MHAL - Manhattan Architectural Language

MHAL is GAAD's language for Manhattan: the city read as a grammar rather than a picture. It captures how the 1811 grid persists while parcels merge and buildings come and go, so that urban form can be grown and mutated at city scale rather than drawn.

• Written 17 Jul 2026 (Miles's word). s1 may never be empty under the contract, and it stood unknown since the file began - the language had never said why it exists.

Unknown - stated, not hidden.

• No R1..Rn set exists yet.

• City
• District
• Block
• Road
• Plot
• Parcel
• Land Use
• Building
• Landscape

• Nine names, no definitions. s4 asks for a controlled vocabulary: every term defined. The distinctions MHAL relies on - Block is not Parcel, Parcel is not footprint - are stated in s5 and nowhere defined here.

City
`- District
   |- Road
   `- Block
      `- Parcel
         |- Plot
         |- Land Use
         |  `- Building
         `- Landscape

Refined hierarchy from the 1811 study. Key distinctions: Block is not Parcel; Parcel is not Building footprint; Development Site is not necessarily one tax lot; one parcel may hold several buildings; one project may assemble several parcels.

• 1811 reference grammar
• City (Manhattan)
• District/neighbourhood
• Grid (avenue+street network)
• Block (perimeter enclosure)
• Parcel/tax lot (legal unit)
• Development site (assembled parcels)
• Building
• Building part (tower/podium/wing)
• Open space (yard/court/passage/setback/residual)

• Was hierarchy plus x_spatial_hierarchy. The refined reading is the one the study uses.

• Six types in use: contains, defines, assembles_into, belongs_to, controls, occupies. T030: the contract names no base vocabulary to validate them against, and MHAL declares none.

1811 = the Commissioners' Plan of 1811, treated as persistent urban GRAMMAR that sits ABOVE parcels and buildings. The grid persists while parcels and buildings mutate over time - MHAL's core mutation model. Contributed by ChatGPT (Architecture Reviewer, D140); anchor facts verified by Claude 11 Jul 2026.

• street/avenue alignments
• street hierarchy & widths
• numbered E-W street sequence
• numbered N-S avenue sequence
• repeatable block envelopes
• predictable framework for subdivision

• tax-lot boundaries
• building footprints/heights
• land use & zoning
• FAR
• podiums/towers/setbacks
• ownership/merger/subdivision
• modern public realm

• Was x_1811_grammar and x_block_families. Values with units and evidence grades - s7's job.

Unknown - stated, not hidden.

• The verbs MHAL supports are not stated.

Unknown - stated, not hidden.

• Status carries 'Detailed urban generation rules: unknown'. Nothing rule-shaped is in the file.

Unknown - stated, not hidden.

Grade C (Calculated). Block frames derived from four verified subway-station coordinates (Wikidata: 51 St and 77 St on the Lexington Av line; 34 St-Penn and 81 St-Museum on the 8 Av/CPW line) plus this file's verified 1811 dimensions (60.96 m street-to-street depth; 100.00 m Lex-Third length; 213.36 m original west avenue interval; 30.48 m avenue width). Grid vector measured independently on both corridors: 80.7-80.8 m per street, bearing 28.8-29.1 deg east of north - mutually consistent and matching the published grid bearing. Frames padded 22 m; positional tolerance ~25 m (whole-arcsecond anchor rounding). Zoom-to-fit links, not survey geometry. ZoLa /bbox route per NYCPlanning/labs-zola README.

• Was x_evidence_classes, x_data_sources and x_reference_blocks.

Unknown - stated, not hidden.

• The prototype runs a self-conformance check; the language states no validation of its own.

Unknown - stated, not hidden.

Unknown - stated, not hidden.

• Held deliberately empty: research and generation proposals are not language content (D170).

Scale (experimental)

• Stated in s3 Identity as 'Scale (experimental)'. Carried here as its own section.

exploring

• Stated in s3 Identity as 'exploring'. s18's provenance schema is not implemented.