Building Without Proof: Is There a Real Urban Design Theory?

Before asking whether urban design has a theory, we need to be precise about what “theory” actually means — because the word gets used loosely, and that looseness is exactly where the confusion in this field begins.

In scientific research, a theory is only considered to have cognitive significance — to be worth taking seriously as knowledge — if it is verifiable and can be tested to be true. That’s the classical, positivist standard: a proposition earns the name “theory” only if someone can check it against observation, and the results can confirm or refute it.

Sharpened further: a theory is a system of hypotheses about a real-world phenomenon that has been tested through observation or experiment and has consistently held up — making it the currently best-supported explanation, subject to revision if new evidence contradicts it.

Three distinctions matter here, and they get conflated constantly in urban design writing:

A fact is a single, verified observation — light travels at roughly 300,000 km/second. Anyone with the right equipment can check it.
A theory is the larger explanatory system built from tested facts — it doesn’t just confirm what’s known, it explains why, and it can predict what hasn’t yet been observed.
A design proposal is neither of these. It doesn’t explain an existing reality — it invents a new one. A designer’s original vision, however compelling, is not a theory simply because it’s called one.

Crucially: a theory does not discover or invent a fact — it explains a fact that already exists in reality, independent of anyone’s awareness of it. The phenomenon comes first; the explanation comes after, and it can be checked by others.

For a theory to be genuinely usable by a practitioner — not just true, but actionable — it needs three further things:

It must identify a real cause, not merely a pattern or correlation.
It must point to a lever — something a practitioner can deliberately manipulate.
It must be tied to a defined, measurable outcome, tested by changing that lever and observing the result — ideally across multiple independent cases, not just one.

This is the test this entire brief applies to everything the field of urban design has called “theory.”

2. Four Definitions of Theory, and Where Urban Design Falls

Before applying this test to history, it is worth checking the definition itself against how other disciplines define “theory” — because urban design does not belong cleanly to any one of them.

Robert Dubin (Social Science) defines a theory by its structure: it must have units (the phenomena), laws of interaction (how the units relate), boundaries (the scope within which it holds), and system states (the distinct conditions under which the system behaves differently). This is a content-neutral test — it asks whether something is built like a theory, regardless of subject.

Nigel Kettey (Applied Disciplines) offers a more permissive, practice-oriented definition: a system of assumptions, principles, and rules of procedure devised to analyze, predict, or explain a behavior. This is the natural home for a design profession — but note it still requires “rules of procedure,” not principle alone.

Swanson & Chermak (Education Research) add an evolutionary requirement: a plurality of interrelated working hypotheses, tested through experiment and empirical evidence, that change slightly over time. Tentativeness and revision are built into the definition itself.

Karl Popper (Science) is the strictest: theories are nets cast to catch the world, refined ever finer — a provisional approximation, never final, always falsifiable.

Testing urban design theory against these four in turn produces an important finding: it does not belong to any single one of them — it is scattered unevenly across all four, which is itself part of why the field’s use of the word “theory” feels so unreliable. Some of what follows in this brief (the geographic growth models, City in Layers) is well-formed by Dubin’s structural test even where it fails other tests. Almost everything qualifies loosely under Kettey’s permissive bar of “assumptions and principles” — but most fails his further requirement of “rules of procedure.” A small minority (the growth models, Lynch, McHarg) behaves the way Swanson & Chermak and Popper require: tested, revised, refined against evidence. Urban design has borrowed legitimacy from whichever of these four definitions was convenient in the moment, without being disciplined by any one of them consistently.

3. Three Different Activities, One Name

Part of why urban design theory is so hard to pin to a single definition is that “urban design” itself is not one activity — it is at least three, bundled under one professional title, and each belongs to a different kind of knowledge.

The social-behavioral dimension — how a human being perceives, moves through, and interacts with a designed surrounding. This is the terrain of Lynch, Jacobs, and the geographic growth models: testable through observation of people, but, as this brief shows repeatedly, only partly and unevenly tested so far.

The engineering/technical dimension — the design of public transport and its nodes, road networks, landscape irrigation, functional street furniture (benches, pergolas, fountains, stairs), and building performance in terms of airflow, ventilation, and daylight to public space, buildings, and green areas. This dimension is fundamentally different from the other two: it rests on real, uncontroversial, tested science — fluid dynamics, solar geometry, transport engineering — with genuine predictive power and defined, measurable performance standards. It is, in fact, the strongest candidate for real “theory” in the full sense anywhere in this brief, precisely because it was never invented by urban design as a discipline — it was borrowed intact from physics and engineering.

The art dimension — urban design as the art of place-making, of shaping places people respond to emotionally and aesthetically. By the very definition of theory this brief has used throughout, this dimension cannot produce theory at all — art creates rather than explains, and is judged by taste, culture, and critical consensus, not verification. That is not a weakness to conceal; it is a precise boundary. This is exactly where the word “theory” should give way to “craft,” “skill,” and cultivated design judgment.

The consequence for everything that follows: no single theory can ever hold across all three of these at once, because they are not the same kind of knowledge. A framework that succeeds as engineering (McHarg’s ecological method) says nothing about the art of place-making. A framework that succeeds as behavioral science (Lynch’s imageability) says nothing about ventilation or transport capacity. And the art dimension, by definition, will never produce a theory that could unify the other two. Urban design’s persistent search for one master theory has, in part, been a search for something that its own three-part composition makes structurally impossible to find.

4. Two Different Kinds of Claim: Form vs. Movement

Urban design theory splits into two families that are often discussed as if they were the same thing, but they aren’t.

A theory of urban form explains what a place is — its shape, its layout, the physical pattern it takes. This is almost always a description of an outcome: the residue of forces — economic, historical, social, or design intent — that have already acted on a place. Form is a noun, not a mechanism.

A theory of people’s movement or behavior explains what a place does to people — how they move through it, use it, and respond to specific, controllable physical conditions. This is a fundamentally stronger kind of claim, because it isolates a variable a designer can actually manipulate (street width, frontage, sightlines) and measures a human response to it (footfall, dwell time, route choice) rather than simply describing a static shape.

This distinction explains, in advance, why some of what follows holds up under testing and some collapses.

5. Before Theory: The Ancient World

Testing every civilization against the strict definition above — an explanation of a real fact, offered by someone, checkable by others — produces a clear pattern.

Babylon and Mesopotamia: cities were enclosed by walls, raised on platforms against flooding, arranged around courtyards. These were functional responses to real problems, refined generation after generation by trial and error — craft tradition, not theory. No one wrote an explanation for others to test; the knowledge simply persisted because it worked well enough not to be questioned.

Egypt and Persia: city form followed cosmological and religious doctrine — axiality, alignment to the Nile, tribute to the gods. This is theology, not theory — belief expressed as form, not an explanation of an observable fact.

Greece — Hippodamus: the real turning point. Hippodamus was described as more of a philosopher than a builder, proposing a grid-planned city tied to a reasoned structure of social order, and this framework was then reapplied by others in multiple cities. This is the first case of someone offering a written, reasoned why behind city form, rather than simply repeating inherited custom — even though it was never tested the way a scientific hypothesis is.

Rome — Vitruvius: the strongest ancient candidate. Rome systematized Hippodamus’s grid logic and, critically, wrote it down as a transmissible technical framework in texts like De Architectura — reasoned justifications for site selection and orientation, meant to be applied and checked by others building new cities across the empire. A genuine proto-theory, even without modern experimental method.

The honest verdict: most ancient city-building was practice and doctrine, not theory. Theory of urban design didn’t exist from the start — it emerged gradually, and even its earliest forms (Hippodamus, Vitruvius) never crossed into genuine testing. They were reapplied through persistence, not verification — and persistence is not the same as proof. A method can survive for centuries purely because no one questioned it, not because anyone confirmed it was correct.

6. The Case-Study Problem

A serious methodological objection needs addressing before going further, because it bears on nearly everything that follows: can a theory be built from a single case study at all?

This is a real, named position in research methodology. Bent Flyvbjerg’s Five Misunderstandings About Case-Study Research catalogues the conventional objection directly: that one cannot generalize on the basis of an individual case, and therefore a single case study cannot contribute to scientific development. Methodologists like Dogan and Pelassy went further, arguing one can validly explain a particular case only on the basis of general hypotheses — everything else is uncontrollable, and of no use. Donald Campbell, in his early work, dismissed single-case studies as having almost no scientific value.

Flyvbjerg’s own paper argues the opposite — that case studies, done well, can and do contribute to theory. But the deeper point for this brief is what actually happened across the theories examined below: the field applied the word “theory” inconsistently across wildly different levels of evidence — everything from a single city treated as a universal law, to genuinely replicated observation across many independent cases. The methodological warning isn’t that case studies can never produce theory. It’s that calling a single, ungeneralized case a “theory” without testing whether the pattern repeats elsewhere is where the field’s credibility breaks down.

That distinction — tested once versus tested as a replicated pattern — is the axis the rest of this brief measures against.

7. Form-Typology Theories: Linkage Theory and Collective Form

Fumihiko Maki’s collective form (compositional, megaform, group form) came from direct observation: Maki traveled and studied traditional villages across the Mediterranean, the Middle East, and North Africa, and inferred a structural relationship between individual houses and the whole settlement — the house unit as generator of the village form, capable of incremental addition without breaking the overall structure.

Tested against the framework:

Real fact? Yes — grounded in genuine observation, not invention.
Tested by others? Partially — the typology has been applied in a variety of projects, including by Maki himself (Hillside Terrace, built in six phases over 24 years).
Defined, measured quality outcome? No. The three forms are a classification, a design vocabulary — not a causal claim tied to a measured result. Maki never ranked one form as producing better outcomes than another.

A striking detail worth isolating: within Maki’s own typology, only two of the three forms (compositional, megaform) involve a deliberate designer applying rules at all. Group form — the one most praised for producing humane, coherent places — is specifically the form that emerged without any designer. The form most admired for quality wasn’t produced by anyone following a method; it was what happens when no one is imposing top-down design at all.

Trancik’s linkage theory (paths, nodes, edges connecting fragmented urban space, building on Kevin Lynch) sits in the same category: a real, reapplied, useful analytical and design tool — but without a defined, tested measure of the quality it produces.

8. Growth Theories from Geography

Human geography produced the field’s closest approach to genuine scientific method — real data, real hypotheses, real testing.

Burgess’s Concentric Zone Model (1925): built almost entirely from one city, Chicago, generalized into a universal claim about how cities grow in rings from a central business district, driven by competition for land value and accessibility.
Hoyt’s Sector Model (1939): built from land-value data across many cities, arguing growth follows wedge-shaped sectors along transport corridors.
Harris and Ullman’s Multiple Nuclei Model (1945): proposed that larger, more complex cities grow around several specialized centers rather than one, because different land uses have different locational needs and some repel each other while others cluster.

Tested against the framework:

Real fact and hypothesis? Yes, genuinely — measured land values, population data, real cities.
Real experiment? Yes — models were checked against data in multiple cities, and this testing did partially falsify Burgess’s rings outside industrial-era Chicago, which is exactly why Hoyt’s and then Harris-Ullman’s models emerged to correct it. This is real scientific self-correction.
Defined quality measure? No — and this is the critical failure point. These models explain pattern, not value. A city that perfectly fits the multiple-nuclei model could be a terrible place to live. Fit-to-model and quality-of-place are entirely separate questions these theories never address.
A lever for the designer? No. These models describe what emerges from thousands of independent, uncoordinated market decisions — no designer is the cause, the market is. There is no step here a practitioner can follow, because no one was designing in the first place.
Used in practice today? No. These models are taught retrospectively, in geography and urban-studies courses, to explain the historical structure of cities that already exist — not applied anywhere today as a method for laying out a new town. Their absence from active practice is itself strong evidence they never functioned as usable design theory.

9. The Social-Reform Vision: Howard’s Garden City

Ebenezer Howard had no training in urban planning or design. He was a social reformer responding to a real, documented problem — the poor living conditions of London’s industrial slums — and he analyzed the real advantages and disadvantages people weighed when choosing city versus countryside life.

His proposed solution — a circular city, radial boulevards, rings of housing around a central garden, industry pushed to the edge, a surrounding greenbelt — was not derived from a functional study of how these components should work together. It was a diagram illustrating a social and economic argument, not a tested design requirement. Howard himself said explicitly that despite the diagrams in his book, each city should be organized according to its own needs and geography — he was telling readers directly not to copy the form literally.

The real, transferable content of Howard’s proposal was social and economic: self-sufficiency, community land ownership, population limits, separation of industry from housing. The geometry was never the mechanism of quality — and mistaking the illustration for the method is a confusion that has followed the Garden City movement ever since.

10. The Perception Theories: Lynch vs. Cullen

Kevin Lynch’s Image of the City (1960) is the strongest empirical study examined in this brief prior to McHarg. It was the product of a five-year study across three independent cities — Boston, Jersey City, and Los Angeles — using interviews, field sketches, and resident-drawn mental maps. The same five categories (paths, edges, districts, nodes, landmarks) recurred across all three cities, and Lynch formulated “imageability” as a measurable criterion for a city’s clarity to its inhabitants.

Tested rigorously — and tested across multiple independent cases, satisfying the replication requirement that broke Burgess. But three serious limits remain:

The retrospective-to-generative gap. Lynch’s method measures the mental image people already have of a place that already exists, built through months or years of lived experience. It cannot be run on raw land with no population. Applying “strong edges, clear landmarks” to a new development is an assumption extrapolated from Lynch’s findings, never itself tested by Lynch or, as far as this investigation found, by anyone since.
The cross-cultural gap. Lynch’s subjects were mid-20th-century Americans, shaped by a specific climate, culture, and mode of urban life. A transient, car-dependent, multi-national population in a hot climate was never part of the tested population — the conclusions may or may not transfer, but that transfer itself was never verified.
Temporal instability. Both sides of Lynch’s relationship — the perceiver and the city — change continuously. A resident’s mental image thickens and shifts with years of exposure; the city’s own landmarks and edges are demolished, redeveloped, or destroyed (the World Trade Center, a textbook Lynchian landmark for lower Manhattan’s legibility, is a stark real-world example). The five elements identified in 1959 Boston describe that Boston, for those residents, at that moment — with no evidence the same relationship held decades later, even in the same place.

Gordon Cullen’s serial vision (Townscape, 1961), by contrast, was never built from systematic, replicated observation. It emerged from Cullen’s own artistic sensibility, illustrated through his own sketches — a genuine and influential design technique, but untested, with no defined outcome measure. Later researchers have occasionally applied it as an analytical lens, but nothing approaching Lynch’s rigor.

11. The Functionalist/Human-Centric Pair: Le Corbusier vs. Jane Jacobs

These two are usually presented as opposites, but under the same test, they turn out to be structurally similar — and it’s worth being honest about that.

Le Corbusier’s Ville Radieuse was a response to real, chaotic, congested industrial cities, but its generative logic was a metaphor — the city imagined as a human body, an organism of separated functional organs — not observed data about how cities actually function well. Where its principles were actually built (Chandigarh, Marseille’s Unité d’Habitation), the results were documented failures against Corbu’s own claims: critics found his zoning ignored local culture and created burden rather than ease; residents nicknamed the Marseille building “the nuthouse.” No quality measure was ever defined or tested — and where it was tested, at scale, it largely failed.

Jane Jacobs, writing The Death and Life of Great American Cities (1961), had no formal planning training. Her method was sidewalk observation from her own window in Greenwich Village — rich, genuinely lived-in knowledge, but a single case, observed by one person, not a structured multi-city study. What separates her from Cullen, however, is precision: her “eyes on the street” claim was specific enough — dense, mixed-use, active streets create informal social surveillance that increases safety — that later researchers, including in criminology and environmental-design studies, could and did test it, with partial confirmation.

The honest comparison: neither passes the full test, but Corbusier’s untested doctrine, built at scale, produced documented human harm, while Jacobs’s untested observation was at least specific enough for others to verify afterward.

12. The One Theory That Fully Qualifies: McHarg

Ian McHarg’s Design with Nature (1969) is the only framework examined in this brief that clears every criterion set at the outset.

Real fact: measurable ecological data — soil type, hydrology, flood zones, slope, vegetation — independent of anyone’s opinion or culture.
Real hypothesis: that land has intrinsic suitability for certain uses based on its ecological characteristics, and development should follow that suitability rather than market or political pressure.
Real, repeatable method: the “layer-cake” overlay mapping technique — transparent maps of each environmental factor layered together, with the most-constrained areas marked to avoid and the most-suitable areas marked for development. Anyone with the same data can rerun the method and reach comparable conclusions.
A controllable lever: where and how a designer places development, protects land, and routes infrastructure.
A defined, measured outcome: flood risk, habitat loss, soil stability — measurable, not aesthetic judgment.
Genuinely replicated: adopted and reapplied at scale worldwide, and directly gave rise to modern Geographic Information Systems.
Stable across culture and time: the underlying physical laws don’t shift by population or era the way human perception and social behavior do.

Its honest limit: McHarg’s method answers where and whether to build — it says nothing about the human, social, or aesthetic experience of the place once correctly sited. It is a genuine, tested theory of a narrow but critical question, not a complete theory of urban design quality.

13. The Common Criterion

Running every case in this brief side by side reveals one identical structural move, without exception:

Every “theory” began as one person’s — or one small group’s — direct observation of a real, existing urban condition, examined through specific, limited case(s), and was then generalized into a broader claim called “theory.”

Hippodamus, Burgess, Maki, Howard, Lynch, Cullen, Jacobs, Le Corbusier, McHarg — every one follows this same pattern. What differed was what happened after that first move: whether the generalization was ever checked against independent cases, whether it named something measurable rather than vague, whether it identified a lever a practitioner actually controls, and whether the underlying fact remained stable across time and culture.

By that measure, only McHarg passes fully. Lynch and the geographers pass partially. Everything else — Maki, Howard, Cullen, Jacobs, Le Corbusier — remains valuable, influential, professional doctrine that the field has simply agreed, by convention and repetition, to call “theory.”

14. The Unanswerable Question

Design is never for an abstraction — it is for a human being who will live in a place and needs to feel comfortable, safe, and satisfied there, while a client needs design quality, cost control, revenue, and added value simultaneously.

Every theory examined in this brief captures only one slice of that whole: McHarg addresses site safety and ecological fitness; Lynch addresses cognitive comfort; Jacobs addresses social safety; Maki addresses belonging and coherence; Le Corbusier addressed (and failed at) hygiene and efficiency; Howard addressed economic and social balance. No study in this entire investigation ever tested the combination — and for good methodological reason: isolating one variable is how testing becomes possible at all. A human being does not experience legibility separately from safety, affordability, or beauty — they experience the whole, all at once, which is precisely what no single study here was designed to measure.

Layered onto this is the temporal problem already identified with Lynch: both sides of the equation — the human being and the city — keep changing. A theory of total human satisfaction in a place would need continuous re-verification, not one-time proof, because neither its subject nor its object holds still the way a physical constant does.

This suggests urban design may be structurally incapable of reducing to a single master theory. What it can offer is a set of separately tested, domain-specific findings — each trustworthy within its own narrow scope — and the act of combining them for one real site, one real client, one real population, at one real moment in time has never itself been tested, and may not be reducible to a formula at all.

15. My Story: From Archive to City in Layers

In recent years I have conducted research on the theory of urban design and the urban design process. the case studies were on three cities which include similar characteristics of industrial era London, Barcelona, Berlin. These cities are couple of similar in the developed countries world and also Latin America. In London case the aim to research the evolution of land-use policy in central London. What I found in the archives was not a story of steady progress, but a documented history of failure: development schemes built by architects that were severely criticized, followed by a retreat from physical-form planning altogether into vague, general text-based policy. That research, and the alternative approach it led to, is set out fully in Rethinking Local Development Plan-Making.

These schemes had tried to solve the destructive living conditions of the industrial age — and in doing so, produced cities that were themselves unsuitable for living. As city elements — roads, underground transport, social housing — were developed and expanded, the consequences reached a point the planners could not control except at the expense of other parts of the city. Social housing schemes placed families alongside unrelated bachelors in shared accommodation, and girls were subjected to rape as a direct, documented consequence of that planning decision.

This is not an abstract failure of “imageability” or an unmeasured claim about “vibrancy.” This is real, recorded harm, caused by design decisions made in isolation from their consequences elsewhere in the city.

City in Layers came directly out of this evidence. It asks the urban designer to treat every layer of the city — transport, green space, waterways, urban form — as something that must be evaluated on its own terms and checked against its relationship to every other layer: how a transport node’s location, size, and capacity connects to the land and the area it serves; how the green space layer is reached and used comfortably; how waterways connect to their surroundings and contribute positively to adjacent layers; how the physical urban form creates a coherent, beautiful extension of what already exists; and how all of this ties back to movement, roads, and people.

Unlike every theory tested earlier in this brief, City in Layers is not a personal observation generalized outward into universal form. Its evidence base is a documented institutional failure, and it exists specifically as a discipline against that proven failure mode: a layer optimized in isolation damages the layers around it. This concept is developed fully, with its London case study, in the City in Layers series, and applied to a real project in the professional practice piece.

16. Comparison Table

Inventor	Name	Purpose	Subject Studied	Centered On	Catch in Application	Replicable?	Practice Lesson
Hippodamus / Vitruvius	Grid planning / De Architectura	Rational, reasoned city layout	Social order, siting logic	Greek city-states, Roman Empire	No empirical testing; reasoned, not verified	Reapplied across many cities, never tested against alternatives	First move from craft to written, transmissible method
Burgess	Concentric Zone Model	Explain city growth pattern	Land value, population distribution	1920s Chicago	Built from a single city; failed to generalize elsewhere	Tested and largely falsified outside its context	Describes market emergence, not design; no lever for a designer
Hoyt	Sector Model	Explain growth via transport	Land value + transport corridors	Multiple US cities	Fits some cities, not others; no quality measure	Partially confirmed across cases	Still descriptive, not a design method
Harris & Ullman	Multiple Nuclei Model	Explain large/complex city growth	Specialized land-use clustering	Larger modern US cities	Explains pattern, not quality	Best-fitting of the three, still descriptive	Same limit — describes unmanaged emergence
Maki	Collective Form	Classify how parts relate to a whole	Traditional villages, own projects	Japan and observed settlements globally	No defined quality measure; group form arose without a designer	Reapplied as design vocabulary	Useful typology, not a quality-guaranteeing method
Howard	Garden City	Social reform — solve slum conditions	Push-pull of city vs. country life	Victorian industrial London	Diagram mistaken for template; he warned against copying it	Underlying social principles transferable, geometry not	Separate the principle from the illustration
Lynch	Image of the City	Understand how people mentally map cities	Human perception of urban form	Boston, Jersey City, LA (1954–59)	Retrospective only; culturally/temporally bound	Genuinely replicated in later studies	Strong diagnostic tool for existing places; weak as generative rule
Cullen	Serial Vision / Townscape	Aesthetic composition of pedestrian experience	Visual sequence while walking	His own sketches, UK townscapes	No systematic testing	Rarely tested by others	Useful design technique, not verified theory
Jacobs	Eyes on the Street	Counter top-down urban renewal	Street-level social life	Greenwich Village, NYC	Single-case, personal observation	Specific claim later tested, partially confirmed	Precise enough to be testable — doctrine maturing toward evidence
Le Corbusier	Ville Radieuse	Efficiency, hygiene, functional order	Zoning, vehicular flow	Idealized/unrealized, later Chandigarh, Marseille	Metaphor mistaken for fact; real-world harm documented	Applied at scale — and largely failed against its own claims	A cautionary case: untested doctrine, built at scale, causes damage
McHarg	Design with Nature	Site ecological suitability	Soil, hydrology, slope, vegetation	Global — method-based, not place-based	Answers where/whether to build, not full human experience	Fully replicated worldwide; underlies GIS	The clearest tested theory in this set — but narrow in scope
City in Layers	Layer-based coordination	Prevent isolated-layer damage	Transport, green space, waterways, urban form and their interrelations	Central London, 1900–2015	Tested in limited urban development projects under the author design management	Structurally designed for reapplication to any site	A discipline against a proven failure mode, not a claim about universal form

17. Conclusion

Every framework this field calls “urban design theory” shares the same origin: one observer, one limited case or a small handful, generalized into a claim presented as universal. That origin is identical across nearly two and a half thousand years of city-building, from Hippodamus to Le Corbusier. What separates a genuine, reliable theory from professional doctrine wearing the name of theory is not where the idea came from — it’s whether the generalization was ever tested against independent cases, tied to a measurable outcome, and shown to hold across time and culture.

By that standard, the field has produced exactly one theory that fully qualifies — McHarg’s ecological suitability method — and it answers a narrower question than most practitioners assume “urban design theory” should answer. Lynch and the geographers pass partially, each within a specific, bounded domain. Everything else — including some of the field’s most cited and taught ideas — remains untested doctrine.

This is not a reason to discard those ideas. Maki, Howard, Cullen, and Jacobs each still contain real, hard-won insight. But a practitioner should know the difference between applying a tested theory and applying a well-reasoned belief — because as Le Corbusier’s Chandigarh and Marseille demonstrate, and as the London archives behind City in Layers demonstrate more starkly still, building at scale on untested doctrine has produced real, documented human harm. The responsible position is not to search for a single master theory of urban design — the evidence here suggests one may not be possible, given how much both the human being and the city itself continue to change. It is to know precisely which tested, partial theory applies to which specific problem, to apply each honestly within its real and limited domain, and to replace what no theory can offer — a guarantee — with continuous, accountable, evidence-checked design judgment.