Shape Grammars

Definition

A rule-based formalism for generating designs: spatial compositions produced by the successive application of replacement rules to shapes. In this wiki the method enters through Anderson & Lovell-Anderson's chapter, which places "simulation of simple shape grammars in visual scripting" as the middle term of its three-part conceptual framework — between "input of micro-level emergent phenomena to delineate spaces" and "an output for visualizing the loop-based calculation by the various micro-level emergent phenomena and macro-level patterns." The chapter's theoretical base is March & Stiny's "Spatial Systems in Architecture and Design: Some History and Logic" (Environment and Planning B 12, 1985), which supplies both the shape < surface < volume ontology (surface-mathematics) and the method's founding claim: "designs and their meanings [can] be viewed as the results of computations carried out according to rules of composition and correlative rules of description" (via note 14). The Gyroid case executes the grammar: "Volume resulting from self-organizing logic, as demonstrated through the shape grammar and assembly of cellular gyroid units. The subdivision of larger surfaces to develop a series of simple modular forms" (Figure 18.2 caption).

Procedure

Define the shape vocabulary: primitives as points and lines (shape), connected into surfaces with "length and breadth," enclosing volume — the shape < surface < volume ladder.
Specify rules, not forms: the designer writes the rules of composition and description rather than drawing the artifact — the design is "the results of computations carried out according to rules."
Simulate in visual scripting: "Systemic (and rule-based) computation and visual-spatial scripting are where a series of nodes or agents at the micro-level are extracted from a data source" — the grammar executed as nodes/agents (see emergence, systems-theory).
Aggregate the outputs: simple modular forms tile or aggregate ("in the mathematical logic of self-organization") into "one larger surface" — the composed whole is never drawn, only generated.
Fabricate: the generated geometry is realized "through both additive and subtractive means," by "an array of modeling or physical prototyping realities" (see digital-fabrication).

Key Thinkers

George Stiny & Lionel March — the formalism's sources via the chapter's most-cited reference (notes 3, 6, 14); Stiny (with James Gips) originated shape grammars in the early 1970s, March & Stiny's 1985 paper supplies the "spatial systems" logic the chapter uses. Not yet paged (see log).
Jonathon R. Anderson & Laura Lovell-Anderson — the method's interior-scale carriers: shape grammars simulated "in visual scripting" as part of the autogenic-topography workflow (2018).
Janine Yeung — the case executor: the "shape grammar and assembly of cellular gyroid units" (Figure 18.2).

computational-design — the method's home cluster: shape grammars as the rule-based strand of algorithmic design, beside parametric and agent-based strands.
surface-mathematics — the ontological base the grammar operates on.
emergence — the generation logic the grammar drives: rules + population → pattern.
parametricism — the continuous-field sibling: both are rule-driven form-generation, but the grammar is discrete and compositional where parametricism is continuous and correlational.
autogenic-design — the process frame: the grammar is the "rules of composition" that make the design self-generating.

Source Support

Sources in the wiki that discuss this method:

Transforming Interior Volumes (Anderson & Lovell-Anderson 2018) — "simulation of simple shape grammars in visual scripting" in the three-part framework; March & Stiny's designs-as-computations thesis (note 14); the gyroid's "shape grammar and assembly of cellular gyroid units" (Figure 18.2).

Open Questions

March & Stiny's claim covers designs and their meanings: "the results of computations carried out according to rules of composition and correlative rules of description." The chapter borrows the first half (computation of form) but not the second (rules of description) — does an interior architecture computed from a butterfly's cuticle have a description rule, or is its meaning undecidable by method?
Are shape grammars and parametric modeling rivals or layers? The grammar generates rules and discrete units; the parameter adjusts values continuously. The Gyroid uses both (shape grammar for assembly, parametric equations for the surface). What is the correct architecture of the two?
Stiny later argued design computation cannot be reduced to description-calculation (the "counting" problem in his later work). Does the chapter's undergraduate studio — grading "self-generated" forms — implicitly take a side in that later debate, and which? (Stiny's later work is uningested — see log.)