Решения по реконструкции

Сам журнал решений ведётся на английском — языке, на котором он написан и проверен.

Version 1.0, 29 September 2026. Public summary: /about/reconstruction.

This log records every modelling decision behind the site's Third Temple reconstruction (the "synthesis" layer in the Architecture Lab, the 3D massing model, the plan diagrams and every generated image). It was written before any imagery was generated, as the brief requires.

We are not building fantasy architecture and calling it Torah. Where the sources are silent or disagree, the decision below says so and the site shows it.

How decisions are made

1. Extract dimensions and relationships from Mishnah Middot (the only systematic rabbinic description of the Temple's plan).
2. Read Rambam, Hilchot Beit HaBechirah, which codifies Middot as law for the future Temple.
3. Map Ezekiel 40-48 separately, as its own layer.
4. Consult traditional commentators (Rashi, Radak, Malbim) and note where they differ.
5. Record the choice, the alternatives, the reason and a confidence level.

Every text below was fetched from Sefaria and stored in src/content/generated/source-texts.json. Quotations on the site render only from that file.

Why Middot + Rambam is the base layer

Rambam writes (Beit HaBechirah 1:4) that Solomon's building is explained in Kings, and that "the building that will be built in the future, although it is written in Ezekiel, is not explained or elucidated." He adds that the returnees from Babylon built the Second Temple like Solomon's, together with some features explicit in Ezekiel. In 2:3 he states that the altar the returnees built was made "like the altar that will be built in the future." In Hilchot Melachim 11:1 he states the Messianic king "builds the Mikdash."

So the most defensible model of a future Temple, within the halachic tradition, starts from the Second Temple plan preserved in Middot, as Rambam codifies it, and then shows Ezekiel's vision alongside rather than blended in silently.

Confidence labels used below: high (explicit in Mishnah or Tanach, not disputed), interpretive (explicit in one source, disputed or reconciled differently by others), reconstruction (the site's own choice where sources are silent), open (no basis to decide; shown as a question).

Decisions

D-01 Base source

• Choice: Mishnah Middot as codified by Rambam (Beit HaBechirah chs. 1-6).
• Alternatives: Ezekiel-only reconstructions (as in Malbim's commentary or Rabbi Yom Tov Lipmann Heller's Tzurat HaBayit, Prague 1602); a Herodian model based on Josephus.
• Reason: Rambam 1:4, 2:3 above; Josephus is historical literature, not halachic authority.
• Confidence: interpretive.

D-02 Temple Mount enclosure

• Choice: a square of 500 by 500 amot (Middot 2:1). Most open space on the south, then east, then north, least on the west (Middot 2:1). The model offsets the courts accordingly.
• Alternatives: Ezekiel 42:16-20 measures 500 on each side "with the measuring reed"; read literally that is 500 reeds (3,000 amot). Some commentators read cubits. The Ezekiel layer shows both.
• Confidence: high for the Mishnah's figure; interpretive for Ezekiel.

D-03 Orientation

• Choice: the Sanctuary faces east; the Holy of Holies is at the western end.
• Sources: Ezekiel 47:1 ("the forefront of the house looked toward the east"); Middot 2:4 (the eastern wall was kept low so that the priest on the Mount of Olives could see the Sanctuary's entrance).
• Confidence: high.

D-04 Soreg and Cheil

• Choice: a lattice (soreg) 10 handbreadths high; inside it a terrace (cheil) 10 amot wide; 12 steps, each half an amah high (Middot 2:3).
• Note: The Greek warning inscription found in 1871 matches the soreg's function in the Second Temple (non-Jews did not pass it). Access in a future Temple is shown as an open question, not a rule.
• Confidence: high for dimensions; interpretive for exact placement around the courts.

D-05 Ezrat Nashim (Women's Court)

• Choice: 135 by 135 amot, with four unroofed corner chambers of 40 amot: Nazirites (SE), wood (NE), those with skin disease (NW), oil and wine (SW, per Abba Shaul) (Middot 2:5). A balcony added so women could watch the Simchat Beit HaShoevah from above (Middot 2:5).
• Confidence: high.

D-06 The fifteen steps and Nicanor Gate

• Choice: fifteen semicircular steps from the Women's Court up to the Court of Israel, "corresponding to the fifteen Songs of Ascents," on which the Levites sang (Middot 2:5). The Nicanor Gate at the top, its doors of copper (Middot 2:3).
• Confidence: high.

D-07 The Azarah (inner courtyard)

• Choice: 187 amot east-west by 135 north-south (Middot 5:1). East to west: Court of Israel 11, Court of Priests 11, altar 32, between porch and altar 22, the Sanctuary building 100, behind the kapporet 11.
• Elevation: Rabbi Eliezer ben Yaakov (Middot 2:6) says the Court of Priests stood two and a half amot higher than the Court of Israel. The model follows him and labels it as his view.
• Confidence: high for plan; interpretive for elevation.

D-08 Azarah gates

• Choice: the model shows the seven gates listed in Middot 1:4 (three north, three south, one east).
• Alternative: Abba Yosi ben Chanan (Middot 2:6) counts thirteen gates, matching thirteen prostrations. The Lab's toggle shows his list.
• Confidence: open. This disagreement is recorded in the Mishnah itself.

D-09 The Great Altar

• Choice: 32 by 32 amot at the base, stepping in to a fire-place of 24 by 24 (Middot 3:1). Total height 10 amot, with the precise handbreadth breakdown of Rambam 2:5-8. Four hollow horns, one amah square. A red line around the middle separated upper and lower blood applications (Middot 3:1).
• Ramp: on the south, 32 amot long and 16 wide (Middot 3:3; Rambam 2:13). No steps (Exodus 20:23).
• Conflict: Ezekiel 43:17 says the altar's "steps shall look toward the east" and gives a hearth of 12 by 12 (43:16). Rabbi Yose in Middot 3:1 reads Ezekiel's 12 as measured from the centre, reconciling it with a 24 by 24 fire-place. The model follows Middot; the Ezekiel layer shows east-facing steps.
• Confidence: high for Middot; interpretive for the future.

D-10 Altar position

• Choice: East-west, the altar occupies amot 22-54 from the Azarah's east wall (after the two 11-amah strips). North-south (Middot 5:2), from the north wall: pillars 8, tables 4, rings 24, gap 8 = the altar's north face at 48 from the north wall; the altar spans 32; the ramp's 30 ground amot follow; 25 amot remain to the south wall. Total 135.
• Confidence: high. Rambam 2:1 adds that the altar's location is precise and never moved.

D-11 The Sanctuary building

• Choice: 100 by 100 by 100 amot (Middot 4:6). The front (porch, Ulam) is 100 wide; behind it the building narrows to 70, "narrow behind and broad in front, like a lion" (Middot 4:7).
• East-west breakdown (Middot 4:7): porch wall 5, porch 11, Heichal wall 6, Heichal interior 40, the dividing amah 1, Holy of Holies 20, wall 6, side-cell 6, outer wall 5.
• Doorways: porch doorway 40 high by 20 wide, no doors (Middot 3:7); Heichal doorway 20 by 10 with four doors (Middot 4:1).
• Twelve steps from the court up to the porch (Middot 3:6).
• Confidence: high for dimensions; reconstruction for façade detail.

D-12 Holy of Holies

• Choice: interior 20 by 20 amot (Middot 4:7; Ezekiel 41:4 gives the same). Separated from the Heichal by one amah; the Mishnah (Yoma 5:1) records two curtains with a one-amah space between them, while Rabbi Yose holds there was one curtain. The model shows two curtains and labels Rabbi Yose's view.
• Contents: the model shows the Foundation Stone (Yoma 5:2: "three fingers above the ground") and no Ark. Whether the Ark will be present in a future Temple is open.
• Confidence: high for plan; open for contents.

D-13 Heichal furnishings

• Choice: Menorah on the south (left of one entering), Table on the north, incense altar between them, set slightly outward toward the east (Rambam Beit HaBechirah 1:6-7).
• Confidence: high.

D-14 The Laver (Kiyor)

• Choice: between the porch and the altar, drawn slightly to the south (Middot 3:6; Rambam 1:6).
• Confidence: high.

D-15 Chambers

• Choice: three chambers on the north of the Azarah (salt, parvah, washers) and three on the south (wood, the golah chamber with its cistern, the Chamber of Hewn Stone where the Sanhedrin sat) (Middot 5:3-4). Their exact footprint is schematic.
• Confidence: high for existence; reconstruction for placement.

D-16 Slaughter area

• Choice: north of the altar: 24 rings, eight short pillars with cedar beams, marble tables (Middot 3:5). Diagrams only; no depiction of slaughter.
• Confidence: high.

D-17 Temple Mount gates

• Choice: five (Middot 1:3): two Huldah gates on the south, Kiponus on the west, Tadi on the north (unused), and the Eastern Gate bearing an image of the palace of Shushan.
• Confidence: high for Middot; the future is not specified.

D-18 The cubit (amah)

• Choice: the model is drawn in amot. Conversions are offered, never asserted: 48 cm (Rabbi Avraham Chaim Naeh), 57.6 cm (Chazon Ish), and 45 cm (a common modern academic estimate). Ezekiel's measuring cubit is "a cubit and a handbreadth" (40:5, 43:13); Mishnah Kelim 17:9 records two reference cubits kept at the Shushan gate chamber. The default display is 48 cm, labelled as an assumption.
• Confidence: interpretive.

D-19 Siting and the present-day Temple Mount

• Choice: the reconstruction is drawn on an abstract, idealised Mount. It is not overlaid on satellite imagery or present-day structures, and the site makes no claim about the precise position of the Sanctuary on today's Mount, a question on which traditional and academic opinions differ.
• Reason: educational honesty and respect for present-day holy sites. Generated imagery is prompted to show no Islamic or Christian structures, and never to depict removal, damage or replacement of any existing building.
• Confidence: reconstruction (deliberate abstraction).

D-20 Materials and colour

• Choice: pale Jerusalem limestone for walls and courts; the altar and ramp of whole unhewn white stones, whitewashed (Middot 3:4); the Heichal's interior overlaid with gold (Middot 4:1); gates plated with gold except the copper Nicanor doors (Middot 2:3); a golden vine at the Heichal's entrance (Middot 3:8). Rambam 1:11 says the building should be beautified as far as the community can afford, with gold if possible.
• Façade ornament: unknown. The model uses plain ashlar coursing and restrained cornices. Labelled reconstruction.

D-21 Heights and sightlines

• Choice: all enclosure walls high except the east (Middot 2:4); the Sanctuary façade, 100 amot, is the tallest element; the Ulam doorway is visible from the east.
• Confidence: high.

D-22 Elevation profile

• Choice: derived from step counts at half an amah per step: 12 steps up to the Cheil (Middot 2:3), 15 up to the Court of Israel (Middot 2:5), a 2.5-amah rise to the Court of Priests (Middot 2:6, Rabbi Eliezer ben Yaakov), 12 up to the porch (Middot 3:6). The absolute datum of each level is not stated in the Mishnah and is drawn schematically.
• Confidence: interpretive.

D-23 People in imagery

• Choice: kohanim in white linen (Exodus 28:39-42); the Kohen Gadol's garments per Exodus 28 with no invented insignia; Levites in plain garments on the steps; pilgrims in simple, modest, timeless clothing. No modern uniforms, flags or weapons. No faces of real people.
• Confidence: reconstruction.

D-24 The Ark

• Choice: never shown as present in the Third Temple. Rabbinic tradition holds it was absent from the Second Temple (Bavli Yoma 21b; Mishnah Yoma 5:2), and traditions differ on where it was hidden (Mishnah Shekalim 6:1-2; Bavli Yoma 52b). Its presence in the future is open.

Ezekiel layer (shown separately, never merged)

Bavli Shabbat 13b records that Chananiah ben Chizkiyah worked to reconcile Ezekiel with the Torah because its words seemed to contradict it. Bavli Menachot 45a, on one difficult passage, says "Elijah will interpret it in the future." Rashi, on Ezekiel 42:3, writes of one architectural term: "I do not know what this is." Radak and Malbim read Ezekiel 40:2's "city on the south" differently (adjacent to the Mount, or far to the south). These are the reasons the site shows layers instead of one confident answer.

Change log

• 1.0 (29 Sep 2026): initial log, written before image generation.