Architecture and the Human Sciences (ch. 4 of Design in Architecture)

Author: Geoffrey Broadbent Year: 1973 Source type: chapter Raw file: raw/chapters/broadbent_1973-architecture-and-the-human-sciences.md

Summary

Chapter 4 of Design in Architecture (1973) opens the book's engagement with the human sciences. Broadbent first traces two opposed attitudes the designer may take to his task — the extreme rationalist concerned with "abstract, self-consistent geometry," the extreme empiricist with sensory experience — and argues that the twentieth century's founding fathers were rationalists whose search for fundamental design standards extended, at worst, to prescribing the lives people should live in their buildings. He then audits sociology's own model-building traditions (physical-science, evolutionary, organismic, equilibrium, conflict, mathematical) to show how each trades rigour for distortion, before cataloguing the full alphabetical range of human sciences — anatomy to sociology — as possible sources of design information, concluding that most are still "embryo sciences" more anxious for methodological respectability than design usability. The chapter is the premise-setting document for the design-methods movement's empirical turn: the human sciences as the place to look for what people actually want.

Key Claims

Two traditions, two extreme errors. The rationalist tradition runs from Laugier (1753, Cartesian method — architecture "consisted, essentially, of columns, beams and pedimented roofs," all else 'licences') through Schinkel to Mies van der Rohe, "the supreme exponent," who filled the spaces of his columns and beams with glass. The empiricist tradition runs from the empiricist philosophers through aestheticians (Addison, Gilpin, Price, Payne-Knight) to Picturesque designers (Repton, Nash), conventionally dismissed as irresponsible.
Rationalist buildings underperform on user comfort. Mies's Farnsworth House (1950) and Crown Hall (1955) "confirm that other aspect of architectural rationalism — they contribute rather less to user-comfort in terms of environmental control… than the average greenhouse."
The Picturesque was multi-modal, not merely visual. Stourhead (1743), in Banham's view "the finest work of art which was ever created in England," stimulated smell (planting), hearing (water, grotto), heat and cold (Flitcroft's Pantheon — one of the first central-heating plants since Roman times) and even equilibrium (a timber bridge designed to 'give' as one walked across it). "The finest of them were concerned not just with visual effect but with pleasurable sensory experiences of a multi-modal kind."
Technical innovation was empiricist work. Throughout the eighteenth and nineteenth centuries technical innovation in architecture was largely the work of empiricists; in 1970 an empiricist approach is re-emerging through environmental science, and some would have the architect design "not just for comfort but for delight."
The founders' rationalism prescribed lives. Le Corbusier's 'Manual of the dwelling' from Vers une architecture demanded a south-facing glazed bathroom, "Never undress in your bedroom," and built-in fittings to replace furniture — which "fixes once and for all the arrangement of the room," contrasted with the three-piece suite's "extraordinary number of permutations." The closer the founders approached pure rationalism, the less satisfactory their buildings from the user's point of view.
Le Corbusier also asked for an empirical science of needs. "By needs," he said, "I mean utility, comfort and practical arrangement"; standards should rest on "analysis and experiment"; "All men have the same organism, the same functions. All men have the same needs" — an almost behaviourist programme, with form then deducible from the standards.
Gropius's norm and doubtful diagrams. Gropius defines the standard as an impersonal "norm" (from the carpenter's square) fusing "the best of its anterior forms" minus the personal; his ten-storey-vs-three-storey daylight-angle diagrams (30° vs 17° 50') Broadbent calls "rather doubtful."
Hannes Meyer's reduction and Mies's retort. Meyer's list of "the only requirements to be considered when building a house" (sex life, sleeping habits, pets, gardening, personal hygiene, car maintenance, cooking, heating, insulation, service) omits reading, writing, conversation — living. His "All life is an urge towards harmony… Life is oxygen + carbon + sugar + starch + protein" drew Mies's verdict: "Try stirring that together, it stinks."
Standards are culture- and moment-bound. Meyer's cultural standardization list (bowler hat, bobbed hair, the tango, jazz, Harold Lloyd, DIN…) survived only "as the cult objects of minority groups" — his standards represented "only one cultural group (western Europe and America) at a particular moment in time (the mid-1920s)."
The Bauhaus invited the human scientists. Meyer invited psychologists (largely Gestalt), sociologists and others to the Bauhaus (1927–28) — an initiative Gropius later claimed as his own; Richard Neutra's Survival through Design (1954) made the same bid; by the 1960s the traffic ran both ways.
The new human-science interest in design was political. Cohn-Bendit's Why sociologists? (May 1968) attacked Mayo's Hawthorn factory studies as the origin of sociology selling itself to management: "industrial sociology seeks, above all, the adaptation of the worker to his work" — "intolerable" from a phenomenological view; environmental design offered the chance of "trying genuinely to find out what people want rather than manipulating them to accept what is offered."
Sociology's models each carry dangers. Collating Buckley (1967) and Inkeles (1964): physical-science models import "phoney precision" drawn at the wrong level; evolutionary models (Comte, Spencer, Marx and Engels's five modes, Durkheim, Sorokin) were "largely abandoned… because after certain events of the twentieth century no one can seriously believe that society has evolved"; organismic models (Spencer) draw analogies at the wrong level — societies should be compared with species, not single organisms (Deutsch); equilibrium models (Parsons, Homans) mistake society's ability to maintain itself, opposed by Mills's conflict model; mathematical models (Bales's 45% interaction curve; Simon's stochastic processes for city size, income, word frequencies) mistake technique for neutrality — Simon himself warns against borrowing "analogical models from the natural sciences" wholesale.
Weber's resolution. Between the rational position (insights without fit) and the empiricist position (correlations without causes), Weber set up theoretical models testable yet described through "an individual view — ascribed to a hypothetical observer in a specific historical context," rejecting the statistically average: what relationship has "the average frog" to the particular frogs a biologist observes? Hitchcock and Johnson (The International Style) made the same point architecturally: International Style housing "implies preparation not for a given family but for a typical family. This statistical monster, the typical family, has no personal existence and cannot defend itself against the sociological theories of the architects"; the functionalists "provide what ought to be needed, even at the expense of what is actually needed."
What statistics can legitimately fix. Physiological limits can: exposure ranges outside which 90%, 80%… 10% of humans would die, and comfort ranges — though "it will never be possible to establish a simple temperature at which everyone… would feel comfortable." Such studies are "the province of physiology" — some human sciences can present the designer with useful information.
The verdict on the human sciences. An alphabetical catalogue of twenty sciences (anatomy, anthropology in three registers, anthropometrics, archaeology, demography, human ecology, ergonomics, ethnography, ethnology, ethology, linguistics, parapsychology, pathology, physiology, psychiatry, psychoanalysis, psychology, social psychology, sociology) "differ greatly in scope, methods and aims"; many are "embryo sciences, anxious above all to establish their respectability as true sciences," and "some of their practitioners seem far more anxious to demonstrate their grasp of a methodology than to produce results which are usable in design. They collect large quantities of data and submit it to minute analysis, without really stopping to think what their analyses are for."

Connections

Concepts: standardization · ergonomics · environmental-psychology · empirical-methods · functionalist-myth · user-participation · commodity-firmness-and-delight (design for delight) Authors: broadbent-geoffrey · le-corbusier · banham-reyner Debates: rationalism-vs-empiricism-in-architecture · design-methods-movement · werkbund-debate Related source notes: source-notes/broadbent-1980-signs-symbols-and-architecture (the semiotic successor volume; the four types of design and the functionalist demolition) · source-notes/scott-2001-architecture-or-techno-utopia (the social-science/techno-utopia audit of modernism, twenty-eight years on) · source-notes/deyong-2018-history-style-modern-interior (the Werkbund standardization debate the chapter's standards-question continues)

Direct Quotes

"The supreme exponent, of course, was Mies van der Rohe, who succeeded in building a Laugier-like architecture of columns and beams, eliminating even the pedimented roof." — on the rationalist tradition's endpoint

"The buildings he thus achieved… also confirm that other aspect of architectural rationalism — they contribute rather less to user-comfort in terms of environmental control, than the average greenhouse."

"In Banham's view (1962) Stourhead is the finest work of art which was ever created in England."

"By needs," he said, "I mean utility, comfort and practical arrangement." — Le Corbusier, quoted

"A standard is established on sure bases, not capriciously but with the surety of something intentional and of a logic controlled by analysis and experiment. All men have the same organism, the same functions. All men have the same needs." — Le Corbusier, quoted

"Life is oxygen + carbon + sugar + starch + protein … Try stirring that together," said Mies, "it stinks." — Mies on Hannes Meyer

"In other words, Meyer's examples represented standards only for one cultural group (western Europe and America) at a particular moment in time (the mid-1920s)."

"Industrial sociology seeks, above all, the adaptation of the worker to his work." — Cohn-Bendit et al., quoted

"…trying genuinely to find out what people want rather than manipulating them to accept what is offered."

"The precision of expression which characterizes physical science, with its dimensions of space and time, its forces and vectors, greatly tempt those who weary of the ambiguity of so many sociological terms…" — Inkeles, quoted

"The most obvious, and most often cited, explanation for the appeal of the physical science model is that the success of physicists and chemists has given their approach an aura of power and prestige so great that people are inevitably attracted to it." — Inkeles, quoted

"…it will be safer to notice them after we have developed our theories than to attempt to employ them as a basis of theory construction." — Simon, quoted, on analogical models borrowed from the natural sciences

"…implies preparation not for a given family but for a typical family. This statistical monster, the typical family, has no personal existence and cannot defend itself against the sociological theories of the architects." — Hitchcock & Johnson, quoted

"The idealism of the functionalists too often demands that they provide what ought to be needed, even at the expense of what is actually needed." — Hitchcock & Johnson, quoted

"Weber tried to resolve this by setting up theoretical models which could be tested but describing them at the same time in terms of an individual view — ascribed to a hypothetical observer in a specific historical context." — on "the average frog"

"Clearly these sciences differ greatly in scope, methods and aims. They differ greatly in degree of rigour and the extent to which they can help the designer with positive results. Many of them in any case are still embryo sciences, anxious above all to establish their respectability as true sciences."

"…some of their practitioners seem far more anxious to demonstrate their grasp of a methodology than to produce results which are usable in design. They collect large quantities of data and submit it to minute analysis, without really stopping to think what their analyses are for."

Note on quotations: the markdown scan of this chapter carries no printed page numbers, so quotes are cited by source of the voice within the text (Broadbent's own, or the named figure he quotes) rather than by page.

Open Questions

Which of the twenty catalogued sciences can actually yield design standards, and at what level of generality? Broadbent defers the selection to later chapters; the chapter itself establishes only that physiological limits and comfort ranges are the proven case.
Broadbent twice promises the fuller history "elsewhere (Broadbent, forthcoming)" — the rationalist tradition from Laugier to the present, and the empirical Picturesque's rehabilitation. The wiki does not hold that work; its rehabilitation of Picturesque multi-modalism should not be conflated with later atmosphere aesthetics.
Is the human-science turn's political motive (find out what people want, don't manipulate) realizable in design research, or does any method adopted inherit management's uses? Cohn-Bendit's critique of Mayo is left standing.
The chapter stops at cataloguing the sciences without adjudicating between them — does its design-methods optimism survive its own critique of the "embryo sciences"?