。Gilgamesh XI 中 eleppu 的 ešer nindanā（十根）尺寸、Homer 中 Achilles 盾牌的描述、《山海经》各地的距离数字、印度 Veda 中 yojana 数字、中国《考工记》中的营造尺寸——三类判据提供一个通用的诊断框架。

---

两个候选工具的共同姿态

这两个工具是本次 tevah 应用工作中浮现的副产品。

关于通用性。两个工具的适用范围远超本文的 tevah 问题。附录 E.1（指称漂移下的 Via Negativa 适用边界）可以应用到任何经历长期指称漂移的神话对象——西王母、Prometheus、Odin、盘古、后羿、Cthulhu（作为人工神话对照组）、《山海经》各地名、昆仑、二里头-夏指称对应等等。附录 E.2（数字虚实判据三类诊断）可以应用到任何古代文本中的尺寸数字——金字塔、ziggurat、希腊神庙模数、罗马引水道、中国宫殿营造尺寸、印度 yojana 距离、Homer 史诗中各类尺寸描述等等。两个工具独立于 tevah，本文只是它们的第一次应用场合。

关于定位。本附录是记录不是推销。两个工具是否值得纳入 SAE 方法论系列（非之道、余之道或未来其他方法论论文）的正式工具库，由方法论系列自身的发展决定。它们作为 input 提交，不预设被采纳。如果方法论系列发展中发现其他更好的工具，这两个工具可以被替代；如果发现它们确实有用，可以被吸收进正式工具库。本附录不对未来走向做任何预设。

献给 SAE 方法论系列的未来版本。

---

参考文献

关于本参考文献列表的诚实声明：下列条目按作者姓氏字母序排列。每条都附有作者可获得性和可访问性声明。对于作者亲自读过全文的文献，标注"全文已读"；对于仅读过摘要或部分页的文献，标注"部分页已读"；对于通过二手转引使用的文献（本文作者未直接读取原文），明确标注"据二手转引"并给出转引来源。这种透明度是本文作为方法论应用论文的基本卫生，读者可据此判断每一条引用的证据等级。

---

Aqrawi, Adnan A. M. 1995. "Correction of Holocene Sedimentation Rates for Mechanical Compaction: The Tigris-Euphrates Delta, Lower Mesopotamia." Marine and Petroleum Geology 12(4): 409-416. DOI: 10.1016/0264-8172(95)96903-4. [部分页已读：摘要和关键沉积速率数据部分（1.3-2.2 mm/yr）]

Aqrawi, Adnan A. M. 2001. "Stratigraphic signatures of climatic change during the Holocene evolution of the Tigris-Euphrates delta, lower Mesopotamia." Global and Planetary Change 28(1-4): 267-283. DOI: 10.1016/S0921-8181(00)00078-3. [全文已读，ResearchGate 开放访问]

Bar-Matthews, Miryam, Avner Ayalon, Aaron Kaufman. 1997. "Late Quaternary Paleoclimate in the Eastern Mediterranean Region from Stable Isotope Analysis of Speleothems at Soreq Cave, Israel." Quaternary Research 47(2): 155-168. [据二手转引，转引自 Dickin 2018。本文未直接读取原文，仅在 Dickin 论述背景中提到]

Bar-Matthews, Miryam, Avner Ayalon, Aaron Kaufman. 2000. "Timing and Hydrological Conditions of Sapropel Events in the Eastern Mediterranean, as Evident from Speleothems, Soreq Cave, Israel." Chemical Geology 169(1-2): 145-156. [据二手转引，转引自 Dickin 2018]

Brown, Francis, Samuel Rolles Driver, Charles A. Briggs. 1906. A Hebrew and English Lexicon of the Old Testament [BDB]. Oxford: Clarendon Press. 通常简称 BDB。תֵּבָה 条目见相关词源段。BDB 有多个开放数字化版本（如 blueletterbible.org, biblehub.com），本文使用其开放在线版。[部分页已读：תֵּבָה 条目与埃及语源讨论]

Brugsch, Heinrich. 1892 (或相关年份)。关于埃及语 ṭbt 到希伯来 tevah 的早期语源学讨论，见 Zeitschrift der Deutschen Morgenländischen Gesellschaft 46 (1892) 或同期相关文章。Erman 相关讨论见同时期文章。诚实标注：本文关于 Brugsch-Erman 词源学论点的引用基于 BDB 和 HALOT 的二手转引，未直接读取 1892 年原始德文文章。期刊数字化版本见 MENAdoc 图书馆（https://menadoc.bibliothek.uni-halle.de/）。

Cooke, G. A. 1987. "Reconstruction of the Holocene coastline of Mesopotamia." Geoarchaeology 2(1): 15-28. [仅摘要已读]

Dalley, Stephanie. 1989 (2000 修订). Myths from Mesopotamia: Creation, the Flood, Gilgamesh, and Others. Oxford: Oxford University Press. [部分页已读：Atrahasis 和 Gilgamesh XI 相关章节，主要通过 Dickin 2018 的引用核对]

Dickin, Alan. 2018. "New Historical and Geological Constraints on the Date of Noah's Flood." Perspectives on Science and Christian Faith 70(3): 176-193. [全文已读，PSCF 开放获取：https://www.asa3.org/ASA/PSCF/2018/PSCF9-18Dickin.pdf]

Einwag, Berthold, Adelheid Otto, Jörg W. E. Fassbinder. 2025. "Harbours, Water Gates, and Canals in Šuruppak and Ur: New Results from Old Sites." In Licia Romano (ed.), Sumer and the Sea: Deltas, Shoreline, and Urban Water Management in 3rd Millennium Mesopotamia (ARATTA 3), pp. 91-100. Turnhout: Brepols. DOI: 10.1484/M.ARATTA-EB.5.143624. [部分页已读：第 96 页精确引用 "115 × 15 米线性结构，西侧矩形平面 138 × 24 米"。该章节全文受 Brepols paywall 保护，本文对第 96 页内容的引用基于 LMU Munich 学术仓库公开的摘录页，具体页面其他内容未完整读取]

Finkel, Irving. 2014. The Ark Before Noah: Decoding the Story of the Flood. London: Hodder & Stoughton. [部分页已读：关于 Ark Tablet (BM 78941) 的核心章节通过 Internet Archive Lending 和 Dickin 2018 的引用核对，大众读物性质，本文只引用其关于圆形船直径约 70 米的数据点]

George, Andrew R. 2003. The Babylonian Gilgamesh Epic: Introduction, Critical Edition and Cuneiform Texts. 2 vols. Oxford: Oxford University Press. ISBN: 978-0198149224. [据二手转引：本文未直接访问 George 2003 原书（受版权保护，需专业图书馆）。对 Gilgamesh XI 第 24-31 行、57-63 行船形描述的引用来自 SOAS BAPLAR 开放译文（https://baplar.soas.ac.uk/）和 George 2000 Penguin 大众版 The Epic of Gilgamesh: A New Translation 的核对]

George, Andrew R. 2000. The Epic of Gilgamesh: A New Translation. London: Penguin Classics. [部分页已读：Tablet XI 章节核对 ešer nindanā 翻译]

Hahn, Stefanie E., Jörg W. E. Fassbinder, Adelheid Otto, Berthold Einwag, Alaa A. Al-Hussainy. 2022. "Revisiting Fara: Comparison of merged prospection results of diverse magnetometers with the earliest excavations in ancient Šuruppak from 120 years ago." Archaeological Prospection 29: 623-635. DOI: 10.1002/arp.1878. [全文已读，Wiley 开放获取]

Heinrich, Ernst / Andrae, Walter (eds.). 1931. Fara: Ergebnisse der Ausgrabungen der Deutschen Orient-Gesellschaft in Fara und Abu Hatab 1902/03. Berlin: Deutsche Orient-Gesellschaft. 数字化开放扫描版由 Stony Brook University Archive of Mesopotamian Archaeological Reports (AMAR) 提供，条目号 585，入口 https://commons.library.stonybrook.edu/amar/585/，PDF 直链 https://commons.library.stonybrook.edu/cgi/viewcontent.cgi?article=1584&context=amar 。[部分页已读：第 6-7 页（扫描 PDF 第 18-19 页）关于 Mound E 厚墙的 Andrae 原文，本文引用基于此一手扫描版直接读取。全书其他部分（目录、Tafel 图版、器物描述等）未完整读取]

Hill, Alan E. 2006. "Quantitative Hydrology of Noah's Flood." Perspectives on Science and Christian Faith 58(2): 130-141. [仅摘要已读，作为 Dickin 2018 的参考文献背景]

Hill, Carol A. 2002. "The Noachian Flood: Universal or Local?" Perspectives on Science and Christian Faith 54(3): 170-183. [仅摘要已读]

Hill, Carol A. 2006. "Qualitative Hydrology of Noah's Flood." Perspectives on Science and Christian Faith 58(2): 120-129. [仅摘要已读]

Jensen, Peter. 1889. 关于 Akkadian 词源的早期语源学讨论，Zeitschrift für Assyriologie 4: 272ff。诚实标注：本文对 Jensen 论点的引用基于 BDB 转引，未直接读取 1889 年原始德文文章。ZA 早期卷次数字化版可能通过 Archive.org 或 DigiZeitschriften 访问。

Köhler, Ludwig, Walter Baumgartner. 1994-2000. The Hebrew and Aramaic Lexicon of the Old Testament [HALOT]. Translated by M. E. J. Richardson. 5 vols. Leiden: Brill. [据二手转引：本文对 HALOT 关于 tevah 词源判断（倾向埃及借词）的引用基于 Google Deep Research 综述转引，未直接访问 HALOT 原书（Brill 订阅数据库，需大学图书馆权限）]

Landmann, Gunter, Andreas Reimer, Stephan Kempe. 1996. "Climatically Induced Lake Level Changes at Lake Van, Turkey, during the Pleistocene-Holocene Transition." Global Biogeochemical Cycles 10(4): 797-808. [据二手转引，转引自 Dickin 2018 关于 Lake Van 沉积记录的讨论]

Mallowan, M. E. L. 1964. "Noah's Flood Reconsidered." Iraq 26(2): 62-82. DOI: 10.2307/4199766. [全文已读，Cambridge Core 开放访问]

Martin, Harriet P. 1988. Fara: A Reconstruction of the Ancient Mesopotamian City of Shuruppak. Birmingham: Chris Martin Associates. [据二手转引：本文对 Martin 1988 的 Fara I/II/III 分期系统的引用基于 Hahn et al. 2022 和其他二手来源的转引，未直接访问 Martin 1988 原书]

McCann, Jason M. 2013. "'Woven-of-Reeds': Genesis 6:14b as Evidence for the Preservation of the Reed-Hut Urheiligtum in the Biblical Flood Narrative." In Jason M. Silverman (ed.), Opening Heaven's Floodgates: The Genesis Flood Narrative, Its Context and Reception, pp. 131-158. Piscataway, NJ: Gorgias Press. [据二手转引，转引自 Dickin 2018 关于 tevah 的 reed-hut 解释]

Morozova, Galina S. 2005. "A review of Holocene avulsions of the Tigris and Euphrates Rivers and possible effects on the evolution of civilizations in lower Mesopotamia." Geoarchaeology 20(4): 401-423. [部分页已读：核心段落关于 avulsion 与年度洪泛沉积的区分]

Postgate, J. Nicholas. 1992. Early Mesopotamia: Society and Economy at the Dawn of History. London: Routledge. [据二手转引：本文对 Postgate 关于苏美尔度量衡标准化时代的引用基于学术综述转引，未直接读取原书]

Qin, Han. 2026a. Myth as Remainder Intuition: A Mythology Application of SAE. Zenodo. DOI: 10.5281/zenodo.18917796. 可从 https://self-as-an-end.net/papers/myth.html 获取中英文版。

Qin, Han. 2026b. Methodology VII: Via Negativa. Zenodo. DOI: 10.5281/zenodo.19481304.

Qin, Han. 2026c. Methodology 00: Via Rho. Zenodo. DOI: 10.5281/zenodo.19657439.

Qin, Han. 2026d. Methodology 0: Via Negativa Foundations (非/Negativa). Zenodo. DOI: 10.5281/zenodo.19544619.

Raikes, Robert L. 1966. "The Physical Evidence for Noah's Flood." Iraq 28(1): 52-63. DOI: 10.2307/4199796. [仅摘要已读，作者通过 Dickin 2018 和其他综述核对 Raikes 关于 Shuruppak 60 cm / Kish 40 cm 的精确对照数字]

Rossi, Corinna. 2020. "On Measuring Ancient Egyptian Architecture." Journal of Egyptian Archaeology 106(1-2): 229-238. DOI: 10.1177/0307513320975782. [仅摘要已读，本文只引用其关于古代建筑测量与象征/功能区分的基本论点]

Ryan, William B. F., Walter C. Pitman. 1998. Noah's Flood: The New Scientific Discoveries about the Event That Changed History. New York: Simon and Schuster. [据二手转引，转引自 Dickin 2018 关于 Black Sea Flood 假说的讨论]

Schmidt, Erich F. 1931. "Excavations at Fara, 1931." The Museum Journal (University of Pennsylvania) 22: 193-245. [部分页已读：University of Pennsylvania Museum 开放获取档案中关于 Shuruppak 洪水沉积层（约 60 cm）的描述段落]

Seely, Paul H. 2004. "Noah's Flood: Its Date, Extent, and Divine Accommodation." Westminster Theological Journal 66: 291-311. [据二手转引，转引自 Dickin 2018 关于 2900 BC 洪水假说的批评]

Steinkeller, Piotr. 2003. "An Ur III Manuscript of the Sumerian King List." In Walther Sallaberger, Konrad Volk, Annette Zgoll (eds.), Literatur, Politik und Recht in Mesopotamien: Festschrift für Claus Wilcke (Orientalia Biblica et Christiana 14), pp. 267-292. Wiesbaden: Harrassowitz. [据二手转引，转引自 Dickin 2018 关于 Ur III Sumerian King List 版本的讨论]

Genesis 6:15 四语文本来源：

Hebrew Masoretic Text 基于 Biblia Hebraica Stuttgartensia (BHS) 或标准 MT 数字化版本（如 sefaria.org, mechon-mamre.org 的开放版本）。[全文已读]

Septuagint (LXX) 希腊文本基于 Rahlfs-Hanhart 校勘版或开放数字化版本（如 biblehub.com/interlinear）。[全文已读]

Peshitta 阿拉姆文本基于 British and Foreign Bible Society 版或其他开放数字化版本（如 dukhrana.com）。[全文已读]

Vulgate 拉丁文本基于 Nova Vulgata 或 Clementine Vulgate 开放数字化版本（如 vatican.va 或 thelatinlibrary.com）。[全文已读]

关于缺失文献的诚实声明：本文在若干节点依赖二手转引。以下列表涉及的作者和文献本文作者没有直接访问，请读者据此评估相应引用的证据等级：

• Brugsch 1892 原始德文文章（通过 BDB 转引）
• Jensen 1889 原始德文文章（通过 BDB 转引）
• Halévy 1888 相关讨论（提及但未引用）
• HALOT 原书（通过 Google Deep Research 综述转引）
• Martin 1988 原书（通过 Hahn et al. 2022 转引）
• Postgate 1992 原书（通过学术综述转引）
• Steinkeller 2003 Ur III 版 King List（通过 Dickin 2018 转引）
• George 2003 Critical Edition 原书（通过 SOAS BAPLAR 和 George 2000 Penguin 版核对）
• Lake Van 数据、Sapropel 数据、Soreq Cave 数据的相关原始文献（全部通过 Dickin 2018 转引）

本文的所有核心论证（Via Negativa 筛选、15 非推导、数字虚实判据、Area D 尺寸匹配）基于作者直接读取的一手材料（Hahn et al. 2022、Einwag et al. 2025 第 96 页、Heinrich & Andrae 1931 第 6-7 页、Aqrawi 2001、Morozova 2005、Schmidt 1931、Mallowan 1964、Dickin 2018、Genesis 6:15 四语文本）。二手转引的文献主要用于背景支持和语言学注释，不承担核心论证责任。

---

本文完。

本文是 SAE 神话学系列的第一篇应用篇。工作基于公开材料，Zenodo 开放发表。CC BY 4.0 许可。