paper
active
2014
paper:2023-12-21-minsik-kim-purplsoc14-properties-pdf-98eb34

Understanding Christopher Alexander's Fifteen Properties via Visualization and Analysis

ByTakashi Iba·Shingo SakaiKao Corporation, Keio University
Fifteen PropertiesChristopher Alexander's 15 propertiesChristopher Alexander's design theoryAlternating RepetitionFifteen Properties of Living StructureBoundariesCentersContrastCorrespondence AnalysisDeep Interlock and AmbiguityEchoesField effectGood ShapeGradientsLevels of ScaleLife (Alexander's broad definition)Life (Alexander's sense)+9 more

TL;DR

Correspondence analysis applied to Alexander's own dependency table (from *The Nature of Order, Book One*, p. 238) reveals that the 15 fundamental properties of wholeness cluster into 4 or 5 coherent groups rather than operating as a flat, undifferentiated list. Using SPSS version 21.0 with symmetric normalization on a binary contingency matrix encoding Alexander's asterisk-marked inter-property dependencies, the analysis projects χ² distances onto a two-dimensional map and yields a stable 4-cluster solution: {CONTRAST, NOT-SEPARATENESS, ROUGHNESS, ALTERNATING REPETITION, GOOD SHAPE}; {LOCAL SYMMETRIES, THE VOID, LEVELS OF SCALE, GOOD SHAPE, POSITIVE SPACE}; {BOUNDARIES, STRONG CENTERS, DEEP INTERLOCK AND AMBIGUITY}; and {SIMPLICITY AND INNER CALM, ECHOES, GRADIENTS, POSITIVE SPACE}. GOOD SHAPE and POSITIVE SPACE function as bridging properties spanning multiple clusters, signaling higher structural centrality than Alexander's prose treatment implies. The paper introduces illustration-mapping — the superimposition of purpose-built schematic diagrams of each property directly onto the correspondence analysis plot — as a dual-mode representation tool for the properties. This implies that the 15 properties are not cognitively or structurally equivalent: three tightly coupled foundational properties (BOUNDARIES, STRONG CENTERS, DEEP INTERLOCK AND AMBIGUITY) may constitute a structural core, while properties like SIMPLICITY AND INNER CALM and ECHOES occupy a more peripheral, dependent role, a structural hierarchy that Alexander's original presentation obscures.

What to take away

  1. 1. Correspondence analysis of Alexander's own 15×15 binary dependency table (The Nature of Order, Book One, p. 238), run in SPSS 21.0 with symmetric normalization, places the 15 properties into 4 broad clusters and 5 finer clusters on a χ²-distance map.
  2. 2. GOOD SHAPE (property 6) and POSITIVE SPACE (property 5) are the only properties that appear in more than two clusters simultaneously, marking them as structural bridges across the property network.
  3. 3. The tightest, most internally coherent cluster across both the 4- and 5-cluster solutions is {BOUNDARIES, STRONG CENTERS, DEEP INTERLOCK AND AMBIGUITY}, suggesting these three properties form the relational core of Alexander's framework.
  4. 4. CONTRAST and NOT-SEPARATENESS form their own 2-property sub-cluster in the 5-cluster solution, co-located on the correspondence plot despite representing apparently opposing spatial intuitions (sharp distinction vs. merging).
  5. 5. The illustration-mapping method — placing 15 schematic diagrams directly on the correspondence analysis biplot — is introduced as a tool for simultaneously encoding structural proximity and visual semantics of Alexander's properties.
  6. 6. Alexander's original dependency encoding rule (an asterisk in cell AB if property A requires property B for complete understanding, from p. 238 of The Nature of Order, Book One) was directly digitized as a binary matrix and used as the contingency table input, making the analysis exactly reproducible from a publicly available source.
  7. 7. The 5-cluster solution separates SIMPLICITY AND INNER CALM, ECHOES, and GRADIENTS as a peripheral cluster, while the 4-cluster solution absorbs POSITIVE SPACE into this group, indicating POSITIVE SPACE's placement is sensitive to the resolution of clustering.
  8. 8. Alexander defines 'center' not as a geometric location but as a source of living power with a fuzzy boundary, a distinction that the 15 properties operationalize but that was absent from his earlier works A Pattern Language (1977) and The Timeless Way of Building (1979).
  9. 9. An open question the paper raises is whether the cluster structure revealed by correspondence analysis is merely an artifact of Alexander's subjective dependency assignments or reflects a deeper structural grammar of wholeness that could be tested across independent annotators or applied domains.
  10. 10. The paper predicts that mapping illustrations onto correspondence analysis plots will open new modes of discussing the 15 properties and facilitate their application to domains beyond architecture, though no empirical test of this transfer is provided.

Peer brief — for seminar discussion

Presented at PURPLSOC 2014 and published in the Krems workshop proceedings (ISBN 978-3-7375-5458-9), this paper takes Alexander's 15 fundamental properties of wholeness from The Nature of Order, Book One (Alexander, 2002) and subjects their internal dependency structure to correspondence analysis, a multivariate technique that projects χ² distances from a contingency table onto a low-dimensional map. The contingency table was constructed directly from the asterisk-coded dependency matrix Alexander himself published on page 238 of that volume, encoded as a binary 15×15 matrix where a '1' in cell AB indicates that property A requires property B for complete understanding. The analysis was executed in SPSS version 21.0 using symmetric normalization. The load-bearing finding is that the 15 properties are not structurally flat: they resolve into either 4 or 5 clusters depending on granularity, with {BOUNDARIES, STRONG CENTERS, DEEP INTERLOCK AND AMBIGUITY} forming the most cohesive core cluster across both solutions, and GOOD SHAPE (property 6) and POSITIVE SPACE (property 5) functioning as cross-cluster bridges. The paper introduces illustration-mapping — overlaying purpose-built schematic diagrams of each property onto the correspondence analysis biplot — as a dual-mode representation that is not achievable through network diagrams or simple dependency tables alone. An alternative method that could have been used is network-theoretic centrality analysis (e.g., eigenvector or betweenness centrality on the directed dependency graph), which would have made bridge-property status quantitatively explicit rather than visually inferred from plot proximity. The implication the paper argues for is that combining statistical visualization with schematic illustration produces a richer understanding of the properties than reading them sequentially in Alexander's text, and that the cluster structure may guide domain transfer of the framework. A hypothesis left open is whether the clusters reflect a genuine structural grammar of wholeness or are an artifact of Alexander's idiosyncratic dependency judgments. The most contestable element for a critical reader is the single-annotator provenance of the contingency table: Alexander's original asterisk assignments were never subjected to inter-rater reliability testing, so the correspondence analysis inherits whatever subjective biases structured those assignments, making the resulting clusters descriptive of Alexander's own cognitive map rather than of wholeness per se. This conflation of author-intent with structural validity is not addressed.

Frameworks (1)

  • Fifteen Properties of Living Structure
    The set of geometric properties that appear in all living structure: levels of scale, strong centers, boundaries, echoes, gradients, deep interlock and ambiguity, local symmetries, roughness, inner calm, not separateness, and others.

Findings (10)

Claims (7)

Hypotheses (2)

Questions (3)

Related work— refs + corpus + external arXiv

Cited / in-corpus / arXiv badges show which signals surfaced each row. Multi-source rows weighted higher.

+1 more

Similar preprints — Semantic Scholar

Cross-corpus bridges (12)

same_concept_as · Nomic cosine

External markdown files that talk about the same concept as this entity.

  • alexander
    Wholeness and the Theory of Centersframeworks/wholeness-and-centers.md0.816
  • alexander
    CHAPTER FOURcanonical/chapters/vol-1/04-chapter-four.md0.805
  • alexander
    Mirror of the Selfframeworks/mirror-of-the-self.md0.799
  • alexander
    Source: https://nature-of-order.stefan-lesser.com/wholenessarticles/garden/2024-12-12_B-Harrell_nature-of-order.stefan-lesser._wholeness_af313d.md0.796
  • alexander
    Source: http://nature-of-order.stefan-lesser.com/telegram-grouparticles/garden/2024-06-05_Theresia-Tanzil_nature-of-order.stefan-lesser._telegram-group_dc0f49.md0.795
  • alexander
    Unfolding and Structure-Preserving Transformationsframeworks/unfolding-and-structure-preserving-transformations.md0.791
  • alexander
    Source: https://stefanlesser.substack.com/p/dynamical-systemsarticles/substack/2023-02-04_Stefan-Lesser_stefanlesser.substack.com_p-dynamical-systems_0b8687.md0.790
  • alexander
    Source: http://nature-of-order.stefan-lesser.com/centerarticles/garden/2021-09-17_Stefan-Lesser_nature-of-order.stefan-lesser._center_ac8f9c.md0.786
  • alexander
    Christopher Alexander Corpus — Document IndexDOCUMENT-INDEX.md0.776
  • alexander
    Source: http://nature-of-order.stefan-lesser.com/levels-of-scalearticles/garden/2022-06-27_Stefan-Lesser_nature-of-order.stefan-lesser._levels-of-scale_7e926d.md0.774
  • alexander
    PURPLSOCpapers/extracted/2023-12-21_Minsik-Kim_PURPLSOC14_Properties.pdf_98eb34.md0.774
  • alexander
    Phase 2 Categorized Extraction — FRAMEWORKStmp/agent-frameworks-2026-05-09.md0.772