Design Thinking

Design Thinking
 Digitized by the Internet Archive
In 2023 with funding from
Kahle/Austin Foundation

https://archive.org/details/designthinkingO000pete
Design Thinking

Peter G. Rowe

The MIT Press

Cambridge, Massachusetts
London, England
Sixth printing, 1995
© 1987 by The Massachusetts Institute of Technology

All rights reserved. No part of this book may be reproduced in any

form by any electronic or mechanical means (including
photocopying, recording, or information storage and retrieval)
without permission in writing from the publisher.

This book was set in Helvetica by Achorn Graphic Services and

printed and bound in the United States of America.

Library of Congress Cataloging-in-Publication Data

Rowe, Peter G.
Design thinking.

Bibliography: p.
Includes index.
1. Architectural design—Decision making—Case
studies. 2. Architecture, Modern—20th century—
Decision making—Case studies. |. Title.
NA2750.R68 1987 729 86-8402
ISBN 0-262-68067-X
To Niki and Anthony
Acknowledgments ix

1) 2)
Designers in Action 1 Procedural Aspects of Design Thinking 39

Case Study 1: Making an Urban Place 3 Some General Characteristics of Design

Problems 39
Case Study 2: Making a Building from a
Formal Type 13 Early Theoretical Positions 41

Case Study 3: Reconciling Two Large Staged-Process Models of Problem Solving

Ideas 20 in Design 46

Other Accounts 27 The Information Processing Theory of

Problem Solving 50
Observations and Questions about the
Protocols 34 Heuristic Reasoning and Design
“Situations” 74
Types of Rules and Constraints at Work in
Design 80

Aspects of Design Behavior 94

Limitations of a Procedural View 110

Contents

3) O
Normative Positions That Guide Design Architectural Positions and Their Realms of
Thinking 115 Inquiry 153

Normative Positions 115 Two Realms of Inquiry 153

Surface Features and Broad Inclinations 116 Architecture from a Naturalistic

Interpretation of Man and His World 154
Further Differentiating Features 122
Architecture from a Referential
Problems of Substantiation 134
Interpretation 175
Theory and Practice 149
A Convergence of Issues 199

Notes 203
Bibliography 209
Illustration Credits 221
Index 225
Acknowledgments
This book is largely the consequence of seminars on methods of
inquiry that Peter McCleary invited me to give to students in the
Ph.D. Program in Architecture at the University of Pennsylvania.
To Peter and several classes of seminar students at Penn | owe a
considerable debt for their frank criticism and constant encour-
agement. Without them the book would never have materialized.

| also want to acknowledge the strong support and encourage-

ment of Anderson Todd, my former colleague at Rice, and our
rather special graduate seminar of 1984. And | would like to ex-
press my appreciation to Peter Papademetriou, at Rice, for his
encouragement, and to Timothy Williams, for his early contribu-
tions and support of the enterprise.

Acknowledgment must also be made of some of my new-found

colleagues at Harvard. In particular, | would like to thank Richard
Krauss and John Whiteman for their interest and constructive
criticism. And | am grateful to Bill Mitchell, my old friend and
now colleague, for his very constructive suggestions about or-
ganizing the material and for his encouragement.

Finally, my thanks to Doris Anderson and Debbie Alber for their

work in preparing the manuscript and to Fisseha Wegayehu for
his support on the graphic material.
 7 a - ; v2 7 a ; = 1 r¢ i"

os vy oat 7 oma olin fha

hc 2oat) ee _-s
ary aii No bl ae roe pty rg 8 = a

w ued er etegis paste Tuyagayeg a erie pia | 71%

0 DB 6 ‘Sines ) ioe lamh om tt) hie? Giger |.
Waogl WT agCaare Hiya, WIGS | o ) cots erature ead Geena eee
o (Ul AP IMRT al epee 1 ear a
<_ mualinre Pas oP ie SRT Da oe) Jarre
> na : ey ot Sa ceyuale pew aie x
ce Die fe elem rte lk rr
iment real b See Vee eas | RT| 10 wa Tey ilon)
vu Se) pe 0 - =p veeed> Ye! eon Wh SY saqht y YiEL |

eS (Oe TL (Ra agrine pe: Helles

De eR a
; Heioeaps una se Wh taliaiaer Ach eClninndy
baee; awe a a6 ae
Re) Wie Git] «i oA ar On : pool
SPER CAEN arn ai. il heed Neda
S aii kits ey offs aoupees

ig

,
Design Thinking
| Chapter
Designers in Action
The design of buildings can be viewed in a number of ways.
First, it can be seen from the perspective of the historical record
of production—the lines, shapes, and masses of past buildings
and urban artifacts—interpreted according to various aesthetic
canons, social circumstances, and technical opportunities. Sec-
ond, it can be examined for its conformity with theoretical pre-
scriptions of what constitutes ‘‘proper’’ architecture and ‘‘good’’
design. Some recent examples of this approach are Krier’s Ur-
ban Space, Lynch's A Theory of Good City Form, and Rossi’s
The Architecture of the City (Krier 1979, Lynch 1981, Rossi
1982). Third, its study can take the form of observing what de-
signers do and how they undertake their tasks. Seen in this last
way, design has often occupied an ambivalent position, being
characterized as either a form of fine art or a form of technical
science. From all perspectives, however, design appears to be a
fundamental means of inquiry by which man realizes and gives
shape to ideas of dwelling and settlement. Furthermore, design
is a practical form of inquiry insofar as it is concerned with mak-
ing and a certain commonplace usefulness, quite apart from its
more esoteric benefits (Harrison 1978).

This book is an attempt to fashion a generalized portrait of de-

sign thinking. A principal aim will be to account for the underly-
ing structure and focus of inquiry directly associated with those
rather private moments of ‘‘seeking out,” on the part of design-
ers, for the purpose of inventing or creating buildings and urban
artifacts.

Although they certainly affect design thinking, other important

aspects of practice, such as the organization and administration
of professional activities, will be of litthe concern here. Nor will |
be dealing with the nature of client-architect relationships or
other formal and informal institutional dimensions of practice,
although in other renderings of the general subject such rela-
tionships and dimensions might profitably have a central role
(Kostof 1977, Tafuri 1976). Their influence on the outcome of
design is undeniable in most architectural and urban design
undertakings. Nor shall | dwell on the manner in which the re-
sults of architectural and urban design might be construed by
some specific audience or by society as a whole, although again
these are undoubtedly pertinent issues (Jencks and Baird 1969,
Bonta 1979). Nor shall | advance a theory about architecture in
the sense of how design should be carried out and what makes
for “good” architecture. Mine will be a neutral account in this
regard.
 My subject is more narrowly defined. | am concerned with the
interior situational logic and the decision-making processes of
designers in action, as well as with theoretical dimensions that
both account for and inform this kind of undertaking.

A useful way to begin the development of a general portrait of

design thinking is by looking at some actual examples of design-
ers at work. It is here, in the give and take of problem-solving
situations in the real world, that we start to see the complex tex-
ture of decision making. And it is here that we discover there is
no such thing as the design process in the restricted sense of an
ideal step-by-step technique. Rather, there are many different
styles of decision making, each with individual quirks as well as
manifestations of common characteristics. Sometimes the un-
folding of a design is strongly influenced by constraints derived
from the initial setting of the problem, such as the context in
which the building is to be-built or its social purpose. On other
occasions the process seems more determined by a designer's
personal attitudes and prejudices toward such things as func-
tional expression or modes of fabrication technology. More of-
ten than not, there is a mixture of both orientations, as designers
move back and forth between the problem as given and the ten-
tative proposals they have in mind.

Three case studies of designers in action will be presented. Each

involved lengthy periods of observation and documentation,
where designers described their activities in detail with the aid of
sketches and other drawings. The aim of these interview ses-
sions was to faithfully reconstruct the sequence of steps, moves,
and other logical procedures that were employed. This type of
reconstruction is referred to as a protocol; and while it omits the
minute-by-minute detail and other real-time variations in design
_ behavior, it does represent a plausible record of major events
(Hayes 1981, pp. 51-57; Pohiman 1982). Furthermore, the sub-
jects had the opportunity of providing detailed explanations of
the rationale behind each step, thereby providing greater insight
than might be gained from mere observation of surface activity.'

The three case studies were chosen, from among a more exten-
sive collection of protocols, because they illustrate three differ-
ent styles of design thinking. One protocol seems more
constrained by information derived from the immediate context
of the design problem. By contrast, another is almost totally
dominated by the a priori use of a particular building type as a
model for resolving the problem at hand. In the third example

Designers in Action 2
 two large ideas seem to dominate the process of designing, as
much through their conflicting influence as through their com-
plementary effect. The full range of design thinking is by no
means represented by these three examples, but they do provide
sufficient material and variation to serve as an informative back-
drop for later theoretical discussion.

Further examples of the design process, taken from various writ-

ten sources, will follow the case studies. Certain features will
emerge as characteristic of the problem-solving behavior of ar-
chitects and urban designers, particularly its episodic structure
and the fact that the designer relies on presuppositions and
hunches about architecture at least as much as on intormation
furnished during orderly confrontation with the constraints
found in a given design problem. It will, | hope, be made evident
that in order to clarify the essential method and style of inquiry
encompassed by design thinking, we must venture beyond the
traditional boundaries of explanatory theories about creative
problem solving. We must engage the normative realm of dis-
course about architecture if we wish to gain insight into what
shapes the content rather than the strictly procedural aspects of
design activity.

FSi eSPe et Ny Dee ee ee

Case Study 1: Making an Urban Place
The first case study concerns the problem of designing a com-
mercial complex for some 1.5 million square feet of office space
on a suburban site of a large American city. The area of the site
is 16.5 acres, spanning roughly north-south between a major
thoroughfare and.a small river. The site is irregular in shape,
with the widest extremity bounding the adjacent thoroughfare.
Other characteristics of the site include a small area in the
floodplain next to the river, extensive wooded areas, and the
remnants of old buildings. The surrounding uses consist of a
single-family residential subdivision to the east, a wide utility
easement to the south, vacant land to the north, and an apart-
ment complex under construction across the river to the west.
The site is some ten miles from the central business district of
the city, although in an area that has already seen considerable
development of commercial office facilities. In short, the pro-
gram of uses and the site are not at all unusual for commercial
office developments in outlying suburban areas of many modern
American cities.

Designers in Action 3
 27422“) fe&

1
Case study no. 1: early sketches.

Designers in Action 4
 aTavHy

bE og

As shown by the early sketches (figure 1, parts 1 and 2), the

designer began to frame his intentions with the ideas of provid-
ing a strong sense of corporate image, giving a strong sense of
address on the major thoroughfare, and making use of the
wooded amenity of the site. Rather than electing to explore
these ideas in the more traditional manner of isolated tower
buildings arranged within the site boundaries, at the outset the
designer attempted to develop a strongly urban form that ac-
knowledged the public domain of the street by pulling the build-
ing masses forward on the site and creating a plaza along the
thoroughfare boundary. As we shall see, this was a persistent
aspect of the design that became more highly resolved as the
process unfolded. In making these moves the designer mani-
fested an intention to establish a context for future develop-
ments in the area, especially those along the major thoroughfare
next to the site.

The designer’s attention then turned to the issue of the various

kinds of office space and supporting commercial areas that
could be provided. Here the emphasis was on the accommoda-
tion of major and minor corporate tenants and the making of
subordinate spaces within the development that would give each
some sense of identity. In addition, the requirement of parking
some 5,000 cars was addressed, together with modes of entry
and egress and the question of the appearance of the complex

Designers in Action 5
 STREET Nee BL0g |
SINCE FRWATE secre

ELEWIPLS /2 Freer PMA MOVTS

EOLE INTEGLAT/ON
KETA/L : Ar BASES of Buie PiMgs
SERVICE
INTERSECTIONS
WHAT 13 ROLE OF OPEN SPACE
CIWIL GESTURE — MAN/Clee, GARDEN
DAKCE =SUTPLY
WYIE THEATRE a

=
Wont
ow
"NE AND

se
FATLAT. - MOTIFS (ie:
\
— , BEE AW MLIWATE
WO SQUARE CAR PARKS
ae FS SmeeT
4 FUNC HONAT
Fe
CAN Deel
PARKING one Fleer.

2
Case study no. 1: exploration of design ideas.

Designers in Action 6
 when viewed from the upper floors of the buildings (figure 1,
part 4).

At this stage a linear arrangement of buildings was provisionally

established, dictated in large measure by the size of the program
of accommodations and the geometry of the site (figure 1, part
5). There was also an attempt to orchestrate a sequence of
public spaces, emanating from the major plaza, on which each
office building had an individual address, onto the street, on
which the complex had a common address. Evaluation of this
plan, however, indicated that the linear scheme was problematic
from the standpoint of efficient distribution of rentable office
space in conjunction with circulation, especially in the vertical
direction. This realization on the part of the designer initiated a
flurry of experimental arrangements (figure 2, parts 1—4). First,
two public spaces were created and the buildings rearranged
with the lower-rise structures placed at the rear of the site so as
to minimize destruction of the site’s natural amenity. A relatively
large number of plausible arrangements seemed possible, how-
ever, and there was a sense of backtracking in the direction of
the earlier linear scheme. Generally, the probiem at this stage
seemed to be underconstrained and lacking a specific direction.

The apparent deadlock was addressed by systematically evaluat-

ing various aspects of the scheme (figure 2, parts 5 and 6; figure
3. part 1) for the purposes of retrenching and more thoroughly

Designers in Action 7
 VEFIAILLDAD 19 AY
Th, RC fs
AN a y
2 Sere

Blew are
MECH. E-

Case study no. 1: development of design ideas.

Designers in Action 8
 3 4

validating some of the initial design ideas. For example, the

plaza on the street was reasserted as a primary idea. The rela-
tionship of car parking to concentrations of building mass was
also reappraised. Essentially, guidelines or rules were estab-
lished that helped the designer plan and prepare for subsequent
exploration. Indeed, the combination of the results from these
evaluations began to give shape to a more definitive organiza-
tion for the scheme. First, harking back to an earlier proposal,
two major exterior spaces were employed, around which the
building program became organized. Each space was seen to
have a landscape quality befitting its location on the site (a
public plaza on the street and a more informally landscaped area
in the interstices of the site; figure 3, part 2). Second, the prob-
lem of providing an appropriate sense of entrance and address
to the major office components was addressed by placing the
primary vertical elements of circulation around a common base
that, in turn, defined the street plaza. This base was also used to
accommodate main entrances and retail commercial uses, with
structured car parking rising up behind, also forming a base for
the office towers (figure 3, part 3). Aspects of this general ar-
rangement were then explored and elaborated in more detail,
including the provision of separate means of entry, egress, and
parking for office workers and visitors. The architectural treat-
ment of the building base was also studied further, with an ac-
centuation of the vertical circulation expressed to mark the

Designers in Action
 building entrances, within the horizontal banding of the plaza’s
enclosure. Attention was also directed toward the landscape
qualities of the street plaza (figure 3, part 4). The potential func-
tion of this major exterior space was investigated further, with
the result that it became cast in the role of outdoor amphitheater
and storm water detention and retention device. Parenthetically,
this latter technical function was a general requirement of the
development program.

Exploration of the arrangement and articulation of the building

base, on which the office towers were to be placed, was con-
tinued, with special attention to the transition between the two
major exterior spaces (figure 3, part 5). Generally, the base was
regarded as building fabric, with the primary purpose of defining
these public exterior spaces. Resolution of this aspect of the
scheme involved a dimensional interplay between the shape and
scale of the outdoor spaces and the planning requirements for
efficient circulation and for adjacency of car parking to the of-
fice towers. It should be noted that a number of more precise
and dimensionally accurate drawings, not shown here, were
made to test these assumptions and to help firm up the scheme.

Having resolved the layout and formal arrangement of the lower

floors, at least for the moment, the designer proceeded to inves-
tigate various methods of architecturally articulating the office
towers (figure 3, part 6). Uppermost in his mind were the provision
of a distinctive image quality, the making of distinctions among
various tenant functions, and the clarification of entrance and ad-
dress. In addition, considerable attention was paid to matters of
structural design and building erection technology. The decision
was made to employ lightweight steel framing for the office tow-
ers because of the economic advantages of speedy fabrication
and because of a preference for the architectural qualities of the
glazed fenestration and spandrel panels that would automati-
cally follow by the same line of reasoning. At this stage a num-
ber of detailed design investigations, not shown here, were
undertaken to more properly incorporate these technical consid-
erations and to address the matter of providing a variety of effi-
ciently planned office suites.

With the front portion of the project approaching resolution, sev-

eral issues introduced by the arrangement were identified and
pursued further. First, the car-parking requirements effectively
dictated construction over much of the site, obliterating most of
its natural character. Second, another consequence of the ex-
tension of the building base across the site was the comparative

Designers in Action 10
 lack of acknowledgment of the ‘‘edge conditions”’: the perimeter
of the overall building mass essentially walled off the site along
its boundaries. Finally, the phasing of construction was acknowl-
edged as potentially problematic in an otherwise unified and in-
tegrated proposal.

Recognition of these aspects of the design led to a radical re-

consideration of the back portion of the complex, toward the
river. The natural landscape was reinstated in the form of a ter-
raced garden on top of the parking structure, flanked by lower-
rise office accommodations relating more directly to the edge of
the site (figure 4, part 2). Another office tower complex was
located on an axis with this formal garden, at the back of the
site, thus alleviating the problem of distributing the parking in
close proximity to office circulation (figure 4, part 3). This ar-
rangement also had the advantage of rationalizing the phasing
and construction of the development, from the southern portion
of the site forward to the street, in easily identifiable units. More-
over, it provided ample accommodation for the development’s
primary tenant in a secluded location with the most outdoor
amenity but still with a strong presence, because of the reduc-
tion in the massing of the three towers on the street plaza.

Further development of this overall arrangement was under-

taken, including systematic studies of vehicular and pedestrian
circulation (figure 4, part 5) and the formal treatment of towers
and public spaces. Incremental adjustments were made, and the
scheme began to be finalized and moved forward into detailed
delineation (figure 4, part 6).

In summary, we see a design scheme that developed from a few

well-chosen urban concepts that were more an interpretation of
what the project might be like than pragmatic responses to pre-
vailing site conditions and program. The protocol exhibits at
least three or four major sequential episodes of design activity,
in which the full consequences of earlier proposals became evi-
dent and were explored. Through these efforts the framework
for further design was clarified. Periods in which the problem
appeared underconstrained to the designer were immediately
followed by systematic reevaluation of his position and an
assessment of the potential outcome of various lines of rea-
soning. Not all the initial intentions were completely satisfied,
however, particularly with regard to the site’s natural amenity.

Designers in Action 11
4
Case study no. 1: sketches from the final stages.

Designers in Action 12
 Aes yr PHASING
ope |PARAS
TECHNICAL

Case Study 2: Making a Building from a

Formal Type

In the second case study a somewhat different approach was

taken by the designer, although many shared characteristics
are evident. The project called for construction of a hotel and
comprehensive health facility on a naturally landscaped site in
a relatively undeveloped area. The site included a lake, several
wooded areas, and an existing hospital to which the proposed
building complex was to be connected. The primary function of
the new health facility was health maintenance through exercise,
dietary control, and ‘‘clean living.”

For preliminary planning purposes the designer divided the pro-

gram of uses into three parts (Rowe 1982): the hotel, the health
facility proper, and a third element housing shared functions
such as restaurants, entrance ways, and car parking. The de-
signer then proceeded with the intention of ‘‘addressing and
preserving the lake on the site... as a major focus of the
scheme.” The first move was to create a ‘‘formal space next to
the lake,’ but shielded from the portion of the site where the
main access was to be found (figure 5, part 1). To accomplish
this, a ‘‘classical villa’ served as the point of departure, linking
the two major programmatic elements—the hotel and the health
facility—around the lake.

Designers in Action 13
5
Case study no. 2: early concepts.

Designers in Action 14
 This decision to deal with a functionally subordinate element
first may appear curious. The choice was made, however, be-
cause of the apparent correspondence between the villa type
and the problem at hand. This correspondence was seen in
terms of the location on the water, a certain symmetry and for-
mality in the building program, and the presentation of a formal
facade to the public side of the site and an informal facade to
the lake. The adoption of the villa type as a model also furthered
the initial intention of simply defining a special place by the lake,
but the intention seems to have been derived as much from the
model as from a study of the site in the absence of the model.
The treatment of the space between the building complex and
the lake was to become a continuing theme throughout the
design.

As shown in figure 5, part 2, problems were immediately en-

countered with this arrangement, due in large measure to the
overpowering physical presence of the two larger elements
—the hotel and the health facility—on either side of the villa.
Consequently, the designer generalized the symmetrical ar-
rangement of the building form, linking these two parts into a
semicircular plan (figure 5, part 3). This move also gave rise to
the idea of defining a circulation path through the pieces at the
base of the building masses and thus having an ordering impact
on the scheme. Considerable attention was then directed toward
reconciling the asymmetrical program with the symmetrical spa-
tial conception by elongating one arm of the plan composition
and shifting the inward lake front focus around to the eastern
edge of the site (see figure 5, part 4).

Evaluation of the scheme at this stage revealed problems with

the scale and shape of the hotel, as well as the lack of any formal
difference between it and the health facility. As the designer ob-
served, ‘‘it was a problem of thinking of it [the health facility] just
as a piece in a formal arrangement, so | started analyzing it as a
separate piece.” At this point a shift occurred in the type of or-
ganizing principle being employed, from a preoccupation with
formal composition derived from the model type to a concern
with the geometrical properties of structural bays and with the
abstract functional arrangement of required program facilities.
To continue the narration, ‘‘Basically | was trying to take a struc-
tural diagram, a functional diagram, and a circulation diagram
and combine them” (figure 5, part 5). These considerations
amounted to a redefinition of the problem at hand. The result
was an alternating plan sequence of large and small structural

Designers in Action 15
 bays, within which circulation and functions of various sizes
could more appropriately be accommodated (figure 5, part.6).

The same level of strategy was then used to plan the service
functions of the hotel. By ‘‘working strictly in sort of square bub-
bles, and within an understanding of proximity requirements
among the uses,” the designer ‘‘tried to arrange the pieces so
that they would make sense.” That is, having organized the con-
ditions of immediate interest into a well-defined problem, the de-
signer proceeded to explore various ways by which building
elements might be arranged and then turned to a more system-
atic form of exploration, once this initial sense of the problem
had clarified certain principles and procedural rules.

With the problem of space-planning the facility at least partially

resolved, attention was again shifted to the overall plan of the
building complex. As the designer put it, “One of the conse-
quences of adopting the semicircular symmetrical form was that
it brought in an [implied] axial arrangement that really had no
relationship to everything else that was going on. If | was going
to stay with that form, | would have to develop a relationship,
and so | went back to looking at the whole site.’’ Perhaps not
surprisingly, an axial composition of building and landscape ele-
ments evolved, with a progression from the site’s main entrance
through to the lake. Adoption of these ordering principles
quickly led to a realization that a formal outdoor space was ap-
propriate for the entrance to the building complex, both for
functional reasons such as car parking and in order to complete
the composition (figure 6, part 1). Further, the designer now felt
that she ‘‘was dealing with two major outdoor spaces: one at the
end of the axis when you arrive and the other facing the lake”;
she characterized the first as ‘‘a formal public arrival space” and
the second as “‘a more informal space’’—something of a reversal
of the earlier data about placing the formal space next to the
lake.

These decisions made, the most pressing issue was which way
the curvilinear form of the hotel should face—out toward the
lake or out toward the street. The designer also realized that she
“‘was trying to use the hotel piece to solve all problems and was
having difficulty with this.” At this point many different experi-
mental arrangements were made with alternative plan shapes
(figure 6, part 2). Ultimately, the designer decided on a more
simplified formal arrangement for the hotel, simply running a
straight slab block across the site (figure 6, part 3). This arrange-

Designers in Action 16
 6
Case study no. 2: concept exploration.

Designers in Action 17
 ment resolved the overall scheme for the major elements of the
project in the designer’s mind.

Within the now well-established framework, solutions to more

detailed problems were developed. For instance, the open-ended
expression of the primary public circulation corridor, running
beneath the slab block and defined by the columnar structure of
the hotel, was terminated at the junction with semipublic func-
tions in the form of an alcove (figure 6, part 4). The columns
themselves and the section through the base of the hotel were
articulated so as to give a sense of grandeur to the public entry
space and to allow for a view through the building toward the
lake (figure 6, parts 5 and 6). The northern, public facade of the
building was composed from a proportional grid with intervals
determined so as to diminish the apparent scale of the building,
always a problematic condition to the designer (figure 7, part 1).
This facade was further divided into a distinct base, middle, and
top, no doubt as an outgrowth of the classical references em-
ployed earlier in the process (figure 5, part 5). A more informal
textural treatment was proposed for the southern facade facing
the lake (figure 7, part 2), consistent with the quality of its loca-
tion, although here again the organizational properties of a grid
are evident.

This process of refinement, adjustment, and embellishment

continued. The only subsequent modifications to the overall
arrangement of the complex arose through attempts to extend
the strategy of expressing various site constraints, a strategy the
designer had earlier found successful. For example, several lines
in plan at 45 degrees were introduced, in order to better relate
the building complex with the existing hospital across the lake
(figure 7, part 3). This feature had the additional purpose of help-
ing to organize a greater sense of informality in the architecture
along the lake front (figure 7, part 4).

In this case study, several distinct lines of reasoning can be iden-

tified, often involving the a priori use of an organizing principle
or model to direct the decision-making process. The use of the
villa type at the outset of the process was as much a precon-
ceived notion as it was a response to the broad array of site
conditions and the building program. It functioned as a domi-
nant organizing principle through which both the site conditions
and the building program were subsequently conformed. More
abstract relational models, dealing with conditions of structural
organization and an expressed preference for functional proxim-
ity, were used effectively to solve specific layout problems. The

Designers in Action 18
 Case study no. 2: project resolution.

Designers in Action 19
 symmetrical compositional principles, apparently engendered by
the earlier preoccupations, seemed to persist, exerting a pro-
nounced influence over subsequent lines of investigation. As in
the first case study, a certain amount of backtracking and con-
solidation of the problem constraints was evident, giving a dis-
tinctly episodic structure to the process.

Peres Leora uedalle ae NP aera Ls i |

Case Study 3: Reconciling Two Large Ideas

The third case study concerns a design for a world bibliographic

center on a waterfront site, adjacent to the downtown area of
Chicago (see figure 8, part 1). The program for the center called
for a ‘‘hard-copy” library; a computer and data-processing facil-
ity, including telecommunication linkage with similar complexes
in other parts of the world; and several theaters, or auditoriums.
The site itself was interesting in the light of Burnham’s plan for
Chicago from the “‘City Beautiful Movement’: the designers had
to assume completion of the area of parkland and civic buildings
begun by Burnham to the south of the site.

From the outset the designers recognized the site as a potential

point of symmetry with Burnham’s scheme and entertained the
idea of establishing this symmetry by extending out into the lake
on a pier structure, similar to those adjacent to the site on the
river side (figure 8, parts 2 and 3). Attention then became
focused on the overall shape that the library facility should take
on the site location opposite Burnham’s planetarium. Initial
speculation was strongly influenced by parallel studies of Burn-
ham’s plan, at least with regard to the massing of the building in
the form of regular solids, although other historical references
could also be seen at work (figure 8, part 4).

This episode was followed by a more systematic evaluation of

the site and its environs. For this purpose a model was con-
structed (see figure 8, part 2). The features of primary interest
were the relatively tight grid pattern of Chicago’s downtown de-
velopment contrasted with the larger grid blocks of the parkway,
the bounding of streets by buildings, and the presence of low-
rise structures on the waterfront (figure 8, part 5). Generally
speaking, the project’s context was defined in these terms.

Two themes emerged from these contextual studies (figure 8,

part 6). One was the creation of a landmark at the point of sym-
metry with Burnham’s plan. The other was the development of a
scheme in the form of a linear system of buildings and interstitial
spaces that would thematically extend the Chicago grid pattern

Designers in Action 20
 Web SA) ES od eg 0) LOPLI
WN ps JUOG 4 Ge His =I f=
LeetEt OU

8
Case study no. 3: site constraints and early sketches.

Designers in Action 21
 out into the lake. In figure 8, part 6, we also begin to see ex-
perimentation with various axial arrangements of objects and
three-dimensional figural experimentation that again makes
strong references to the buildings of the Chicago Exposition.

The process of exploring and elaborating these two themes con-

tinued (figure 9, parts 1 and 2), with attempts to position the
“landmark building” within the linear grid system and to vary the
meter of the grid pattern in order to more fully understand its
potential properties for ‘‘place making.” Gradually the scheme
became resolved as a linear form, protruding into the lake, ex-
tending the Chicago grid and terminating with a rotunda-like
structure reminiscent of the regular solids postulated at the be-
ginning of the process (figure 9, part 3).

At this stage of development, however, a programmatic evalua-

tion of the scheme suggested to the designers that the linear
proposal would both require more building facilities than were
available and prove to be an inefficient arrangement for library
use. Here the need for vertical library stacks.and close proximity
to supporting services dictated a more compact shape. Conse-
quently, organization of the scheme became more concentrated
at a single location and the linear structure began to recede (fig-
ure 9, part 4). Further concentration of the building elements fol-
lowed, until a single building emerged (figure 9, part 5). Vestiges
of the linear scheme can still be seen, however, in the manner in
which the building mass is “‘flattened,”’ addressing the drive out
from downtown Chicago to the site.

Having resolved, for the moment, the matter of the dominance of

the two earlier themes, or large organizing principles, the de-
signers then began to explore alternative programmatic arrange-
ments and formal configurations for the center’s building mass.
Here a straightforward proposal of vertical library stacks, in the
form of a tower resting on a base of administrative and other
facilities, began to emerge. In the minds of the designers this
arrangement was both efficient and consistent with the idea of
creating a landmark (figure 9, part 6, and figure 10, part 1).
Throughout these explorations there was a conscious prefer-
ence for adopting a strongly formalistic, symmetrical, and neo-
classical conformation for the building. The only remaining
linear plan element was a road leading from the building, on
axis, toward downtown Chicago.

Other aspects of the facility's program were incorporated into

the design investigations, as much to help resolve choices be-
tween formal alternatives as for the sake of comprehensive

Designers in Action 22
 9
Case study no. 3: investigation of design themes.

Designers in Action 23
 = =e
Air arl (ntimt )

Sena —aaaw)

ie Cie

10
Case study no. 3: development of design concepts.

Designers in Action 24
 study. A myriad of compositional principles were invoked, as
well as a separation of the library proper from the facility hous-
ing the electronic media and transmission devices (figure 10,
part 2). The incipient notion of enforcing a duality between elec-
tronic media and printed matter was soon abandoned, the de-
signers having concluded that this separation ‘‘made the
computing and technical facilities stand out too much’; they
once again retrenched to proposing a single building (figure 10,
part 3).

The next episode of design activity, not shown here, involved

extensive experiments with various kinds of structural arrange-
ments, the accommodation of vertical circulation within the
tower, and refinement in the distribution of programmatic ele-
ments. During this process a square building base with a round
drum of library stacks was established. A distinction was made
between ‘“‘ceremonial” circulation to and through the building
and “expedient” circulation. Geometrically, the square base was
divided, in plan, into nine parts for the symmetry and general
centering effect afforded by such an arrangement. Further sub-
division resulted, albeit in a subtle form, until a “reasonable
structural bay of 27 feet’’ was determined. Throughout this pro-
cess of fitting structure and circulation and use areas together in
plan, the three-dimensional consequences were also explored
for the purposes of making necessary adjustments.

The organization of the facades and general external form of the

building was pursued next, although to some extent these inves-
tigations had been performed concurrently with the studies of
structure and circulation. To start with, the elevation was
geometrically organized, within a square approximately corre-
sponding to the outline of the plan. A vertical division of one-
third for the base and two-thirds for the tower was adopted
(figure 10, parts 4 and 5), at least provisionally, as representing
the ‘“‘right’’ proportions for the building mass. During this pro-
cess the facility for the electronic media returned, again as a
conspicuous (linear) element in the overall composition (figure
10, part 5). It was later relegated to a more minor role, although
now its location became fixed on the minor axis through the li-
brary facility (see figure 11, part 2). Exploitation of geometrical
characteristics, either seen to be inherent in the square building
form or deriving from the plan configuration, were finally aban-
doned as arbitrary. Instead a more straightforward articulation of
entrance, reading rooms, and library stacks was adopted (figure
10, part 6).

Designers in Action 25
11
Case study no. 3: final drawings.

Designers in Action 26
 At this stage a number of sketches were made, as the designers
put it, ‘‘to examine the overall effect of the proposed scheme”
(figure 11, part 1). As a consequence, the theaters, formerly part
of the building base, were placed underground, on the pretext
that their distinctive shape detracted from the clarity of the over-
all composition. This placement also allowed the profile of the
shoreline to be shaped in a circular manner around the building
base (figure 11, part 2).

Finally, the proposal was delineated in more detail, as shown in

parts 2 through 6 of figure 11. The earlier concept of a grid pat-
tern extending into the lake returned in the guise of parkland
with reserve tracts for later development (figure 11, part 2). The
minor axis through the main library building was used to orga-
nize the location of the computer facility and the entries to
the theaters (figure 11, part 3). The earlier formal and figural
preoccupations on the part of the designers are evident in the
resulting elevational and sectional treatments (figure 11, parts
4and 5).

Perhaps the most distinctive feature of the protocol is the atten-

tion paid by the designers to the two large themes of creating a
focal point, or landmark, and extending the grid pattern of
Chicago, in a linear fashion, out into the lake. Throughout, these
two themes seem almost to compete with one another. First one
dominates, only to recede again as the process unfolds. In the
end, design effort was focused on the proposal of a single land-
mark building, although even then its immediate site environs
were clearly controlled by the idea of the grid pattern. An ex-
plicit concern with context is another hallmark of the case study.
Here homage is paid to Burnham’s earlier plan and to the build-
ings of the Chicago Exposition, as well as the general layout of
contemporary Chicago.

Other Accounts

The historical record provides few case studies in similar detail

involving well-known designers in action. This is to be expected,
given the rather private nature of the activity and the ‘‘messy”’
quality of early speculations when compared to the finished de-
signs. As Robert Venturi remarked on some recent work in prog-
ress, ‘One ought not to watch the sausage being made at the
factory’ (Holmes 1985).

Designers in Action 27
 Still, there are a variety of sources from which we can gain
insights into the act of designing that range from spasmodic
glimpses to reasonably thorough reconstructions. We have, for
instance, autobiographical accounts from major designers that
at least provide some outline of situational and experiential
factors that have shaped their output. Frank Lloyd Wright's
own account of his early years, formative experiences, and
preoccupations during various projects is a case in point (Wright
1943). In a less personal vein, designers have offered retro-
spective analyses of projects on which they worked. At times
these are more than rationalizations for the purposes of ex-
plaining and justifying a scheme and deal with what really hap-
pened—with the inner conflicts inherent in their particular
design philosophies. In Suckle’s recent By Their Own Design, for
example, architects openly discuss their process of design and
construction (Suckle 1980). We also have public presentations
and archival material closely chronicling the design thinking that
went into major architectural projects. Finally, we have numer-
ous reconstructions by historians and design critics who have
attempted to piece together the designing of various buildings.

Le Corbusier’s Venice Hospital

One reasonably complete account of prominent designers in
action concerns the scheme by Le Corbusier and Julian de la
Fuente for the Venice Hospital (1964—1967). Undoubtedly this
thoroughness is the result of the project’s having been com-
pleted in Le Corbusier’s atelier after his untimely death, thus
necessitating the explicit documentation of earlier steps and the
disentanglement of the web of design ideas. After all, to faith-
fully pick something up where another has left off requires more
than interpretation ‘“‘in the manner of” that other. It requires sub-
stantiation through reconstruction of the prior thinking.

The primary themes for the project are the city of Venice and,
secondarily, advances in modern medicine, particularly in the
realms of postoperative and outpatient care (Julian de la Fuente
1968). The hospital is viewed less as an instrument for the effi-
cient technical practice and institutional delivery of medicine
than as part of the city, where patients and visitors alike can
partake in something that, as closely as possible under the cir-
cumstances, approximates urban life. Here Le Corbusier is very
explicit. When describing the more public realm of the hospital,
or what he terms the calli and campielli (both words are used by
Italians particularly to describe features of Venice: the narrow
streets and the small squares, respectively), he states that they
not only are “places ... for circulation but are equipped with all

Designers in Action 28
 necessary to permit social life of patients.” ‘These spaces,” he
continues, ‘‘are covered and glazed in such a manner that the
patient would experience the same feeling that he would have in
the city” (Julian de la Fuente 1968, p. 22).

Within this broad and strongly guiding idea, the development of

the scheme starts with a number of observations about the city
of Venice. ‘Venice is formed by her plan of water,” which is
“one of the most favorable sustaining elements for architec-
ture,” as shown in figure 12, part 1 (Julian de la Fuente 1968,
p. 8). From which Le Corbusier becomes more focused and par-
ticular in the orientation of his interpretation. ‘‘What is basic in
Venice,” he states, ‘‘is the ordering of natural and artificial circu-
lation, the pedestrian and the gondola. . . . Venice [is] a tes-
timony to functional precision” (Julian de la Fuente 1968, p. 8).
Further on he describes a revelation that came to him about this
circulation: “‘The pedestrian ways are amazingly economical, all
very efficient. .. . | disentangled the puzzle of the houses, the
networks of pedestrians. ... It is like a cardiac system, pure,
perfect in its workings. Nothing is straight... the plazas are
imposing reservoirs, lakes to accommodate the throngs” (Julian
de la Fuente 1968, p. 9). And so the idea for the hospital's circu-
lation and public spaces came into being.

Early developmental sketches reveal that the scheme had an-

other point of departure as well, deriving from the technical ac-
complishments of modern medicine and health care delivery. To
quote from Le Corbusier’s report, ‘‘The extremely profound and
swift evolution of modern medicine, combined with the growing
obligation to offer equally modern services to the sick, are two
factors, among others, which hasten the obsolescence of exist-
ing hosptials” (Julian de la Fuente 1968, p. 20). In particular, Le
Corbusier and his collaborators were concerned with the inten-
sive care and accommodation of hospitalized patients as well as
with ambulatory and outpatient care and the most appropriate
environment for this service.

A three-level hierarchy of spaces and activity areas was fash-

ioned out of the program of uses (figure 12, part 2). At the most
fundamental level was the unité /it, or ‘bed unit,’”’ sometimes re-
ferred to as /a cellule, ‘‘the cell’ (Julian de la Fuente 1968, p. 21).
This consisted of a private 3-meter by 3-meter space at the dis-
posal of the patient, complete with necessary accommodations,
a hanging garden, and the possibility of altering the ambient
lighting of the space through the use of colored panels (figure
12, part 3). These unités lits, in turn, were arranged into the unité

Designers in Action 29
 (Oe amma ge
all Ahoy cage
fe rd jen ee “p ve

YU, Abs kA
D
Fal ey
pons

12
Le Corbusier and Julian de la Fuente’s Venice
Hospital project—drawings of Venice, plan
developments, and model.

Designers in Action 30
 des soins, or the ‘‘unit of care.’ Here 28 patient spaces were
organized into a single, relatively independent administrative
unit under the supervision of a nursing staff. Finally, the units of
care, together with other service functions, were arranged ac-
cording to the principles of circulation and community spaces
described earlier through the use of the calli and campielli.

In the development stages of the plan, attention seemed to be

focused on the basic patient spaces and their arrangement into
a coherent unit of care. Consequently the overall disposition of
the plan at this stage is regular in its layout (figure 12, part 4).
Further on in the process the plan becomes far more differ-
entiated as specific site requirements begin to make their pres-
ence felt and as the interpretation of the morphology of Venice
comes into play (figure 12, part 5). Instead of having the appear-
ance of happenstance interstitial realms, the public spaces—
‘the domain of the upright man’’—now take on a more positive
and consciously arranged pattern (Julian de la Fuente 1968,
p. 22). Accommodation is also made for future expansion and
functional rearrangement of the facility. The final result is a
building cornplex of considerable spatial variety and systematic
order (figure 12, part 6). It is also a facility that appears to blend
in with the city, not through superficial imitation but through
thorough reckoning with the morphology of surrounding areas,
adapted to different institutional purposes.

The Use of Analogy

As we Saw in the earlier case study examples, initial design
ideas appropriated from outside the immediate context of a spe-
cific problem are often highly influential in the making of design
proposals. Quite often references are made to objects already
within the domain of architecture. On other occasions, however,
an analogy is made with objects and organizational concepts
that are farther afield and outside architecture. Sometimes these
analogies serve a designer’s purposes for more than a single
project and thus become incorporated as a central part of that
individual’s design thinking.

In Suckle’s anthology, John Johanson candidly provides an ac-

count of his basic premises.in-architectural design (Suckle 1980,
pp. 67-77): ‘I have more recently come to see buildings and
building complexes in terms of their parts: that is, individua-
tion,” he explains; he goes on to describe, albeit somewhat in
caricature, a procedure of ‘‘place it [the enclosures], connect it
[provide access], and support it [hold it together structurally].”
The underlying analogy, which gave substance to the Goddard

Designers in Action 31
 Library at Clark University (1968) and the Oklahoma Theater
Center (1970), is drawn from the field of electronic circuitry.
Johanson observes, ‘‘| wanted to borrow the underlying ordering
principles and their systematic logic and use them as a model
for architectural methodology.” The specific relevance of the
analogy to architecture is seen to be sustained through the or-
ganization of three principle elements: the circuit’s chassis,
“representing the structural frame’; the circuit components,
representing ‘‘functional enclosures” in architecture; and the
circuiting system, which represents ‘“‘channels for the circulation
of people and mechanical systems.”’

Setting aside any qualms we might have about the somewhat

mechanistic and strictly functional character of this analogy,
Johanson does seem to use it in order to address several issues
that he sees as important in contemporary building. The first is
the need for ‘‘direct solutions,” for an “economy of means’”’ and
an improvisational approach, or what Johanson calls ‘“‘ad hoc-
ism.” Here the analogy serves him well. After all, in the design
and assembly of electronic circuitry these are among the major
design parameters. The second issue is the impermanence, or
possible impermanence, of a building’s spatial and technical or-
ganization. Again the concept of open-ended planning and the
provision of facilities is well modeled in the analogy of electron-
ics. Third, there is the question of accommodating the obsoles-
cence of certain parts of a building and the various maintenance
and replacement schedules of service components. Here the
“plug in and clip on” aspect of the analogy for interchanging
and rearranging parts is evident. Finally, there is the modern
technological prospect of dynamically changing a building en-
velope to accommodate different environmental conditions or
periodic requirements of use. Again the analogy with electronic
circuitry, at least as far as it goes, serves Johanson well. This is a
technology that has explicitly incorporated notions of planned
adaptation through changes in system configuration and hence
function.

Big Ideas and Operational Constraints

Another aspect of design thinking that was evident in the
foregoing case studies is the tenacity with which designers will
cling to major design ideas and themes in the face of what at
times might seem insurmountable odds. Often the concept the
designer has in mind can only come to fruition if a large number
of apparently countervailing conditions can be surmounted. For
instance, a new building technique might have to be invented, as

Designers in Action 32
 was the case in Utzon’s Sydney Opera House, or antiquated in-
Sstitutional rules and codes might have to be overcome.

Again in Suckle’s anthology, Richard Rogers provides a detailed

and unabashed account of the give and take that took place be-
tween the central design idea and the technical requirements for
the well-known Plateau Beaubourg or Centre Pompidou project
in Paris (Suckle 1980, pp. 107-121). According to Rogers, the
central design idea was the inversion of the traditional notion of
facade and the conception of the project as a ‘‘flexible container
capable of continuously adapting, not only in plan but also in
section and elevation, to whatever needs should arise.” In short,
Plateau Beaubourg was seen as an ‘inside-out building.” To
these ends, ‘‘Beaubourg [was] constructed completely from pre-
fabricated dry elements [and] becomes in essence a gigantic
ever-changing erector set, as opposed to the more common
doll’s house with its precious, nonadditive, tailor-made detail-
ing and its inherent lack of freedom and choice.” Plateau
Beaubourg was to be an “‘architecture of possibilities,” in which
“flexibility should be communicated by the legibility of [the]
building.”’

An enormous amount of effort had to be invested in order to

realize this architecture of possibilities. For instance, within the
time allotted for completion of the project it was impossible to
“debug the initial idea of moving floors held by friction clamps.”’
Consequently the idea, though theoretically feasible, was aban-
doned. The prefabricated nature of all building components
necessitated advances in building technology in a number of
areas, not the least of which was the use of special cast-steel
components (geberettes) for the building’s frame. Furthermore,
the hoped-for use of standardized components added to the
necessity for unusual engineering accomplishment. Consistent
with the theme of an inside-out building, the decision to expose
the mechanical components of the building required a signifi-
cant realignment of traditional attitudes among architects, en-
gineers, and contractors. The appearance of each component
became a paramount consideration, calling for unusual coopera-
tion among all the parties concerned and ultimately for unor-
thodox methods of fabrication and construction.

As Rogers was to observe, however, ‘Probably the major flaw in

the scheme lay in our inability to develop a new technique for
fireproofing structural steel. ... We actually believed that we
would make a major breakthrough and find a system of making
the steel itself incombustible without resorting to the consider-

Designers in Action 33
 able expenditure of creative effort required in the search of ele-
gant ways to ad hoc fireproofing. It didn’t happen.” Instead, the
codes and practices of Parisian bureaucracy prevailed. In order
to bring the building within range of firemen’s ladders, it was
reduced in height by some 19 meters. Further, the allowable fire
compartment size required modifications to the layout of interior
spaces. Finally, a design principle of ‘separate and extinguish”
was invoked, whereby individual components were fireproofed.
For example, trussed beams of steel were wrapped in a blanket
of fibrous material and further encased in stainless steel in order
to achieve the required rating. Indeed, the visual effect of almost
every element in the building had to be thoroughly considered
from this perspective.

Quite apart from its unusual and perhaps visionary appearance,

Plateau Beaubourg stands as a testament both to the design
team’s original ideal and to their technical and administrative
prowess. ‘‘It is impossible,” pronounced Rogers, “‘to divorce the
building from its legal, technical, political, and economic con-
text. At the same time, a major part of any design approach is
the way constraints may be absorbed and whenever possible in-
verted into positive elements.”

ER ee a
Observations and Questions about the
Protocols

From these cases and others that have been documented (Con-
sortium of East Coast Schools of Architecture 1981, Sch6n 1983,
Sch6n 1984), several observations can be made about the nature
of design thinking. To begin with, it is apparent that the un-
folding of the design process assumes a distinctly episodic
structure, which we might characterize as a series of related
skirmishes with various aspects of the problem at hand. Usually
the results of these investigations cohere into a more singular
direction for the design activity, although not necessarily as a
linear progression of reasoning.
This episodic structure manifests itself in a number of ways.
First, there is the ‘‘to and fro’”” movement between areas of con-
cern—a movement perceived at the time by the designers in our
three case studies. In all three protocols there was movement
back and forth between exploration of architectural form and
evaluations of program, structure, and other technical issues.
Second, there seem to be periods of unfettered speculation, fol-
lowed by more sober and contemplative episodes during which

Designers in Action 34
 the designer ‘‘takes stock of the situation.”’ Third, each episode
seems to have a particular orientation that preoccupies the de-
signer. We might say that the organizing principles involved in
each episode take on a life of their own, as the designer be-
comes absorbed in exploring the possibilities that they promise.
Here a ‘‘dialogue’”’ between the designer and the situation is evi-
dent (Sch6n 1983, ch. 3). In our case studies these episodes,
such as the various massing exercises with building volumes,
often became very speculative as the designer ‘‘pressed on,” as
it were, when information from another quarter might have re-
solved the problem at hand more economically. Such situations
often subsequently gave rise to a certain amount of backtrack-
ing, as the designer retrenched to what seemed a more advanta-
geous position. Finally, as the scope of the problem became
more determined and finite for the designer, the episodic
character of the process seems to have become less pro-
nounced. During this period a systematic working out of issues
and conditions took hold within the framework that had been
established. This phenomenon is not at all surprising when we
consider the fundamental difference between moments of prob-
lem solving when matters are poorly defined and those with clar-
ity and sufficiency of structure.

These observations raise questions about the inherent structure

of the episodes, the kind of information that is being included,
and its relationship to the procedures being employed. They also
raise questions about the manner and means by which one epi-
sode gives way to another. These episodes are not happen-
stance events. They possess an interior logic that seems deter-
mined partly by the subject matter at hand and partly by the
organizational procedures being used. They also have a con-
sequential connection with one another. Without such logic and
closure among episodes the emergence of design proposals
would be difficult to imagine.

Within the episodic structure of the process, the problem, as

perceived by the designer, tends to fluctuate from being rather
nebulous to being more specific and well defined. Furthermore,
moments of ‘‘blinding”’ followed by periods of backtracking take
place, where blinding refers to conditions in which obvious con-
nections between various considerations of importance go un-
recognized by a designer (Newell, Shaw, and Simon 1967, pp.
107-108). Sometimes this characteristic seems to have a great
deal to do with the manner in which a designer represents the
problem at hand. For example, in the second case study the de-
signer was preoccupied at one moment with making a formal

Designers in Action 35
 arrangement of the building elements in plan and then suddenly
realized that one of the elements was a very large structure and
not at all in keeping with the plan concept. To remedy such a
situation, designers invariably seem to return to an earlier point
of departure—they backtrack. We saw this in the third protocol
when the designers pursued the linear plan form of the complex
to what they regarded as an illogical conclusion and then turned
back to considering the buildings as a more concentrated formal
composition.

By contrast, during moments of clear problem definition more

straightforward procedures are used. For example, in our case
studies there were episodes in which variations of a particular
organizing principle were systematically explored, immediately
followed by an evaluation of their relative success or appropri-
ateness. On other occasions, although the desired result was
clear, the precise means for achieving that end remained a mat-
ter for consideration. When difficulties were encountered, de-
signers resorted to various forms of reconsideration, although
rarely did they attempt to break down the conditions at hand into
their basic components.

Having observed these phenomena, can we identify and define

certain generic problem-solving procedures? Are these proce-
dures exclusive to design, or can they also be ascribed to other
forms of creative problem-solving activity? Consider the domi-
nant influence that is exerted by initiai design ideas on subse-
quent problem-solving directions—another prominent feature
revealed by the protocols. Designers inevitably bring certain or-
ganizing principles to a problem at the outset. Even when severe
problems are encountered, a considerable effort is made to
make the initial idea work, rather than to stand back and adopt a
fresh point of departure. In the third case study, on the site in
Chicago, this phenomenon was particularly evident in the persis-
tent competition that took place between the two underlying de-
sign themes. We can also see some very real distortions of
otherwise systematic procedures in these attempts to adhere to
the “big idea.’’ For example, there were the instances of design-
ers seeming to cram the building program into the architectural
object that they were shaping, and the attempts to conform
otherwise straightforward structural systems to architectural in-
tentions about movement and entrance. Sometimes it was as if
the inherent relational structure of subproblems took on an en-
tirely malleable quality in the face of other design prejudices.
The persistence and technical virtuosity with which Rogers ap-
proached the Pompidou Center also had these hallmarks.

Designers in Action 36
 Clearly 2 distinction can be made between the constraints and
opportunities that are found in the problem conditions as given
and the enabling prejudices’ that designers bring to bear on the
situation. Moreover, the protocols discussed in this chapter
showed interpretation of the context solely as given to be exer-
cising a subordinate influence over decision making, compared
to ideas and references supplied by the designers from else-
where. This phenomenon was particularly evident during the
early stages of the projects, as the designers searched for con-
cepts around which to construct frameworks for reinterpreting
the design problem. These concepts or hunches were to provide
insight and direction for further information processing.

What, we must ask next, is the source of these design ideas?

Why is it that ideas from outside the immediate problem context
are so influential? Without a doubt design is to be seen as a
normative enterprise; the resulting proposals are about what is
proper. Even if the presence of such normative reasoning is
clear, however, its role and character remain vague. What qual-
ities snould we look for in a normative position about architec-
ture, and how might we adjudicate among positions that
compete for our attention? In addition, what is the structure of
the interplay between normative design ideas and procedural
aspects of design thinking?

Finally, it should be noted that the ‘style’ of the completed proj-

ects seems to have been determined primarily by two factors.
First, there was the sustained influence of initial design ideas in
the form of organizing principles, rules, and references. Second,
there was the influence exerted by the particular structure of the
problem-solving process itself. In this iatter regard, the sequence
in which design principles were applied seems to have mattered
the most. Contrary to sore earlier accounts by so-called design
methodologists, the kind of theory we need if we are to explain
what is going on when we design must go beyond matters of
procedure. We need to move directly into the realm of normative
discourse about what constitutes architecture and urban design
in order to clarify both the inherent nature of the enterprise and
the direction in which procedures are inclined. Futhermore, we
need to ask how this discourse is shaped and what frames of
reference are appropriated in its shaping. Definitions of prob-
lern-solving procedures are not enough. We must also explore
what we can say about the service in which these procedures are
placed.

Desires in Action 37
Procedural Aspects of Design Thinking
 Beneath the surface irregularities of designers’ modes of opera-
tion, common information-handling procedures can be iden-
tified. Furthermore, theoretical developments in cognitively
based interpretations of creative problem solving have provided
us with tools for analyzing such procedures (Hayes 1978, Hunt
1982). In spite of variation, these interpretations all seek to ex-
plain creative problem solving under the conditions of bounded
rationality that are characteristic of design. Here bounded ration-
ality refers to the concept that human problem solvers are rarely
in a position to identify all possible solutions to the problem at
hand and therefore settle for choices that seem to satisfy the
required solution properties of a problem, as they see them at
the time. Generally, they make decisions that might otherwise be
seen as suboptimal, or what Simon refers to as ‘‘satisficing”’
(Simon 1957; Simon 1969, pp. 64-76; Simon 1979, p. 3). This
condition certainly holds for most design exercises, at least of
the kind involving invention and novelty, with which we have a
primary interest. As we shall see in more detail later on, the lat-
ter are invariably problems whose solutions are neither strictly
true nor false.

Before continuing, it is well to address a point that usually arises

as soon as the so-called creative aspect of design is introduced
into a discussion of problem solving. After all, some might main-
tain, design is much more than mere problem solving. The
veracity of this observation, however, clearly depends on our
understanding of the word problem. To paraphrase Thorndike’s
venerable definition, a problem can be said to exist if an organ-
ism wants something but the actions necessary to obtain it are
not immediately obvious (Thorndike 1931). It is hard to imagine
circumstances under which the impetus for design is not cov-
ered by this definition, which subsumes both problems that are
predefined and placed at our doorstep and those that are
brought into our consciousness de novo, requiring definition
and redefinition.

LR Lae 2 ne
Some General Characteristics of
Design Problems
In the world of design problems, a distinction can be made be-
tween those that are well defined and those that are i// defined.
In the latter category further distinctions can be drawn, resulting
in the subclass wicked problems (Churchman 1967).
 @ Well-Defined Problems
Well-defined problems are those for which the ends, or goals,
are already prescribed and apparent; their solution requires the
provision of appropriate means (Newell, Shaw, and Simon 1967,
p. 70). Rittel, who refers to this class of problems as ‘‘tame,”’
states that they can be ‘‘exhaustively formulated .. . and solved
by a knowledgeable man without the need for further informa-
tion’ (Rittel 1972, p. 392). A common example is the solution of
two algebraic equations with two unknown values. Here the aim
of the exercise is to find the values for x and y, or some other
similar designation of variables. The solution requires applica-
tion of the rules of algebra to the specific equation structure that
is given.

This class of problems can be specified under the rubric ‘Given

a set P of elements, find a subset S of P having specified proper-
ties’’ (Newell, Shaw, and Simon 1967, p. 70). Other common and
not-so-common examples of this class are crossword puzzles,
finding the combinations to safes, and making moves in check-
ers and chess. In architecture and urban design the class would
include the space-planning problem, in which a set of building
spaces is prescribed, together with a site in which they are to be
assembled and some expression of adjacency requirements
among the spaces. Here, the problem can be formally expressed
by letting P represent all possible combinations of building
spaces, S;,...,S,, with a subset S of combinations that satisfy
the adjacency requirements, A(s;,s;). The object is to find a satis-
factory combination, S. Other problems, such as those involving
the compositional arrangemeni of building elements in design or
the allocation of natural resources in planning, may also have
sufficient clarity in their goals to be formulated in a similar man-
ner. It should be understood, however, that when actually
confronting even these problems designers must define and
redefine them in a manner that is sufficient for the proposal of a
solution.

@ Ill-Defined Problems
For ill-defined problems, on the other hand, both the ends and
the means of solution are unknown at the outset of the problem:
solving exercise, at least in their entirety (Newell, Shaw, and
Simon 1967, p. 71; Bazjanac 1974, p. 8). Most architecture and
urban design problems are of this type. A client or a citizen
group comes to the designer with the desire to build a house or
improve the quality of a neighborhood. Although the general
thrust of the problem may be clear, considerable time and effort
are usually spent with the client in order to clarify what is re-

Procedural Aspects of Design Thinking 40

 quired. A large part of the problem-solving activity, then, con-
sists of problem definition and redefinition.

@ Wicked Problems
Many design problems are so ill-defined that they can only be
called wicked problems (Churchman 1967, Rittel 1972, Bazjanac
1974). This class has, among others, the following characteris-
tics that are most pertinent to our discussion. First, they are
problems without a definitive formulation, or indeed the very
possibility of becoming fully defined. Additional questions can
always be asked, leading to continual reformulation. Second, as
a corollary to this first characteristic, they are problems with no
explicit basis for the termination of problem-solving activity—no
stopping rule. Any time a solution is proposed, it can, at least to
some significant extent, be developed still further. Third, differ-
ing formulations of the problems of this class imply differing
solutions, and vice versa. In other words, the problem’s formula-
tion depends on a preconception that, in turn, implies a definite
direction toward the problem’s solution. Finally, solutions that
are proposed are not necessarily correct or incorrect. Plausible
alternative solutions can always be provided. This characteristic
follows logically from the first property—the impossibility of de-
finitive formulation. Reformulation can take place beyond the
realm of considerations within which the original proposals were
made, thereby opening up avenues of approach to other solu-
tions.'
Such distinctions were not always a part of our understanding of
problem-solving activity. In fact, as we shall see, earlier theories
made few if any references to the kinds of problems confronting
designers.

eeae ee
Early Theoretical Positions
Without venturing too far back into history, two distinct themes
can be seen at work in the development of theory about problem
solving. Beginning toward the end of the nineteenth century with
associationism, a mechanistic type of doctrine can be observed
to recur that sought to explain problem-solving behavior
through the use of irreducible lawlike relationships deemed to
govern mental processes. By contrast, other efforts were made
to explain problem solving in more behavioral and nonmental-
istic terms.

Procedural Aspects of Design Thinking 41

 @ Associationism
At the turn of the century associationism was the prevailing doc-
trine in most quarters concerned with theoretical speculation
about problem solving. It held that the sole mechanism of hu-
man learning consisted in the permanent association of impres-
sions that had been repeatedly presented to the senses in
combination.” The associationist view of creative problem solv-
ing was both atomistic and mechanistic. It was atomistic in the
sense that it postulated that ideas took the form of elements,
analogous to basic physical entities, and that these elements
were hooked together to form thoughts or insights about prob-
lems. It was mechanistic in the sense that simple laws of conti-
guity, again based on models of atomic structure in the physical
world, were used to account for the association of elements—
ideas—to form thoughts (Newell, Shaw, and Simon 1957).

Under the doctrine, creative problem solving was seen to pro-

ceed as a stream of associations, where each association pro-
duced successive new attachments and thus new insights about
problems. The source of these new associations, or attachments,
was experience of the environment external to the mind. The
doctrine thus embodied an empiricist view of the mind as a
tabula rasa upon which experience was recorded. In the final
analysis, creativity was seen to be largely a matter of happen-
stance, a random kind of event.

Around 1900 a controversy developed between two opposing

camps of experimental psychologists (Boring 1950, chs. 17-19).
On one side were the adherents of many of the basic tenets of
associationism, such as Wilhelm Wundt, who maintained that
mental images, sensations, and feelings were a necessary part of
understanding and learning.* For instance, after its introduction
into our language a word will call up a picture in the mind of
what it stands for. Having this image is what understanding the
word amounts to (Bloor 1983, pp. 7-10). On the other side were
those, like the Austrian Brentano, who maintained a psychologi-
cal theory that stressed the role of mental acts such as intending
and focusing attention. For them a word, to use the same ex-
ample, has meaning not because it is accompanied by a mental
image but because it is accompanied by a mental act (Bloor
1983, p. 9).

@ The Wiirzburg School

During the first decade of this century, speculation from the
Wurzburg School in southern Germany, principally under Kulpe,
Ach, and Buhler, began to replace the doctrine of associa-

Procedural Aspects of Design Thinking 42

 tionism (Humphrey 1963, chs. 3-4). Although originally follow-
ers of Wundt, this group, using extended introspective methods,
produced results that suggested that subjects given complex
tasks such as translation, mental arithmetic, and problem solv-
ing arrived at answers without any accompanying imagery or
other sensations. Thus they tended to side with Brentano and
the ‘‘act theorists.’
From these findings a number of new and useful concepts
emerged. Instead of the association of ideas, the Aufgabe
(“task” and hence ‘‘determining tendency’) was seen to be the
controlling mechanism in problem-solving behavior (Humphrey
1963). Here a task gave direction to thought, especially in prob-
lem-solving situations. Creative problem solving was thus re-
garded as essentially purposeful and, therefere, controlled in a
manner that was far more circumscribed than by pure chance.

Another major technical contribution of the Wurzburg School

was the use of systematic introspection as a means for describ-
ing problem-solving behavior. Subjects were asked to recon-
struct their sequence of thoughts during some problem-solving
exercise, while the experience remained fresh in their memories.
This kind of approach toward data gathering gained wide adher-
ence during subsequent technical developments of the field.

@ The Gestalt Movement

The Gestalt movement in psychology, beginning in the 1920s
with the work of Kohler, Koffka, Wertheimer, and others, has
made a wide variety of contributions to our understanding of
human thinking, not the least of which is their celebrated work
on visual perception (Kéhler 1929, Koffka 1935, Wertheimer
1945). In the specific realm of our interest, proponents of the
movement rejected the mechanistic doctrine of associationism,
although they maintained the central value of phenomenal ob-
servation of environmental stimuli in explaining problem-solving
behavior. An idea of holistic principles for organizing informa-
tion, embodied in the concept Gesta/ten, replaced the discrete
mechanistic view of prior positions.

Speculation and experimentation within the Gestalt movement

also extended the Wurzburg School’s idea that creative problem
solving was directed behavior: from the notion of the Aufgabe
developed the notion of the set, or what became more loosely
referred to as the mind set of the problem solver in confronting a
task. Clearly, this extension moved the controlling influence on
the mental processes associated with probiem-solving behavior
beyond a singular focus on the task at hand by admitting other

Procedural Aspects of Design Thinking 43

 kinds of influences, say from related experiences, that were gen-
erally less intrinsic to the particular task situation.

A logical progression of this concept was the mechanism of the

schemata, or organizational frameworks for structuring informa-
tion. In the case of visual perception, these frameworks were
overlaid on external visual stimuli, vastly facilitating the organi-
zation of these sense data into meaningful information about the
source of the stimuli, namely the object being perceived. Among
others, Bartlett sought to explain creative thinking in terms of
the idea of schemata, suggesting that there are certain fixed ar-
rangements within the brain, strongly associated with past re-
sponses to general types of stimuli or cognitive experiences
(Bartlett 1961, Broadbent 1966). To Bartlett, imagination, or cre-
ative insight into a problem, consisted of free construction on
these fixed arrangements or schemata. Once again we can see
that the associationists’ reliance on random mechanisms was
largely rejected and that the Wurzburg School's concept of Auf-
gabe was generalized into the realm of prior experience. We can
also see that this position remains fundamentally mecha-
nistic and mentalistic in its doctrine.
@ Behaviorism
The behaviorist position began as a reaction to what proponents
termed the mentalism of earlier doctrines. It was a fundamental
rejection of all attempts to study inner mental processes in
which distinctions were made between a concept of mind anda
concept of body. Instead, the behaviorists postulated that human
behavior, including problem solving, could only be adequately
explained in nonmentalistic, concrete terms. By concrete terms
they meant observable, measurable, and replicable patterns of
physical behavior.

Investigations within the position quickly gave rise to the now

familiar stimulus-response, or S-R, models of behavior, founded
on the assumption that given a particular external stimulus, one
could predict a certain response with complete assurance. From
the standpoint of the working methods of the behaviorists, men-
tal processes didn’t matter. Questions of the ‘‘mind”’ were irrele-
vant. Behaviorism consequently rejected the Wurzburg School's
use of introspection as unstable and thus suspect, although ad-
herents were strongly wedded to the use of experimental tech-
niques. Instead, they embarked upon a far-reaching program of
correlating environmental stimuli with what they could docu-
ment as consistent patterns of behavioral response.

Procedural Aspects of Design Thinking 44

 The position took strong hold in psychology and related disci-
plines, especially in the United States after Watson returned
from Europe in 1913, bringing the central ideas out of the labo-
ratories of Pavlov and the like (Watson 1930 [1924]). The move-
ment reached its apogee in the work of Skinner during the 1930s
and 1940s and became generalized, even popularized, into the
idea that behavioral malfunctions and pathologies could be
cured by appropriate environmental modification (Skinner 1953).

In the more specific realm of problem-solving theory, the behav-

iorist position gave rise to, or certainly supported, the develop-
ment of phase, or rigid-state, models of creative problem-solving
behavior (Arieti 1976, ch. 2; Ghiselin 1952; Gregory 1966). Such
behavior was widely acknowledged as conforming to an epi-
sodic process consisting of distinct and almost discontinuous
phases of activity. The aim behind the models was to identify
and describe each phase and the sequence of phases.

Although a number of variations were developed, each model in

One way or another incorporated four fundamental steps, or dis-
criminable phases of activity. They were (1) preparation for the
task or situation at hand, (2) incubation, (3) illumination or inspi-
ration, and (4) verification, involving the testing of proposed so-
lutions. In the structure of the models the illumination phase
occurs when the problem solver becomes aware of a potential
solution to a problem. Such awareness can arise as a sudden
intuition, a flash of insight—the so-called ‘Eureka!’ phenome-
non. It can also transpire as a result of sustained efforts at explo-
ration involving the more systematic pursuit of various hunches.
Furthermore, the awareness of a solution may not be complete;
sometimes it is merely a glimmer from a propitious direction in
which a more fully developed solution might be sought. The in-
cubation phase of activity was usualiy defined simply as the pe-
riod of preparation for illumination. It is the stage in which
thoughts about the situation or task are allowed comparatively
free rein. Clearly, the steps in the process could be repeated un-
til all aspects of the problem at hand were addressed. Within the
models’ structure, however, the progressive relationship among
the steps remained immutable.

This general view of problem-solving activity dominated the liter-

ature in the field, at least within the United States, for several
decades, roughly until the end of the 1950s. We can see that the
theoretical enterprise was strongly based on the observation of
human behavior. There was no attempt to venture into the realm
of mental processes. In fact, as mentioned earlier, the very possi-

Procedural Aspects of Design Thinking 45

 bility of making such an excursion was antithetical to the behav-
iorist doctrine.

At the same time there were, of course, some exceptions to this

position. Koestler’s speculations about creative thinking arising
from the ‘‘bisociation of two mutually incompatible contexts”
hark back to the Gestaltist idea of schemata. To summarize
Koestler’s view: (1) normal thought proceeds within a frame of
reference, associative context, or type of logic; (2) in normal per-
sonal dealings we operate within many such frames of reference,
but only one at a time; and (3) creating involves relating two
normally independent frames of reference, or in Koestler’s terms
the ‘‘bisociation of matrices’ (Koestler 1964, Perkins 1981).’
Bruner’s use of ‘‘puzzle forms” in attempting to explain prob-
lem-solving behavior has similar hallmarks. Here the problem
solver, when confronted with a new and yet unsolved problem,
overlays the structure of the unsolved problem with an appar-
ently similar problem with which he or she is experienced
(Bruner 1961; Bruner, Goodnow, and Austin 1967). Finally, Gor-
don’s “‘synectic theory” and its associated techniques for at-
tempting to enhance creativity by ‘‘making the strange familiar
and the familiar strange” are also principally based on the use of
analogy. In certain other respects, however, his position gener-
ally acknowledged a phased, episodic structure for creative be-
havior (Gordon 1961).

Paes densa oie moma kl Seed ae gens

Staged-Process Models of Problem
Solving in Design

In explorations that seem to have drawn heavily on rigid-state

models from the behaviorist doctrine, a number of significant
contributions were made that are pertinent to design. During the
late 1950s and the 1960s, attempts were made to describe the
creative problem-solving process at work in design by way of the
logical structure of overt activities that appears to take place. In
other words, design was regarded as a series of stages charac-
terized by dominant forms of activity, such as analysis, synthesis
evaluation, and so on.

Evidence of this kind of concept about design can also be found

within the tradition of architecture in the form of project organi-
zation and pedagogical doctrines that evolved in the ateliers of
the Ecole des Beaux Arts and the Ecole Polytechnique during
the eighteenth and nineteenth centuries (Egber 1980, Carlhian

Procedural Aspects of Design Thinking 46

 1979). Here a rigidly formalized staging of work activities was
imposed on the students by their masters during the conduct of
an esquisse. In the teachings of Blondel this sequence took the
form of a systematic reading and interpretation of the program,
followed by exploration of ways in which the program could be
met, culminating in a parti. Subsequent phases were directed
toward elaboration and detailed presentation of this parti in
plan, section, and elevation. Throughout, each episode of
activity was carefully monitored. For our purposes, however, the
work of Morris Asimow, an industrial engineer prominent during
the 1950s and 1960s, will serve to introduce and illustrate the
theoretical contributions of this kind.

@ Asimow’s Model
In a text entitled /ntroduction to Design, Asimow distinguished
two structures in the design process: a vertical structure involv-
ing a sequential phasing of activities, and a horizontal structure
in the form of a decision-making cycle common to all phases
(Asimow 1962). The chronological sequence of steps, or phases,
in the vertical structure proceeded from a definition of need,
through feasibility study, preliminary design, detaiied design,
production planning, and finally production itself. Furthermore,
within each design phase there was a sequence: preparation for
design, design of subsystems, and so on. Overall, the general
process, or sequence of activities, was seen by Asimow to ad-
vance from abstract considerations to those that are more con-
crete and particular. Numerous feedback loops—relationships
between phases along which information about the design situa-
tion was seen to flow—were incorporated to account for the ob-
servable tracing back through the process in order to respond to
new information or difficuities.

Asimow represented the horizontal sequence as a cycle that be-

gan with analysis and proceeded through synthesis and evalua-
tion to communication (Asimow 1962, ch. 3). He saw this cycle
as repetitious, or iterative, both within and between the various
phases of activity. Parenthetically, Asimow’s speculation about
the structure of design activity is roughly congruent with the
“iconic model” proposed in various forms by Mesarovic and
others, as shown in figure 13 (Watts 1966, Mesarovic 1964).
Throughout this kind of account runs the assumption that it is
possible to discriminate distinct phases of activity and, further-
more, that such distinctions have relevance to our understand-
ing of design. Here we can see the influence of behaviorist
doctrine. In addition, the account exhibits a strongly determin-
istic posture. The very maintenance of distinct phases of activity,

Procedural Aspects of Design Thinking 47

 =~
Communication eS S Concrete

E
(Host Environment) Abstract

Analysis ——> Synthesis ——>~ Evaluation ——> Communication

13
An iconic model of a design process.

with a beginning and an end, and with feedback loops among

them, requires that objective performance criteria can be explic-
itly stated in a manner that fundamentally guides the procedure.
Moreover, there is a strong implication that the eventual synthe-
sis of information in the form of some designed object follows in
a straightforward fashion from an analysis of the problem at
hand together with likely performance criteria. Therefore, once a
problem has been defined, its solution is made directly accessi-
ble in terms of that definition. In many ways this is also a
prevalent view in ‘‘operations-research”’ circles, in which prob-
lem-solving activity is defined through the application of techni-
cal procedures, largely of a mathematical kind, that have been
developed to solve general classes of problems.

@ Models of the Hochschule fiir Gestaltung at Ulm, Archer, and Others

Similar propositions were advanced during the late 1950s and
early 1960s by designers and theoreticians at the Hochschule fur
Gestaltung at Ulm in southern Germany (Maldonado and Bon-
siepe 1964). In the work of Hans Gugelot, Tomas Maldonado,
and others, various staged-process models of design were pro-
posed and used as a basis for design education and product
designs for clients like Lufthansa and Braun (Maldonado 1972,
Broadbent 1973, ch. 13). Furthermore, here such speculation
moved beyond description and explanation of design behavior

Procedural Aspects of Design Thinking 48

 into the realm of idealization. Not only was the possibility of a
“scientific” and totally objective approach toward design seri-
ously entertained, it became a goal in itself. A confident sense of
rational determinism prevailed; the whole process of design, it
was believed, could be clearly and explicitly stated, relevant
data gathered, parameters established, and an ideal artifact
produced.

Advocates of the position were also very mindful of the social

and political consequences of such an enterprise in allowing the
esoteric and, to them, heretofore subjective realm of design to
be emancipated and made accessible to many. The Ulmers’ ex-
periment proved short-lived, however: as Frampton’s account
makes clear, unsuccessful reckonings with both the technical
and the sociopolitical issues involved in their approach con-
spired to force members of the Hochschule to disband the in-
stitution on the nineteenth of February 1968 (Frampton 1974).

Bruce Archer, an industrial designer from the Hochschule fur

Gestaltung at Ulm and the Royal College of Art in London, also
proposed an “operational” model of design, although in slightly
different terms (Archer 1963-1964). In a simplified form, Ar-
cher’s model is schematically represented by figure 14. Here
again design is seen as a sequence of activities defined by their
orientation and by the general type of task involved. Whether
intentional or not, strains of Aufgabe can be discerned. Further-
more, the process can be described in a general form, irre-
spective of particular circumstances. Feedback ioops, or rela-
tionships between activities, are more in evidence than in pre-
ceding models, with the result that the staging of activities is
perhaps less discretely defined. With the enumeration of three
interrelated realms for the process, namely external representa-
tion, process of activities, and the problem solver, a distinction
begins to be made between overt behavior and the cognitive
realm—a departure from the behaviorist position. The emphasis,
however, is still on the sequence of activities and on the behav-
ioral realm.
Among similar proposals by other designers and theoreticians,
the work of Denis Thornley at the University of Manchester de-
serves mention. His model, which clarifies the design process for
educational purposes, was incorporated among the professional
practices of the Royal Institute of British Architects (Thorniey
1963).

Procedural Aspects of Design Thinking 49

 Training

Analytical Brief Programming Experience Observation

Phase Measurement
Inductive
Data Collection
Reasoning

Analysis Evaluation
Judgment
Creative
Deductive
Phase Synthesis Reasoning
Decision
Development

Executive Solution
|
Communication
Description
Translation
Phase Transmission

14
Archer's model of the stages of a design process.

PDUs Sn nis RNAi titiue easraneoene eae uate eliocsceeriee |

The Information Processing Theory of
Problem Solving

From about the 1930s onward, there accumulated a body of ex-

perimental evidence that behaviorists found increasingly difficult
to explain (Hunt 1982). For example, results from maze experi-
ments involving rats and other animals—a popular empirical
technique of the behaviorists for testing the stimulus-response
mechanism—suggested that the animal subject paused at the
choice points along the maze’s pathways before making deci-
sions about which way to travel. Under the behaviorist paradigm,
the animals’ conditioning should have led to a straightforward,
uninterrupted traverse through the maze. It was concluded by
Tolman and other experimenters that the subjects were going
through a vicarious process of trial and error as they attempted
to traverse the maze; they must, therefore, hold a mental map or
picture of the maze (Tolman 1938, Hunt 1982). This finding
seemed to call for a cognitively based explanation rather than
one that was purely behavioral.

Questions also arose with reference to the phase or staged-

process models. What is it about a definite progression of activi-

Procedural Aspects of Design Thinking 50

 ties that automatically results in a specific, further activity?
Through what mechanism or means do we advance from analysis
to synthesis? How is it that unique solutions are often rendered
to problems, when the information processing that takes place
seems so straightforward? As far as they go, staged-process
models do illuminate certain commonly observable features of
design activity; yet the illumination is at a comparatively low
level.

In 1957 Newell, Shaw, and Simon published a paper entitled

“Elements of a Theory of Problem Solving” that ushered in quite
a different line of explanation. This position quickly became
known as the information processing theory of problem solving
(Newell, Shaw, and Simon 1957, 1967). Instead of regarding the
cognitive realm as eluding analysis and therefore irrelevant, pro-
ponents of the new theoretical perspective sought to explain
problem-solving behavior by way of basic information processes.
They maintained that an adequate explanation of observed hu-
man behavior can be provided by a “‘program’’ of primitive infor-
mation processes that account for the cognition associated with
an action. In other words, the primacy of essentially cognitive
processes in explaining problem-solving behavior was
reasserted.

The research agenda of this theoretical position was the descrip-

tion and analysis of human cognitive processes, with the aim of
explaining a wide variety of observable activities, including cre-
ative problem solving (Hayes 1978, Newell and Simon 1972, Si-
mon 1979). Proponents couched their explanations in terms of a
small, finite number of basic mechanisms for processing infor-
mation, mechanisms that could be grouped or arranged into
strategies, or programs in the computer sense, that allowed
complex problems to be solved. Considerable emphasis was
also placed on experimental evidence gained from the analysis
of step-by-step narrations, or protocols, provided by problem
solvers about their own behavior (Hayes 1981, pp. 51—57).

@ Postulates and Directions

The following postulates emerged from Newell, Shaw, and Si-
mon’s early work and became central to the information pro-
cessing position (Newell, Shaw, and Simon 1967, pp. 63-75).
First, there is a problem space whose elements are knowledge
states, some of which represent solutions to a problem. Second,
there are one or more generative processes, or operations, that
allow one to take knowledge states as input, or as starting posi-
tions, and produce new knowledge states as output. In other

Procedural Aspects of Design Thinking 51

 words, the problem space, composed of knowledge states, is
transformed during the course of problem-solving events. Third,
there are one or more test procedures that allow the problem
solver to compare those knowledge states that are presumed to
incorporate solution properties with a specification of the solu-
tion state that is not the knowledge state itself. Test procedures
are also assumed to exist for comparing parts of knowledge
states to detect differences among them. Finally, there are fur-
ther processes enabling a problem solver to decide which
generative processes and which test procedures to employ, on
the basis of the information contained in available knowledge
states.

Perhaps these postulates will become clearer when connected

to the following general definition of a problem (Newell, Shaw,
and Simon 1967, p. 70): “Given a set P of elements, find a subset
S of P, having specified properties.” Here the elements of both P
and S are knowledge states, and their arrangement in P repre-
sents the problem space. Finding subset S among P is thus a
matter of moving, by way of generating potential solutions, from
one knowledge state to another, until one having the specified
properties is found or developed. Throughout this procedure the
direction of search is governed by information gained from com-
parison of the properties of some knowledge state S; to those of
S and by the use of this information in making a selection from
among a range of potential generative processes.

A commonplace example can be found in the manner in which

we solve word-arithmetic problems of the type ‘‘John is twenty-
four years old. His brother was half his age fourteen years ago.
How old is his brother?’ Here the generative processes are es-
sentially algebraic, allowing symbols representing substantive
aspects of the problem to manipulated and transposed accord-
ing to acceptable rules or practices (Eastman 1969). Test proce-
dures, in the form of questions about relationships among
various aspects of a problem under consideration, allow prog-
ress toward a solution to be ascertained. Particular sequences
and specific choices from among the generative processes and
evaluative procedures that are available to a problem solver are
also important, for they represent the overriding means by which
a solution can be sought. In the course of post-hoc analysis, they
also allow particular approaches to be distinguished, even for
relatively simple problems. As we are well aware from everyday
experience, essentially the same solution to a problem may be
arrived at from a number of directions involving a variety of
means. Other commonplace examples can be found in doing

Procedural Aspects of Design Thinking 52

 ©

2 0 N = Nodes
0 O

L My @
aethiys
©
Cate
O © ca
Na, (is @ Decision
O O Sequence

O ©)

Nb> O
@) @

©)
15
A general diagram of a decision tree.

crossword puzzles, arranging furniture in a room, and even such

extemporaneous exercises as cooking.

Formal representation of problem space and of knowledge

states is often achieved through the device of a decision tree,
through which a search for a solution is conducted. In figure 15
the nodes Na,, Nb;, Nbz, and so on, represent points at which
decisions must be made, or decision points, and the links L,, Lo,
and so on, represent branches—the courses of action associ-
ated with different outcomes at each decision point. A particular
problem-solving approach, or protocol, is thus specified by a
particular sequence of nodes and links beginning on the left-
hand side of the tree and moving toward the right-hand side.

A simple example from the field of transportation planning will

serve to illustrate this conceptual scheme. Consider the various
choices involved when we contemplate making a trip to work
with the aim of saving time and expense. First, we may decide
whether or not to go to work at all, although presumably our
state of health, the traffic conditions, or some other fundamental
aspect of the trip would have to be very poor indeed to prompt a
negative outcome. Second, with the approval of our employer,
we may decide whether or not to travel during the “peak hour.”
Third, depending on a variety of factors such as the availability
of vehicles and of mass-transit service, relative cost, and waiting
times, we may choose to take our car, travel with a friend, take a
bus, or walk to work. Finally, again within the means available to
us, we may choose one route over others. Using a decision tree,
the choices associated with a trip to work can be arrayed as

Procedural Aspects of Design Thinking 53

 Route b

{Frequency | Time — Mode vl Route |

|

16
A decision tree of typical transportation choices.

shown in figure 16. The possibilities among the general classes

of decision about frequency, time of day, mode of travel, and
route of travel may be represented by as many as 12 nodes, or
decision points, and 28 branches, or courses of action. Clearly
the comparative advantages—measured, say, in terms of time
and expense—for each combination of the available courses of
action could be computed, and the one or ones that meet the
criterion of minimizing time and expense could be identified. It is
also apparent that when we make this kind of decision in actual
practice we seldom, if ever, systematically evaluate the out-
comes of all possible decision sequences. Rather, we choose
combinations that are habitual, seem to offer satisfactory results,
or perhaps seem to involve the least amount of risk. Further-
more, we might combine the decision components (frequency,
time, mode, route) in a different order from the one represented
here, say from initiation with what appears to us to be the most
critical (for instance, availabililty of modes of transportation),
and never really move beyond this node of the tree in our practi-
cal decision making. In other words, having once decided
whether to take a car or a bus, we become largely indifferent to
other aspects of the work trip. What has transpired is the appli-
cation of a simplifying assumption that makes the decision pro-
cess more tractable.

Procedural Aspects of Design Thinking 54

 It should also be clear from the form of extensive representation
offered by the decision tree that many ordinary tasks (such as
seating people appropriately for a dinner party) may turn out to
be fairly complex, requiring to be tackled in a piecemeal, step-
by-step fashion rather than at a single stroke. Certainly more
complex and specialized tasks, such as finding the optimal ar-
rangement of departments in a large hospital complex, require
concentrated effort and a stepwise progression through a prob-
lem space.

Considerable psychological evidence would seem to support

this kind of observation and its expianation under the informa-
tion processing theory. For instance, the well-documented limits
on short-term-memory—the kind of memory that handles data
associated with immediate recall in connection with a particular
task—are placed at between 4 and 11 ‘‘chunks’”’ of information
(Miller 1956, Simon 1979).° Because the capacity of our short-
term memory is so circumscribed, complex problem-solving ac-
tivities requiring extended attention are likely to be performed in
a serial fashion. Combinations of basic procedures that allow the
generation of information and the assessment of progress to-
ward a solution are applied in a directed manner, as a Sequence
of events rather than as a single episode or experience.

Currently, the information processing theory is the dominant

school of thought about creative problem solving. As mentioned
earlier, experimental research is usually conducted through the
use of elaborate protocol analyses in which subjects engaged in
a problem-solving task faithfully narrate, or otherwise communi-
cate, what they are doing as they work through a problem. Un-
like the introspection of the Wurzburg School, the procedures
are immediate, generally replicable, and external in their mode
of operation. The results from these experiments are then
carefully examined to discern which fundamental information-
processing mechanisms appear to be involved. Often computer
programs are devised that incorporate these mechanisms and
allow solution of the same class of problem in a manner that is
indistinguishable from that of human subjects. In a very real
sense, these computer programs represent the theory and be-
come building blocks in the advancement of the theoretical
position.

A number of architectural researchers concerned with the de-

sign process have become involved with this kind of activity, or
at least in this line of speculation. The information processing
theory has also provided a basis for contemporary work in com-

Procedural Aspects of Design Thinking 59

 puter-aided architectural design, particularly for attempts to de-
velop ‘‘complete and hospitable design environments.” Here the
work of Negroponte, Mitchell, and Eastman provides clear exam-
ples (Negroponte 1970, 1972, 1974; Mitchell 1977; Eastman 1975).

@ Problem-Solving Behavior
Problem-solving behavior can be divided into three subclasses
of activity. The first, the representation of the problem through
structuring and restructuring a problem space, is known as the
problem representation problem. The second, the generation of
solutions, is termed the so/ution generation problem. The third,
the evaluation of candidate solutions, is known as the solution
evaluation problem. Those who study problem-solving behavior
generally make comparisons among problem solvers according
to differences in their methods of problem representation, solu-
tion generation, and solution evaluation. Clearly these three sub-
classes of activity are interdependent. The choice of solution
generation strategy may markedly affect the manner in which a
problem is represented and the manner in which solutions are
evaluated. It is generally in terms of solution generation strategy
that problem-solving procedures are described.

Trial-and-Error Procedures
In the strictest sense, random trial-and-error procedures involve
finding a solution to a problem in an entirely random manner. It
is arguable, however, whether truly random trial-and-error takes
place without some sort of bounding or narrowing down of the
scope of a problem. Obviously the concept of error requires the
presence of some means of testing a solution for desirable prop-
erties. This indicates the operation of at least tacit methods of
both evaluating and representing solutions.

The distinguishing feature of trial-and-error procedures is that

successive trials (proposals of solutions) are made indepen-
dently of the results of intermediate tests. In other words, infor-
mation provided by tests with regard to discrepancies between
the properties of proposed solutions and those deemed neces-
sary is not used to guide further information processing. This
usually means that although the problem solver can detect
whether or not a proposed solution is satisfactory, the means are
not available to use this information in any determinate manner.

Consider, for example, the problem of arranging furniture in a

room to satisfy some specified requirements, or the problem of
laying out the floor plan of a building to satisfy certain require-
ments regarding the adjacency of rooms, or the problem of sub-

Procedural Aspects of Design Thinking 56

IL =J

P=!
ll
=|

=r
l i
17
Pieces of a simple puzzle and the layout field.

dividing a piece of land among various kinds of uses. All three

are specific cases of a class of problem in architecture and ur-
ban design that is commonly referred to as the space-planning
problem.

Another very familiar example of this class of problem is the jig-

saw puzzle. Consider a simple puzzle made up of the pieces
shown in figure 17. Using trial-and-error, one would generate
candidate solutions by arranging and rearranging the pieces,
sometimes in a random manner. The pattern of lines and shapes
that resulted from each trial arrangement would then be com-
pared to a picture of the final solution to ascertain whether the
proposed solution actually matched the picture in all its particu-
lars. After a certain number of attempts at rearrangement the
process might be abandoned and the most complete proposal or
the one with the least discrepancy adopted (see figure 18).

In the hands of a young child this is a plausible strategy for solv-

ing the jigsaw puzzle, and given enough perseverance a line
drawing of Alberti’s early scheme (1460) for the facade of San
Sebastiano in Mantua would result (Wittkower 1971, p. 52). In the
hands of an adult, however, a more direct approach is likely to
be taken, unless the number of pieces is dramatically increased
or the picture of the solution confiscated. Under these circum-
stances it is likely that a considerable amount of trial-and-error
would be required to solve the puzzle, at least until some recog-
nizable pattern of lines emerged.

Procedural Aspects of Design Thinking 57

 < va

H
)
ll
oa
x =

a Se oO@
52)

Tl

=)
gS
—<

oOo : Q ee
a ©

|
= =
2) as=}
j=) aie

tr
= oO

Be he

tt
(an) (ms

=r
T
——
8

——
3

eae!

iT
t

|
S @
= S
— —

|
ve o
= =

=
~~
=
= e i
D> rab) |

=
[IL 15
[om
uw >
=][- N

i
&

Kai
fey (
S
(=)

=i
Yo}
x) =

=i
2 — Wie
iil

ili
Ul —ez
Pt (
Ee Ee |
0
omm©
|
Ag Anz
s
 We tend to resort to trial-and-error procedures when the mag-
nitude of a problem—the number of parts and of interrelation-
ships between them—is overwhelming, and especially when
information from the application of available testing procedures
cannot be used to direct the search for a solution. In the ex-
ample of the jigsaw puzzle it is possible to test matched and
unmatched lines between adjoining pieces, but the outline be-
yond this simple matching procedure may still not be evident.
The point is that solution generation takes place, to a significant
degree, independently from the organization of the problem
space after a prior trial.

Generate-and-Test Procedures
The generate-and-test approach is a variant of trial-and-error,
with the important difference that the results of tests are explic-
itly used to guide subsequent attempts to generate solutions.
Again the procedure takes place in the context of well-defined,
explicitly bounded problems.

Let us again consider the puzzle of Alberti’s San Sebastiano

facade. At first a solution is proposed, randomly or otherwise, by
arranging the pieces in the field provided. The shape of lines in
this arrangement is then compared to a picture of the final solu-
tion. If it does not conform, then the differences between the
arrangement and the solution are noted. This testing may be
done by observing whether lines match, whether they make rec-
ognizable figures, or whether the overall composition appears to
be lopsided in some way. Quite commonly the piece that seems
to be most out of place is chosen to be rearranged next. If the
second arrangement of puzzle pieces is better than the first,
then it is accepted for further improvement. If not, then the ear-
lier arrangement is used as the basis for further improvement
(see figure 19). The process of generating and testing candidate
solutions continues in this manner until a satisfactory arrange-
ment has been found.

Of course, there are several other versions of this general strat-

egy that could be proposed. For instance, rather than selecting
each time the puzzle piece that seems to be most out of place,
the overall arrangement of pieces, among several candidates,
showing the least discrepancy with the solution could constantly
be used as the basis for further improvement. The distinguishing
feature of the procedure, however, is that it makes explicit use of
information regarding the conformance of prior trial arrange-
ments to the required solution properties (here, the lines of the
drawing) as the basis for directing further search for a solution.

Procedural Aspects of Design Thinking 59

ll

rT | OO
if

example of generate-and-test on the puzzle

problem.
 Certainly not a random process, it involves, if in a rudimentary
form, explicit decision rules for guiding subsequent problem-
solving behavior (for instance, ‘‘Select the most out-of-place
puzzle piece and attempt to improve its location and
arrangement’).

Numerous examples of this kind of approach can be found at

work among architects and other designers in action. Many dif-
ferent manifestations of layout problems involve the constant ad-
justment of various parts (rooms, for example) with respect to
One another and with respect to a whole (a building), in order to
achieve a desired effect. Even in the resolution of formal design
problems, where we are operating under the aegis of some pro-
portional rule or compositional concept, we inevitably make
these kinds of incremental adjustments.

This strategy raises the question of how a problem solver deter-

mines whether the best possible solution has been found. Until a
point is reached where there is no discrepancy between a pro-
posed arrangement of spaces and the solution properties (the
criteria for arrangement), it is difficult to decide on a stopping
rule and to discontinue that particular aspect of problem-solving
activity. Furthermore, returning to our jigsaw puzzle example, it
may be the case that several different arrangements of pilaster
bays and openings are plausible, even as variations of the solu-
tion. Therefore, a totally determined arrangement of the puzzle
pieces cannot in fact be reached. One obvious stopping rule in
this kind of circumstance would be that the solver should quit
when no further improvement in the overall arrangement can be
observed under a number of successive arrangements. Never-
theless, there is no guarantee that this candidate solution is not
suboptimal and that with further perseverance we couldn't finda
better solution.

This phenomenon of incrementally moving from worse to better

solutions is often referred to as “hill climbing,’ where the top of
the hill is the location of the best solution. For some kinds of
problems explicit techniques have been developed, allowing as-
sessments to be made as to whether a candidate solution is sub-
optimal or not. For many other kinds of problems, however, no
such techniques exist. In any event, the idea that successive,
small, short-sighted steps can lead toward innovative results
contrasts starkly with the notion of insightful leaps described
earlier. Nevertheless, such procedures remain effective, espe-
cially when we bear in mind that a large number of small steps
may cover more ground than a leap (Perkins 1981, p. 151).

Procedural Aspects of Design Thinking 61

 Means-Ends Analysis
This strategy involves an extension of the generate-and-test pro-
cedure, particularly into the realm of allowing the provision of
alternative decision rules so that we can explicitly meet different
kinds of problem-solving situations. The three essential compo-
nents of this approach are (1) a prescribed set of actions
(means), (2) a prescribed set of goals (ends), and (3) a set of
decision rules. In short, the approach involves the explicit defini-
tion of both ends and means and an analytical framework, via
appropriate decision rules, for connecting these two aspects of
the problem together.

Consider again the facade of San Sebastiano in Mantua, only

this time not in the guise of a jigsaw puzzle but rather as a com-
positional problem framed quite explicitly by basic architectural
elements such as pilasters, wall openings, a pediment, and
steps, as well as by the decision rules (means) illustrated in fig-
ure 20. The ends are also quite explicit, namely to arrange the
elements into a classical form after the fashion of Alberti’s origi-
nal composition.

Along the way toward this kind of resolution, various kinds of

discrepancies, or compositional difficulties, are likely to be en-
countered, requiring various kinds of action and the employment
of various means. For example, the use of decision rules regard-
ing the division of the facade into bays of equal spacing and the
symmetrical distribution of the facade along a vertical center line
strongly indicates placement of the tallest wall opening in a cen-
tral position. The rules do not, however, necessarily provide
guidance as to the arrangement of the arched and squared-off
smaller wall openings. At least two possible arrangements ap-
pear plausible, as shown in figure 21—a dilemma that can be
resolved through the use of another rule that relates the centers
of the radii producing the arched openings with the center of
those producing the similar split pedimental opening, in the
form of an equilateral triangle (see figure 20). The result is distri-
bution of the arched openings in the outside bays and, further,
termination of the angle of incline of the stepped base. In other
words, through successive use of appropriate decision rules the
compositional problem can be resolved. Clearly, other rules
could and did apply to this case; it is hoped, however, that the
example is sufficient to illustrate the procedure.

Without too much suspension of disbelief, we can see how the

structural features of this rather simple illustration can be ex-
tended to many problems that we encounter in practice. It is

Procedural Aspects of Design Thinking 62

 at

a W meat

| a
Bilateral Symmetry
aaa aaa W/2:W/2
Lis
Hi \ ;
~/ | A 4
Wh. \ , Regular and Equal

Vi
Bel } Sided Figures
H = W, b, d.
AT ae GE
D
Regular Subdivision
4. =.
of Parts
| a a, b, ¢, d; @.
‘= +—

| | 5a
w/2 w/2
2a Three and Five
LITT T | part Composition.
a—3a—a, 5a, 5¢.

20
Rules for means-ends analysis of a facade problem.

Procedural Aspects of Design Thinking 63

 en.
Gaae

21
Alternative arrangements of openings for the facade
of San Sebastiano.

probably more likely that architectural design, within the canons

of some accepted or self-imposed aesthetic orthodoxy, coupled
with siting requirements and user needs, proceeds along these
lines than by the use of simpler generate-and-test procedures. In
fact, very strong similarities are evident between such means-
ends analysis and the format of rules and prescriptive devices
found in a number of theoretical works from the classical tradi-
tion, such as those of Perrault, Blondel, and later Durand, as well
as other members of the Parisian éco/es (Perrault 1683, Blondel
1771-1777, Durand 1802). Other practical guides, perhaps with
less theoretical merit and import, are to be found in the work of
people like Halfpenny and Morris (Halfpenny 1968 [1724], Morris
1971 [1734]). In fact, the seventeenth, eighteenth, and early
nineteenth centuries are replete with such treatises (Wiebenson
1982).

Generally speaking, the resolution of appropriate architecture,

from overall organization to the selection and conformation of
ornament, was fastidiously documented, not in a rote step-by-
step fashion from beginning to end, but through the generalized
treatment of a myriad of subproblems that might be encountered
and the rules by which solutions should be sought. We might
quibble over the normative aspects of this approach represented
by both the division of subproblems and the means for their so-
lution. Nevertheless, an explicit protocol for performing means-
ends analysis is very evident.

Procedural Aspects of Design Thinking 64

 Still other examples of the development and use of design rules
that reasonably conform to the framework of means-ends analy-
sis can be found among more recent architectural works. For
instance, in his ATBAT project of 1946, Le Corbusier specified
precise guidelines (‘‘means’’) concerning the resolution of con-
ditions of spatial arrangement toward the ‘‘end” of mass hous-
ing provision, as shown in figure 22 (Le Corbusier 1951, p. 116).
In many ways Habraken’s concept of ‘‘supports” and “‘struc-
tures” is of a similar kind, specifying as it does the means for
individual determination of a dwelling environment within a
comparatively unified system of production (Habraken 1972).

It should be clear from these examples, and from reflection on

the character of the space-planning problem mentioned earlier,
that although the number of rules, or types of moves that can be
made in two-dimensional space, may be quite finite in scope,
each rule may result in a process that allows for several out-
comes. The various concatenations of rules and means may thus
lead to a wide variety of particular renditions of acceptable com-
positional arrangements and spatial layouts. The point is that
such surface variety should not necessarily be construed as re-
sulting from an equal variety of fundamental problem-solving
procedures; nor, conversely, should it be assumed that a com-
paratively small number of procedures will result in a lack of
compositional variety.

Problem-Space Planning
The purpose behind the procedure of problem-space planning is
evident in the earlier definitions of a problem space and of the
decision tree that represents its structure. To reiterate for a mo-
ment, a problem space can be defined as an abstract domain
containing elements that represent knowledge states, some of
which are solution states to the problem at hand. As described
by the decision tree, the structure of a problem space is repre-
sented by nodes for decision points and branches for courses of
action. In considering a simple case where two courses of action
(1 and 2) are applied in sequence to represent an approach to-
ward a solution to a problem, the structure of the resulting prob-
lem space can be displayed as shown in figure 23. If we say that
a 2-1-2 combination, or ordering, of the two available courses of
action reaches a solution, then the minimum, but not the only,
path through the problem space is aj, bo, C3, dg. Thus, problem-
space planning, and the procedure used to accomplish it, aims
to structure the overall search process toward a solution: to help
select, in advance, an appropriate combination of courses of
action that leads economically to a solution. The example also

Procedural Aspects of Design Thinking 65

 CO)
ol
DewjE ra 80hi
ae
2XI6 =2i6 jam

Le Corbusier's guidelines for the ATBAT project of

1946.

Procedural Aspects of Design Thinking 66

23
A tree diagram of a problem space.

illustrates just how extensive the problem space can be for a

problem that appears to be quite straightforward. This further
underlines the need for some kind of planning.

The relationship between this kind of procedure and those dis-

cussed earlier should also be evident. It is through problem-
space planning that we define and subdivide larger problems
into more tractable subproblems and find guidance in applying
more specific strategies and procedures.

Two broad categories of problem-space planning can be distin-

guished. They are the top-down or hierarchical decomposition
methods and the bottom-up or hierarchical decomposition-
recomposition methods. This distinction becomes clearer under
another kind of representation of a problem space (Alexander
and Manheim 1962, Milne 1970).

Let us consider the elements of a problem space with distin-

guishable subproblems in the form of a finite hierarchical sub-
division down to the (primitive) level of singular statements
about specific criteria, or required solution properties (see figure
24). Reformulation and rerepresentation of the problem space
can be made in the form of a hierarchical planar or lattice struc-
ture, denoted by the solid and dashed lines, where the nodes
represent subproblems at various levels of detail and the links
represent (implied) connections between the subproblems. Thus
the node P? represents subproblem number one at the fifth level
in the hierarchy of subproblems, node P3 the second subprob-
lem at the third level, and so on. As shown, nodes P3, P?, and P3

Procedura! Aspects of Design Thinking 67

24
A hierarchical diagram of a problem space.

are linked, and therefore the subproblems that they represent

can be expected to have a common or similar ingredient. For
example, if the graph represents the structure of a problem
about the provision of urban outdoor spaces in a downtown
area, then the subproblems represented by the nodes P} and P3
may concern the design of retail commercial and recreational
areas, respectively. The two may be linked by concepts of pedes-
trian and vehicular movement. The subordinate node P?, in this
case, may deal specifically with something like a restaurant area
in direct proximity to a street and thus related most directly to P?
and more indirectly to P3.
In problem-space planning of the top-down or hierarchical de-
composition variety, the solver’s attention is directed downward
through the lattice structure, creating what tend to be broader
subdivisions of the problem at first, and gradually moving into
more numerous detailed subproblems (see figure 25). The over-
all (implicit) structure of the problem thus gradually becomes re-
vealed as the process unfolds, through times fj, ts, tz,..., tp, and
continues toward a higher level of resolution and eventual termi-
nation (Pena 1977).

In architectural design the reasonably immediate selection of an

overall parti, once the outlines of a problem have been pre-
sented or explained, is an example of this kind of approach.
Likewise, the immediate delineation of a community planning
problem in a direction such as transportation versus, say, com-
munity services is another case in point. Uniess we assume a

Procedural Aspects of Design Thinking 68

 a ee eee >
25
Top-down hierarchical subdivision of a problem
space.

special status for such “‘big ideas,” the process of discovering a

problem’s structure may also be initiated by ‘“‘smaller ideas.” For
instance, the tendency to light upon a familiar or intriguing as-
pect of the problem, even though that aspect may prove in time
to be relatively insignificant, often serves to ‘‘get things started.”

Returning to the facade of San Sebastiano, the composition of

which was presented earlier as an exercise in means-ends analy-
sis, we Can view problem-space planning as a method of or-
ganizing and sequencing the application of decision rules. It is
the process by which we think through the problem at hand, de-
fine it further, and plan our moves. One plausible sequence, il-
lustrated by figure 26, begins by applying general compositional
principles before becoming more particular. This is not to sug-
gest, of course, that Alberti’s scheme of approximately 1460 was
arrived at in quite so straightforward a manner. As Wittkower
points out, the project for San Sebastiano at Mantua, as does the
project for San Andrea in the same town, represents a departure
on Alberti’s part from earlier principles (Wittkower 1971, p. 47).
San Sebastiano, with its very thin pilasters on a continuous wall
surface, heavy entablature, and wall openings, represents a
theoretical shift. It is, in effect, an alternative ‘scheme for reviv-
ing the classical temple front... adapted to the needs of wall
architecture’ (Wittkower 1971, p. 49)—hardly a straightforward
design exercise. We do know, however, that certain of the com-
positional principles presented earlier were explicitly employed
by Alberti—for example, the width-to-height ratio of 1:1, the

Procedural Aspects of Design Thinking 69

 =

Be
eee

26
Sequential recomposition of the facade of
San Sebastiano.

Procedural Aspects of Design Thinking 70

 equal division of bay widths across the front of the facade, and
the general proportional system of regular subdivision of reg-
ulating lines (Wittkower 1971, p. 48). Parenthetically, the scheme
was not executed in the manner described. The two pilasters
next to the central bay were omitted, as well as the broad ex-
panse of steps at the base (Wittkower 1971, p. 49).

In spite of what might be seen as peremptory, even rash deci-

sions embraced by this general approach, in the hands of experi-
enced practitioners it can be very successful. Fluent problem
solvers make use of a large repertoire of plans and planning
schemes. Consequently they appear to tackle problems ‘‘head
on,” or ‘top down.” Furthermore, lack of clarity in the problem
as given may render an alternative approach inconceivable.

By contrast, in the use of bottom-up or hierarchical decomposi-

tion-recomposition methods the aim is to provide a complete, or
at least extensive, description of the inherent structure of a prob-
lem space from the outset by explicitly breaking down—decom-
posing—the problem as given and understood into its most
fundamental components; that is, to the level of the kind of sin-
gular problem statement mentioned earlier. The relations among
these components are then systematically identified, allowing
recombination into a coherent picture of the problem-space
structure.

The procedures described in Alexander’s landmark work Notes

on the Synthesis of Form, and in later works dealing with “‘pat-
tern language,” are essentially problem-space planning methods
of this kind (Alexander 1964; Alexander 1965; Alexander,
Ishikawa, and Silverstein 1968; Alexander et al. 1975; Alexander
et al. 1977; Alexander 1979; Studer 1965; Milne 1970; Owen
1970). In general they proceed in the following manner (see fig-
ure 27). First, the problem as given is restated in a form consist-
ing of numerous discrete problem statements. In other words, it
is described in terms of its singular constituent elements rather
than by broad organizing principles. For example, in Alexan-
der’s case of the Indian village, under the heading of social
forces shaping the community environment we find simple de-
scriptive sentences such as, ‘‘Extended family is in one house,’
‘“‘Need to divide land among sons of successive generations,”’
and so on (Alexander 1964). Second, relations among these dis-
crete problem statements are expressed, usually in the binary
form of ‘‘yes, there is a relationship” or ‘‘no, there is not a rela-
tionship.’’ For example, again for Alexander’s Indian village, the
problem statement ‘‘Rules about house door not facing south” is

Procedural Aspects of Design Thinking 71

1 3 Fak
ENTIRE VILLAGE
Religion and Caste
Harijans regarded as ritually impure, untouchable, etc.
Proper disposal of dead.
Rules about house door not facing south.
Certain water and certain trees are thought of as sacred.
Provision for festivals and religious meetings.
Al A2 A3 Bi B2 B38 B4 Ci C2 D1 D2 03
Wish for temples.
Cattle treated as sacred, and vegetarian attitude.
ONO
CORSE
OS Members of castes maintain their caste profession as far as A1 contains requirements 7, 53, 57, 59, 60, 72, 125, 126, 128.
possible. A2 contains requirements 31, 34, 36, 52, 54, 80, 94, 106, 136.
9. Members of one caste like to be together and separate from A3 contains requirements 37, 38, 50, 55, 77, 91, 103.
others, and will not eat or drink together. B1 contains requirements 39, 40, 41, 44, 51, 118, 127, 131, 138.
10 . Need for elaborate weddings. B2 contains requirements 30, 35, 46, 47, 61, 97, 98.

Social Forces
11. Marriage is to person from another village.
12. Extended family is in one house.
13. Family solidarity and neighborliness even after separation.
14. Economic integration of village on payment-in-kind basis.
ik Modern move toward payment in cash.
16. Women gossip extensively while bathing, fetching water, on
way to field latrines, etc.
ile Village has fixed men’s social groups.
18. Need to divide land among sons of successive generations.
19. People want to own land personally.
20. People of different factions prefer to have no contact.
21. Eradication of untouchability.
22. Abolition of Zamindari and uneven land distribution.
23. Men’s groups chatting, smoking, even late at night.
24, Place for village events—dancing, plays, singing, etc.,
wrestling.
20. Assistance for physically handicapped, aged, widows.
26. Sentimental system: wish not to destroy old way of life; love of
Excerpts from Alexander’s procedure in Notes on
present habits governing bathing, food, etc. the Synthesis of Form.

interacts with 8, 9, 12, 13, 14, 21, 28, 29, 48, 61, 67, 68, 70,
77, 86, 101, 106, 113, 124, 140, 141.
interacts with 3, 4, 6, 26, 29, 32, 52, 71, 98, 102, 105, 123,
133.
interacts with 2, 12, 13, 17, 26, 76, 78, 79, 88, 101, 103, 119.
interacts with 2, 5, 6, 17, 29, 32, 45, 56, 63, 71, 74, 78, 79, 88,
91, 105, 106, 110, 124.
interacts with 4, 6, 10, 14, 17, 21, 24, 46, 102, 113, 116, 118,
131, 133, 140.
interacts with 2, 4, 5, 20, 21, 53, 58, 61, 63, 82, 102, 111, 117,
130, 134, 135.
interacts with 20, 31, 34, 53, 57, 58, 59, 80, 85, 86, 94, 105,
106, 123, 124, 125.
interacts with 1, 9, 14, 15, 21, 22, 25, 27, 48, 58, 59, 61, 62,
63, 64, 65, 89, 95, 96, 99, 111, 112, 114, 115, 116, 121, 129,
136, 140, 141.

Procedural Aspects of Design Thinking 72

 connected with a variety of religious rituals and with ‘“‘Sentimen-
tal system: wish not to destroy old ways of life. . .” (Alexander
1964). In a design problem about an urban space, such as the
one mentioned earlier, identifiable aspects of the problem hav-
ing some bearing on pedestrian movement presumably would be
related to one another. Third, very closely interrelated groups of
problem statements are identified, thus defining the most basic
level of design subproblems. For instance, returning to Alexan-
der’s Indian village, this may be an expression of house design
criteria. Fourth, the subproblems isolated in this manner are
combined, using the pertinent relational information, usually in a
hierarchical stepwise fashion. This process of combination cul-
minates in a single problem statement that is inclusive of all sub-
ordinate statements. It corresponds, then, to the root node P} of
the lattice in figure 24. In short, the procedure allows decisions
made by a designer about parts of a problem, and about their
interrelationships, to be systematically structured so as to pro-
duce an explicit picture or diagram of the problem space at vari-
ous levels of generality. Finally, this structure can be further
exploited in order to identify those aspects of particular sub-
problems of interest that seem to have the most controlling influ-
ence over the configuration of the problem space. For example,
if the problem essentially revolves around concepts of move-
ment and circulation, this will become evident. These may not be
concepts that are apparent to the designer at the outset, and
their revelation may also be consequential by exposing a bias on
the part of the designer that may subsequently become the basis
for reconsideration of the problem’s structure (Silverstein and
Jacobson 1978).

Clearly, alternative representations of and decisions about the

problem space are possible and may even prove more enlighten-
ing. For instance, there are procedures where the relational
structure is seen as representing the ‘‘lines of communication”
along which subproblems are connected. This technique allows
instances of partially contradictory statements to be resolved (De
Leon 1972, Batty 1971). For instance, the assertion that doors of
houses should not face south may be made from one point of
view, and the contention that they should face south from some
other standpoint. Obviously such statements are related through
the contradictory aspect. Nevertheless, appropriate resolution
may not be as simple as taking them to be self-canceling. What
is required is a resynthesis of the information in which the ambi-
guity is either preserved or resolved by reference to other more
determining factors. For example, the preponderance of evi-

Procedural Aspects of Design Thinking 73

 dence from elsewhere in the problem space may suggest that of
the two alternatives the south-facing doorway is the least detri-
mental to expected patterns of behavior.

We can see that both top-down and bottom-up approaches to

problem-space planning apply to well-defined problems, or at
least to problems for which we have sufficient definition and
understanding to make the necessary planning decisions. On the
other hand, the idea sometimes associated with Alexander and
others’ formal treatments of this issue, that the procedures are
strictly a way of defining problems, is a mistaken one (Studer
1965). We can certainly say that they are techniques for structur-
ing problems, but not that they are techniques for defining them.
Clearly, specification of problem elements and of the interrela-
tionships among them requires the use of some overarching
concept or knowledgeable attitude about a problem: a knowl-
edge that, when all is said.and done, remains in some sense
implicit, even tacit. In this regard it further seems inaccurate to
assume that top-down approaches are necessarily inferior to
bottom-up approaches because they provide, at a given moment,
a less explicit structural description of a problem space. Here we
must observe that bottom-up methods, although seemingly more
determined, or determining, are just as vulnerable as their
counterparts to shifts in understanding and orientation to a
problem. The desirability, even the necessity, of explicit problem
structures, according to Alexander and others, lies in the critical
judgment that can be brought to bear on the biases and misin-
formation that are revealed when a problem is explicitly and
comprehensively stated (Silverstein and Jacobson 1978).

Finally, the formal accounts of problem-space planning that

have been discussed are altogether idealized forms of the proce-
dures, just as were the earlier examples of other problem-solving
approaches. In actual practice, designers rarely if ever adopt one
extreme or the other. Nevertheless, at least implicitly and at par-
ticular moments, our problem-solving behavior can seem to lean
one way or the other.

eid COsacs ee oy wee eRe eS EON

Heuristic Reasoning and Design ‘‘Situations”’

One aspect in this discussion that has been alluded to but so

far not really addressed is the class of devices variously termed
decision rules and overarching concepts. Particularly in con-
nection with generate-and-test procedures and means-ends
analysis, the ability to invoke an appropriate rule for relating the

Procedural Aspects of Design Thinking 74

 results from an evaluation of solutions to further courses of ac-
tion was fundamental. Also, rules and principles are necessarily
invoked when one is pursuing the task of problem-space plan-
ning. Such behavior introduces us into the realm of heuristics
and heuristic reasoning.

To Newell, Shaw, and Simon, a heuristic is any principle, proce-

dure, or other device that contributes to reduction in the search
for a satisfactory solution (Newell, Shaw, and Simon 1967, p. 78;
Simon 1969, p. 80).° Perkins in his account is less emphatic
about the results. For him, ‘‘a heuristic is a rule of thumb that
often helps in solving a certain class of problems, but makes no
guarantees” (Perkins 1981, p. 192). It may be compared to the
exact procedures of an algorithm in that regard. Similarly, for
Polya, the term again applies to ‘‘rules of thumb,” or provisional
procedures that prove useful in solving problems (Polya 1957, p.
37). In short, heuristics is a term that is applied to specific prob-
lem-structuring devices ranging from explicit decision rules of
the type discussed earlier to a wide variety of analogies, analogs,
and models. It is also applied to general kinds of procedures for
guiding the search for solutions. For example, the ‘“‘heuristic of
reducing differences,” apart from its particulars that may vary
from one application to another, lies behind the concept of
means-ends analysis (Newell, Shaw, and Simon 1967, p. 87). As
we have seen, the specific decision rules are aimed at reducing
the difference between the properties of proposed solutions and
those of the goal set. In the foregoing accounts of problem-
space planning, the underlying concepts enabling decisions to
be made (that is, the heuristics employed in the overarching con-
cepts) direct specific decisions and thus are instrumental in the
manner in which a problem becomes defined in the first place.

The term heuristic reasoning refers to a problem-solving pro-

cess in which it is unknown beforehand whether a particular
sequence of steps will yield a solution or not. Consequently, it
involves a decision-making process in which we do not know
whether we actually have a solution until the line of reasoning is
completed, or all the steps are carried out. Heuristic reasoning is
part and parcel of most solution generation strategies, may be
characteristic of an individual problem solver, and guides the
overall organization of search through a problem space. So far,
however, it is an area in which no general theory seems to exist.
In fact, there is considerable disagreement about the emphasis
that should be placed on heuristic procedures in the realm of
human probiem solving, particularly with reference to what is
known as “‘heuristic technology’ —the development of proce-

Procedural Aspects of Design Thinking 75

 dures that are superior to those routinely used. Skeptics assert
that the broad organization of problem-solving behavior almost
takes care of itself, once a person masters the contributing per-
formances that are required. Knowing various techniques for
solving layout problems, for example, more or less guarantees
that the overall approach to this type of problem will unfold ap-
propriately. On the other hand, Perkins, among others, presents
evidence and arguments that suggest the importance of heuris-
tics in matching already familiar methods with problems during
the pursuit of more efficient problem-solving performance (Per-
kins 1981, p. 196).
This picture of episodic rule making and application does not
entirely convey the nature of the act of heuristic reasoning, at
least as it occurs in the design process. For one thing, the rules
may or may not be explicit and repeatable in the sense of some-
thing objectively acquired.and learned. The term heuristic is
used here in a broader sense than in heuristic technology, where
subjects learn the application of various problem-solving tech-
niques in order to improve their performance on specific tests.
The heuristics employed by designers may be quite subjective,
having evolved from prior personal experience. They are cer-
tainly explicit, but only in the contexts with which they are asso-
ciated in the designer’s mind. Further, rule making and rule
application have a certain congruence with the concepts of
“framing,” ‘‘frames of reference” (Schon 1984), and ‘“‘apprecia-
tive contexts” (Vickers 1983), although heuristic reasoning is
less potentially contemplative and more immediately associated
with action. Useful knowledge of a rule is as much a matter of
the purposes to which it is to be put as it is a matter of the rule’s
external existence.

This concept of heuristic reasoning, when it moves beyond the

objective and perhaps mechanistic realm of problem solving
theory, bears a striking resemblance to aspects of Merleau-Pon-
ty’s concept of situation (Merleau-Ponty 1962). As interpreted by
Mallin, Merleau-Ponty uses situation to mean ‘‘involvement in
circumstances’ or “‘active concern with sets of natural, cultural,
or human problems” (Mallin 1979, p. 7). A situation occurs when
an individual becomes totally absorbed in something, relates it
to himself, and begins to understand it (Mallin 1979, p. 7). Quite
clearly, various episodes of design presented in chapter 1 have
these characteristics about them. To Merleau-Ponty, when we
are in a situation neither the objective realm of those things out-
side ourselves nor our own subjectivity is primary (Mallin 1979,
p. 8). Both are founded in the situation.'° Therefore, in the mat-

Procedural Aspects of Design Thinking 76

 ter of heuristic reasoning it is not simply a case, as the informa-
tion processing theorists would seem to suggest, of isolating
problem-solving circumstances and setting them over against a
similarly abstracted set of actions and purposes. By the same
token, it is not a case of subjective idealism either (Mallin 1979,
p. 12). For how could we explain the very evident struggle that
took place in the first chapter’s protocols to reorganize, manage,
and be accountable for objective aspects of problem-solving
contexts?

The rule aspect is also present in Merleau-Ponty’s concept of

situations. In a creative endeavor such as designing, an attempt
is made to overcome the situation’s novelty and make some
sense of it. This process leads to the acquisition, as Mallin puts
it, “of a sort of general principle . . . that is habitually reapplied.”
Thus the acquired principle becomes ‘‘sedimented,”’ allowing
one to “learn how to act in certain kinds of circumstances”
(Mallin 1979, pp. 12-13).

The open-ended aspect of heuristic reasoning finds a parallel in

Merleau-Ponty’s definition of situations as being ambiguous.
They are said to be beyond complete and perfect grasp; further
meditation can always lead to “‘a more specific and deeper com-
prehension” (Mallin 1979, p. 14).

In summary, the design process may be seen to be marked by a

sequence of episodes or situations that are, in turn, coincident
with periods of heuristic reasoning through which problems are
defined and solutions sought. During each episode a particular
heuristic device or set of devices can be said to be in operation
and in general control of the reorganization of a problem space.
Further, the orientation of this operation is neither entirely ob-
jective nor entirely subjective. It is both. Between episodes, con-
trol is relinquished, so to speak, from one set of organizing
principles to another.

@ Logical Structure
In simple form, the logical structure behind the application of
heuristic reasoning would seem to be, ‘‘If problem X is encoun-
tered or perceived, then take action Y under conditions Z” (Akin
1978, pp. 27-34; Akin 1982 [1979]). In the case of architectural
design, action Y might take the form of a specific design re-
sponse, such as the prescription and manipulation of the com-
positional qualities of some building elements in response to a
perceived probiem and its surrounding or auxiliary conditions.
For example, the problem might be construed as one of satisfy-
ing the requirement of a distinct and grand formal entry to a

Procedural Aspects of Design Thinking 77

 building complex, given the available components of the build-
ing program and the features of the site (the auxiliary or bound-
ing conditions of the problem). Here, action Y might be an
arrangement and massing of the elements of the building in
which major public areas are combined at the termination of a
principal thoroughfare leading into the site. It would thus resolve
both the problem at hand and specific conditions inherent in the
building program and site configuration, such as adjacency re-
quirements among functional areas and accessibility.

On the other hand, with the earlier definition of a wicked prob-

lem in mind, the logical structure of the reasoning process might
also be seen as, “If conditions Z obtain when viewed from the
perspective of the application of action Y, then problem X is de-
fined.” In other words, a problem becomes defined in response
to a perceived set of conditions by way of a concrete proposal.

At first reading this may seem to be a perverse state of affairs.

Nevertheless, it is the framing and self-referential qualities of this
kind of logical structure that allow more comprehensive judg-
ments about the scope and thrust of the problem-solving situa-
tion to be suspended for a moment, permitting problem-solving
activity to proceed. This logic characterizes the ‘hunch making”
so often observed in design behavior, without which it is difficult
to imagine how any kind of solution could be reached. For ex-
ample, in the first of the case studies presented in chapter 1, a
preoccupation with resolving the street plaza on the front por-
tion of the site allowed the design scheme to be advanced, with
an almost complete absence of concern for any other aspect of
the problem.

@ Probiem-Oriented Constraints and Autonomous Constraints

In connection with the logical structure of the heuristic rea-
soning process, a question arises about the manner in which a
particular action Y becomes associated with either the recogni-
tion of a problem X or the perception of conditions Z. To put it
another way, how is it that the ensemble of problem X, action Y,
and conditions Z becomes recognized in the first place?

Although other aspects of this question will be tackled later, for

present purposes it is useful to distinguish between two types of
problem constraints. According to Simon, these are ‘‘problem-
oriented constraints” and constraints that are autonomously
(that is, independently) provided by a designer in order to orga-
nize a problem space (Simon 1970, 1973).

Procedural Aspects of Design Thinking 78

 Problem-oriented constraints are those that are primarily derived
from consideration of a problem X that has been previously de-
fined. In other words, as far as the problem solver is concerned,
the necessary and sufficient conditions for taking action Y are
determined solely within the ambit of problem X and related con-
ditions of the task environment Z. A form of means-ends analy-
sis may take place, in which the means are defined and specified
in terms of clearly identifiable ends. For example, the point of
symmetry with Burnham's planetarium, which became such a
determining factor in the third case study in chapter 1, was es-
sentially part of the problem as given. So too was the location
and shape of the existing hospital in the second case study,
which subsequently constrained the location and orientation of
building elements along the lake edge.

By contrast, autonomous or independent contraints do not de-

rive from the problem as given and understood. They are, how-
ever, not arbitrary, as their application may lead positively to a
reformulation of the problem in a different light—a reformula-
tion that greatly facilitates further problem-solving acitvity. In
essence, information about the ensemble of X, Y, and Z is in-
troduced into the problem-solving process from somewhere
else, rather than by direct consideration of problem X under con-
ditions Z. An example is the use of the Chicago grid in the third
case study, for the purpose of providing constraints for the over-
all layout of the site. While the grid was certainly present in the
area adjacent to the site, there was nothing in the problem state-
ment, or brief, that required that any reference be made to it.
This constraint, introduced by the designers, usefully tran-
scended the givens of the problem situation.

It is the “covering” characteristics, yet lack of complete corre-

spondence (isomorphism) between autonomous constraints and
their problem-based contexts that renders them so useful in pro-
viding a basis for further problem-solving activity. For unless the
entire problem at hand can be solved using strictly problem-
oriented constraints, we have to step outside the known problem
context in order to continue problem-solving activity. In the vein
of Merleau-Ponty, the “otherness” of the independent facts sur-
rounding the problem can only be broached by the use of “‘sedi-
mented principles’ known to the designer (Mallin 1979, p. 13).

Procedural Aspects of Design Thinking 79

Preis Seek pW NAR UR ene Na amena CR a eas
Types of Rules and Constraints at Work
in Design

One way of describing the kinds of heuristics that are employed

to constrain problem spaces in architectural and urban design is
with reference to the type of information that they provide. At
least five classes of heuristics can be distinguished, in the guise
of common types of analogy, “solution images,” and form-giving
rules. They are as follows: anthropometric analogies, literal
analogies, environmental relations, typologies, and formal
languages.

Anthropometric Analogies
The use of an anthropometric analogy involves employing a
mental construct that describes man’s physical occupancy of a
space. Such constructs can often be seen at work in the deliber-
ations of naive designers or in problem-solving situations where
a designer has little or no experience. For example, a person
without any architectural background might-produce a design
for a staircase based entirely upon the act of imagining someone
ascending into a room in a certain manner. The result might be a
graceful form for which the designer in question appeared to
have no prior reference. Almost anyone who has experience
with moving, sitting, standing, and so on, yet no special knowl-
edge, would seem capable of devising and exercising such a
construct.

Recently, in the face of what they see as an abstract and non-

body-centered architecture, Bloomer and Moore have attempted
to reexamine the significance of the human body in architecture.
In their words, ‘‘We believe that the most essential and memo-
rable sense of three-dimensionality originates in the body
experience and that this sense may constitute a basis for under-
standing spatial feeling in our experience of buildings’ (Bloomer
and Moore 1977, p. x). Certainly in other eras the body and an-
thropometric qualities were more central to architectural
thought (see figure 28).

Literal Analogies
This kind of heuristic involves borrowing known or found form-
giving constructs as a point of departure for structuring a design
problem. Literal analogies are so termed because in all cases the
resulting architectural forms match very closely the conforma-
tion of what the designer sees as the key features of the analog.
Here a useful distinction can be made between what we might

Procedural Aspects of Design Thinking 80

28
An elegant stairway with anthropometric qualities
(the main stair of the Benedictine monastery at
Gotweig, by von Hildebrandt, 1718-1739).

Procedural Aspects of Design Thinking 81

 call iconic analogies and canonic analogies (Broadbent 1973,
chez):
The scope of references for the development of iconic analogies
is extremely broad. Objects from the natural world may serve as
sources—the shell of a crab for the roof of Le Corbusier’s Ron-
champ Chapel (Le Corbusier 1958), hands folded in an attitude
of prayer for Frank Lloyd Wright’s Unitarian church (Broadbent
1973, pp. 40-45; Wright 1943), sails for Utzon’s Sydney Opera
House (figure 29), in a milieu steeped in yachting (Ziegler
1973).'' Iconic analogies can also include imagery from some
scene, painterly conception, or narrative account of real or imag-
ined circumstances. As Waldman clearly demonstrates in his
commentary on “narrative design problems,” we can and do
make use of story lines as metaphors in solving design problems
(Waldman 1982). Hans Hollein presents the elements of such a
scenario in his design for.the Ringturm Travel Agency in Vienna,
which incorporates the motifs shown in figure 30. Artifacts and
elements from more squarely within the realm of architectural
experience represent a third major resource.

In all cases the analogy appears useful to designers by virtue of

the symbolic or iconographic qualities that they attach to it. It is
a physical representation of an intention that, when applied, pro-
vides additional structure to a problem.

Canonic analogies have as their basis “ideal” proportional sys-

tems or formal geometric properties. They are usually mani-
fested as somewhat abstract geometrical patterns or shapes.
Configurations such as ‘‘Cartesian grids’ or ‘‘Platonic solids”’
have quite a venerable history in the shaping of design prob-
lems, although the proposed ideal may show considerable varia-
tion. For example, the Hellenic urban grid layout of the fifth
century B.c., as at Priene, had a modular order of more or less
equally spaced blocks and streets with a central area set aside
for special functions. By contrast, the later layout of Alexandria,
by the Macedonian architect Dinocrates (c. 331 B.c.), accom-
modated variations in both site conditions and local functions.
One of the features of the Alexandrian grid was the deployment
of multiple subcenters, each within an ethnic quarter of the city,
linked together in the form of a hierarchical functional system
(figure 31).

A particular literal analogy may, of course, be capable of simul-

taneously expressing both iconic and canonic qualities. In actual
application the final distinction must be regarded as a circum-
stance of the moment and clearly both a matter of degree and a

Procedural Aspects of Design Thinking 82

 29
A form with the visual quality of a sailing vessel
(original elevations for the Sydney Opera House, by
Jarn Utzon, 1957-1973).

Procedural Aspects of Design Thinking 83

 arare aiaaee
30
Elements of a narrative scenario (the Ringturm
Travel Agency in Vienna, by Hans Hollein, 1979).

Procedural Aspects of Design Thinking 84

 Biieswas
E358 be
y
‘wy? a |

31
Two variations on the grid system of town layout
(1. Priene, 2. Alexandria).

matter of the purposes to which the analogy is put by the

designer.

Environmental Relations
Here use is made of a principle or set of principles, often derived
empirically, that represents what appear to be appropriate rela-
tions between man and his host environment and among com-
ponents of the building fabric itself. Special information is
typically incorporated about human behavior as a determinant of
form; about the influence of environmental factors such as cli-
mate, physiography, and resource availability (see figure 32);
and about engineering factors affecting structural and material
performance.

This kind of heuristic is apt to be highly problem oriented. The

principle involved creates a bridge between a perceived problem
and an ensemble of form-giving characteristics representing its
potential resolution. For example, the application of “‘bubble dia-
grams’”’ in the arrangement of spaces in accordance with given
formulae about human behavior—such as how far apart people
prefer to sit in a waiting room—rarely moves beyond the con-
straints of the problem as given.

Typologies
Typologies embody principles that designers consider unvarying
(Colquhoun 1967); as heuristics they allow us to apply knowl-
edge about past solutions to related architectural problems. For
this discussion it will be useful to divide them into three sub-
classes: building types as models, organizational typologies, and
elemental types. Again, a long and venerable history is associ-
ated with their use (Vidler 1977).

A building type used as a model represents characteristics

worthy of emulation. It seems to provide for the perceived needs,
uses, and customs found in the design situation under consider-

Procedural Aspects of Design Thinking 85

 ation. For example, a courtyard house, a French hétel, or a
basilican church may be closely followed in arriving at a design
solution. Vidler in his apologia for the neorationalists sees our
cities as a typological source of design ideas (Vidler 1978).

The organizational typology is primarily used as a framework

and reference for solving problems concerning the spatial distri-
bution and conformation of functional elements. It may also be
used for the underlying rules of formal composition that it offers.
A case in point would be the use, without complete or literal
emulation, of those organizational principles at work in a classi-
cal villa facade.

Finally, elemental types are prototypes for solving general

classes of design problems—the problem of entry into a build-
ing, for example, or that of rendering the transition between the
ground plane and the rise of a building, or that of resolving the
needs for both a sense of community and a sense of privacy. The
corpus of architectural production and theoretical writing pro-
vides a repertoire of approaches to such problems.

The foregoing classification is, or course, rather arbitrary. A

building, or part of it, may be used as a prototype in all three
regards, serving at once as model, organizational type, and pro-
vider of elemental types. Again, the particular orientation of the
use of a typology in design is largely a matter of the moment and
of the designer’s intentions. Furthermore, typologies implicitly
possess both the iconic and the canonic qualities of literal
analogies—but with the important difference that these qualities
are confined to the realm of existing architectural expression
and thus can be considered ‘tried and true.” This property is not
necessarily applicable to literal analogies. As with other heuris-
tics, however, the effectiveness of a borrowed type hinges ulti-
mately on its demonstrated appropriateness to the design prob-
lem at hand.

So far, in this as in many other discussions of typologies, there is

a distinctly reverential tone. The types referred to are usually
those with time-honored pedigrees from the point of view of
contemporary historical-theoretical preoccupations. Yet other
types must surely be admitted: organizational principles stem-
ming from current practice or lore, without overt historical refer-
ences. Sometimes these arise out of a sober, practical reckoning
with current needs and prevailing circumstances; on other occa-
sions they might stem from forward-looking polemical positions.
In modern urban design enterprises both these latter strains are
often evident among the principles presented to rationalize a

Procedural Aspects of Design Thinking 87

 scheme. Organizational diagrams manifesting such contempo-
rary preoccupations as circulation, market forces, and prevailing
institutional mechanisms of development are cases in point
(figure 33).

@ Formal Languages
These are generalizations of information from other heuristics,
especially from typologies and environmental relations. They are
languages inasmuch as they possess guiding structures or rules
that explicitly direct decisions about the ‘‘correct’”’ functioning
and ‘‘meaningful’”’ ordering of formal design elements. For ex-
ample, treatises on the ‘‘classical language”’ provide a repertoire
of architectonic elements and rules for their composition that
undoubtedly incorporate fundamental aspects of relevant typolo-
gies but go beyond the realm of particular types in both scope
and generality (Wiebenson 1982, Summerson 1979). Such works
present the ‘‘semantic’”’ and ‘‘syntactic’”’ ingredients of an inter-
nally consistent architectural expression.

Alexander’s ‘pattern language’”’ (Alexander et al. 1977) repre-

sents a heuristic repertoire that is also of this kind (figure 34).
Furthermore, it is a no less deterministic method of reasoning,
although ostensibly concerned in a different manner with the
problem of architectural composition (that is, with behavioral
determinism rather than formal determinism). Yet a design
language can be quite idiosyncratic and full of the references,
leitmotifs, and particular approaches of an individual designer.
This quality derives from a habitual way of doing things—the use
of “sedimented principles,” to return to an earlier discussion—
and is most evident in the work of original and experienced
designers.

Generally, the heuristics designers use in practice are principles

that draw on sources from a number of these categories. For
example, in Shadrach Woods's proposal for the Berlin Free Uni-
versity, where the heuristics at work are fairly well documented
(as shown in figure 35), a number of important organizing con-
cepts, particularly those dealing with overall spatial organization
and qualities of flexibility, clearly derive from both the category
here called typologies and that called environmental relations
(Woods 1964, 1965). In the case of Woods’s proposal, the clarity
of the heuristics also derives from the juxtaposition with contras-
tive features that are not to be employed—for example, Woods’s
“groundscraper diagrams’ versus ‘“‘high-rise diagrams.”

In addition, a heuristic may encompass either a ‘‘large idea’’ ora

“small idea’ and may become more developed through repeated

Procedural Aspects of Design Thinking 88

Penthouse/tower Terrace walk-up Terrace
46 units/1BR 2BR 12 units/OBR 1BR 2BR 16 units/2BR 3BR

Terrace Loft space/flexible living, fixed BR Penthouse/flat

20 units/2BR 3BR Number of units unspecified 25 units/1BR 2BR 3BR

33
Variations on the New York block typology (part of
the project from the Welfare Island competition,
New York, by 0. M. Ungers, 1975).

Procedural Aspects of Design Thinking 89

 3. Size based on population B
4. Community territory
7. Entrance location
11. Arena enclosure
43. Waiting diversions

. Arena thoroughfare
. Open to street
. Necklace
. Community projects
. Entrance shape
. Subcommittee watchdogs
. The intake process
. Outdoor seats
. Information

. All services off arena

. Free waiting
. Overview of services
. Self-service
. Pedestrian density
. Building stepped back
. Vertical circulation
. Short corridors
. Arena diameter
. Stair seats . Core service
. Town meeting
. Block worker layout
. Meeting rooms
. Staff lounge
. Square seminar rooms

18. Windows overlooking life 20. Activity pockets 42. Sleeping OK

26. Vertical circulation in services 27. Self-service progression 48. Barbershop politics
33. Service layout 29. Outdoor seats 53. Form-filling tables
40. Office flexibility 32. Child-care position 54. Accessible bathrooms
56. Informal reception 38. Community wall 57. Child-care contents

34
The “pattern language” of Alexander, Ishikawa, and
Silverstein (an application in the San Francisco
area).

Procedural Aspects of Design Thinking 90

 application. Using terminology and concepts from Lakatos’s
writings on the history and philosophy of science, Stanford
Anderson presents a convincing analysis of the early work of
Le Corbusier in which two large ideas can be seen to emerge
(Anderson 1984, Landau 1982). These ideas he calls ‘‘pro-
grammes’; each has at its core a working hypothesis, or set of
related hypotheses, through which Le Corbusier organizes
much of his architectural work. The programmes are not mutu-
ally exclusive and receive varying degrees of emphasis in differ-
ent projects. Further, the core working hypotheses can be seen to
evolve from relatively vague notions to fixed and clear principles
for action.

The programmes Anderson defines are the concepts of prome-

nade architecturale and Maison Domino. |n the promenade
architecturale Le Corbusier's central concern is with “how we
correlate experience and knowledge’’; he proposes ‘“‘an abstract
experience of architecture’ based on immediate perceptual
experience rather than historical association. In the Maison
Domino, Le Corbusier is seen to be seeking a ‘‘rational and eco-
nomic solution” for the provision of emergency housing. Both
programmes come together to influence the later formulation of
the ‘‘five points,’’ one of which deals with the “free plan” and its
possibilities for spatial experience, and another with rationaliza-
tion of structure using the promise of modern technology
(Anderson 1984, p. 3f.).

A final point about the heuristics employed in a design exercise

is that they may have an entirely informal quality and a relatively
short life span. They may be developed for a particular applica-
tion and subsequently discarded. Although this is especially true
of problem-oriented constraints, it can also happen with an
autonomous constraint in the form of a ‘‘wild idea” that comes
to the fore, is useful for a moment, and is then forgotten.

Pais BO ae ee es Oe so aL a RCD
Aspects of Design Behavior

Equipped with the concept of heuristics and heuristic reasoning,

we now can take up the task of identifying and attempting to
elucidate the characteristic features of the problem-solving
behavior of designers in action. Although a more phenomeno-
logical account may at times seem more appropriate to the spirit
of design activity, a general adherence to the information-
processing paradigm of problem-solving theory will be main-
tained because of its breadth and precision.

Procedural Aspects of Design Thinking 91

 1. THE IDEA oF UNIVERSITY G THE EXTERNAL CrPRLEsion oF
THE NEEO FOR AHO MXCHANGE OF DIFFERENCES IN FURnCTION (ARE
GENERAL AND SPECIAL INFORMATION, THESE AB IMPORTENT AE THE
sinwanines 7) ano nosrarcia
SPECIAL on FOR REPRESENTATIVE FOAM ALSO
INFORMATI
TEND TO SEGREGATE THE sNIVER SITY
INTO SPECIALIZED DISCIPLUNES ONLY

nee ott iy
| »8! GENERAL 3°
JLAFORMATION,
1 “4 \
ey r) \ *

aoa
NEED. HHEES
INFORMATION @--- TEED yee?
NE gg:
a8 av
SPECIAL
INFORMATION

{CEA CF UNIVERSITY
DISSOCIATION

THE UNIVERSITY 18 COMPOSED OF

INDIVIDUALS AND GROUPS, wORRWNO 17 ESE RHURAS ERIN BISZEI ONVING
ALONE OR TOGETHER, IN DIFFERENT THE MINIMUM ORGANIZATION
DISCIPLINES. WHEN INOIVIOLALS NECESBARY To AN ASSOCIATION OF
WORK TOGETHER THEY TAKE ON DiScrPLINER THE SPECIFIC
NEW CHARACTERISTICS ANO DEVE NETURES OF DIFFERENT FUNCTIONS
NEW REEDS. RARE ACCOMODRTED wiTHIN A
SEVERRELERO rE ORE WHICK
—
+\t/2== lL EXPRESGES UNIVERSITY

d+C=O
—+1:0
Mie
OIBCIPLINES
* COO
JODCIDIGUL
Se \SOLATION OF SPECIFIC DISCIPLINES.
JOLIDOOLE
ASSOCIATION

G, IN SKYSCRAPER TIPE BulLoInGs

DISCIPLINEL TEND TOBE SECAE CEES
THE RELATIONSHIP FROM ONE FLOOR
TO ANOTHER 1S TENYOUS, ALMOST
FoRTUITOUS, Pressing THROUGH
THE SPACE MACHINE LIFT

Fie
v7 aa
SPECIALS st
INFORMATION Se
———
See ____ FLANZ OF ISOLATION

ee
ATOMIZATION OF THE IDER OF
UNIVERSITY

Bur THE REMOVAL OF BUILT BAR RERS

“AND THE MIXING OF DISCIPLINES
3g IN A GROUNO
GROUN: SC RAPER OPGRNITATO
z nN
1® NOT ENOUGH.
QAERTER POSSIBILITIES OF COMM-NITY
THE GROUP 18 MEANINGLESS WHEN
ANOS EXCHANGE e ARE PRESENT
E wiTsour
THERE (8 HO PLACE POR THE
NECESSARILY BACRIFICING ANT
NDIVIOUAL.
TRANQUILITY.

GROUP 1S EVERYWHERE

THE RELATIONSHIP OF CHSLP HOD

WONWIDUML UBT ALSO BE CONSCERED
REM OF ACTIVITY AND BREAD OF 10, TENTATIVE Use oF minimum
TRAN OOILITY HUST BE PROVIDES. STCTURING SYSTEM WHERE INDIVIOUR
\E THE Gaour 18 EVERYWHene THERES BRO GROUP May CETERMINE
HO GAOsP BECHUBE THERE 18 HO CRSIAABLE RELAT ONSHIPS.
INDIVIDUAL.

PLACED FOR INDIVIOUAL =PLACES POR cae

TRAN QUIUTY AND RETVITT
\POLATION AWO ExCHance

Procedural Aspects of Design Thinking 92

35
Heuristics and model for the Berlin Free University
(Candilis, Josic, and Woods, 1963).

Procedural Aspects of Design Thinking 93

 @ The Phenomenon of “Constancy of Appreciation”
To begin with, it is fairly clear that a given rule, analog, or model
will be used in different ways at different times by different prac-
titioners. It follows from this that the clarity and singularity of
application of a particular heuristic derive from the moment. Its
significance in a design process is thus fixed somehow, at least
at a given point in time. Furthermore, at the moment of applica-
tion, the problem solver’s attention is fixed on discrete and spe-
cific characteristics embodied in the heuristic that is being
employed. This is what Schon refers to as the phenomenon of
“constancy of appreciation” (Sch6n 1983, pp. 272-273).

Although attention may shift, at some other moment, to other

characteristics, the use of the heuristic must be discrete, con-
stant, and fixed. Without these conditions a decision could not
be made at all. The relevant information provided by the heu-
ristic is regarded by a problem solver as infallible. For example,
the designer in the first case study in the preceding chapter had
to have been at least momentarily convinced about the efficacy
of the principle for shaping urban space in order to have pro-
posed a continuous fabric of building out of which a “‘place’”’
was cut. This is not to say that once the moment has passed, the
decision may not be revoked. However, quite apart from whether
it was “correct” or “incorrect,” the value of the decision, once
made, to further problem-solving activity is that the outcome
forms the basis for reflection—for what Schon refers to as
“‘back-talk” between the problem solver and the problem situa-
tion (Sch6n 1983, ch. 3). Thus there ensues an episodic se-
quence of steps, wherein at different moments attention is fixed
exclusively on particular aspects of the problem that seem to
warrant consideration.

An interesting question, as yet unresolved, that stems from this

phenomenon of constancy of appreciation is whether and to
what degree there is congruence in the manner in which design-
ers use particular heuristics. Evidence of very real similarities
among examples of architectural and urban design production
at certain epochs in history would suggest a reasonable amount
of congruence. However, for a more lengthy treatment of this
and related issues we must wait until the next chapter.

@ Generation of New Information

One positive outcome of an appropriate course of heuristic rea-
soning is that it productively moves the problem into a new light.
Consequently, both the scope of the problem (the constraining
conditions) and the relative promise of various solution methods

Procedural Aspects of Design Thinking 94

 (courses of action) can be reinterpreted. In short, the reasoning
process can have the effect of providing valuable new informa-
tion. For instance, in the second case study the line of reasoning
about how to structurally support the building immediately led to
a realization about formal design possibilities for the central
corridor.

Clearly, it is the testing and comparison procedures inherent in

heuristic reasoning that provide this new information. As de-
scribed before, it is by comparison of the state of a problem
space at time twith its state under the rule of an organizing
principle at time t + 1 that new information is furnished. Logi-
cally speaking, this can be accomplished in at least three ways.
First, the results of a test can provide information about the con-
formity of proposed solutions vis-a-vis known constraints. In
other words, they can show how close the proposed solution has
come to satisfying the constraints. Second, a test can provide
information regarding overall progress toward a solution; that is,
it can reveal whether the particular line of reasoning appears to
be working when evaluated against global constraints. Finally,
the application of heuristics can provide new information about
other appropriate constraints, not so far considered as being
part of the problem. Of course, the validity and import of the new
information may not be fully apparent until subsequent steps
have been taken making use of the information. As the new in-
formation becomes available, it must be provisionally considered
useful; otherwise the line of argument through which it was re-
vealed would have to be abandoned. In other words, information
about constraint conditions not so far considered may or may
not be appropriate, but it will not be until the conclusion of sub-
sequent operations that the value of the information can be fully
assessed. For example, again in the second case study, we saw
that it was only when the moves associated with compositionally
arranging pieces of the building were almost complete that the
problem of scale and size became apparent.

Sometimes, through either necessity or a lack of confidence in

the outcome portrayed by new information, a line of argument is
abandoned. This abandonment can result in the problem solv-
er’s backtracking through the problem space to a point that ap-
pears to be non-troublesome and promises to allow forward
progress. This topic will be taken up a little later in our
discussion.

Procedural Aspects of Design Thinking 95

 @ Referential Bases of Heuristics
An interesting aspect of many organizing principles is that,
when logically extended, they seem to provide designers with a
self-referential rule structure (Eisenman 1979), a rule structure
largely governed by the subject matter incorporated within the
rule. For example, in the application of an abstract proportional
device, such as a3 xX 3 cube, there are many possible implied
patterns of orthogonal and diagonal divisions that can be ex-
ploited to give rise to a building (as in Eisenman’s ‘‘Houses’’; see
figure 36). Furthermore, the full implications or extent of the pat-
terns may not be fully understood by a designer when the pro-
portional device is first employed, owing to the phenomenon of
constancy of appreciation. Gradually, however, the latent qual-
ities of the model may become apparent, introducing new infor-
mation into the designer’s problem space. In fact, the device
may initially be chosen for a reason that becomes irrelevant as
other, underlying qualities are more fully exploited: once the
new possibilities inherent in the problem have been revealed, the
probiem itself is reformulated, sometimes radically.

If this exercise is reduced to a kind of ‘analog takeover,” involv-

ing only the willful exploration of the latent properties of the
chosen device, then clearly the appropriateness of the exercise
can seriously be called into question. A very real issue that at-
tends the dominant use of autonomous constraints is the super-
ficial enrichment of the problem under consideration. It is an
enrichment that runs the risk of so distorting the problem that it
loses its original meaning and orientation.

@ End Justification
As defined earlier, the structure of the reasoning process used
in applying a heuristic seems to conform to the general argu-
ment, ‘‘If problem X is encountered, then take action Y under
conditions Z.”’ Following from the subsequent discussion of al-
ternative forms of argument, it seems reasonable to propose re-
writing the expression as, “If conditions A obtain, then action B
is taken for the intention of C,”’ where intention C is clearly
bound up with solution of the problem. Thus, we might say that
another referential base, or form-giving power, of a heuristic de-
rives from the correspondence that can be realized between a
designer’s intention, the prevailing conditions of the problem
situation, and a formal action.

In many cases, however, an intention and an action are ex-

pressed synonymously and may precede the perception of spe-
cific conditions. In other words, the intention can be understood

Procedural Aspects of Design Thinking 96

36
Transformations 4, 13, and 16 (House Il, by Peter
Eisenman, 1969).

solely by reference to the action; they become so thoroughly

intertwined that we cannot distinguish one from the other. Fur-
thermore, although certain general boundary conditions may be
in view, they lack specificity. For instance, in the second case
study the designer’s action of using the model type was indistin-
guishable from any (separate) intentional orientation that she
may have had at the time.

Clearly, attention has somehow been shifted here, and a trans-

position has occurred in the structure of the argument guiding
the reasoning process. It now becomes, “If action B, implying
intention C, then (specific) conditions A must obtain.” The solu-
tion of the problem has taken on the character of being ‘“‘end-
justified.’ Nevertheless, if we recall the earlier delineation of an
ill-defined problem and remember the underconstrained struc-
ture of many design problems, at least at the outset, this seem-
ingly perverse strategy begins to appear plausible, if not
necessary. Taken to extremes, of course, it can force us into a
dilemma similar to that encountered with superficial enrichment
of a problem space—viz., a solution in search of a problem.

The Influence of Modes of Representation

Michael Graves has eloquently described the inevitable reciproc-
ity that occurs in an architectural design process between the
act of drawing and the thinking associated with it—between
“the image” and ‘‘the mind” (Graves 1977). The hand moves, the
mind becomes engaged, and vice versa. If we infer from this

Procedural Aspects of Design Thinking 97

 statement that what is understood about a problem comes to us
via representation of solution images, then the role of media,
such as drawings in the case of architectural design and formal
mathematical models in some urban planning applications, be-
comes a very important consideration. We might then ask, what
is it about drawing or symbolic modeling that allows us to dis-
cover things? Or, conversely, how might the medium of expres-
sion actually constrain a design process?

Clearly, in the case of drawings, as Graves goes on to say, differ-

ent scales and levels of precision reveal different qualities of so-
lutions (see figure 37). Referential sketches, for instance, often
have an idiosyncratic, notational quality about them. They are
the ‘‘marking”’ of concepts and conceptual developments, rich in
meaning to some but meaningless to others. Static standard pro-
jections, such as plans, elevations, and sections, may fail to con-
vey certain spatial qualities, such as those time-dependent
aspects experienced as one moves through a building; whereas
other media—computer graphic displays, for instance—may not
be so limited.

On a related note, Kenneth Frampton, commenting on some re-

cent graphic output from contemporary architectural circles, has
observed, ‘‘The veil that photo-lithography draws over architec-
ture is not neutral. High-speed photographic and reproductive
processes are surely not only the political economy of the sign,
but also an insidious filter through which our tactile environment
tends to lose its responsiveness. ... When much of modern
building is experienced in actuality, its photogenic sculptural
quality is denied by the poverty and brutality of its detailing”’
(Frampton 1980, p. 297). To be sure, these statements are part of
Frampton’s critique of the modern technological order’s debase-
ment of our sense of place, of settlement and lodging. Never-
theless, he is also demonstrating that there is a symbolic value in
our choice of media, one that reveals a great deal about what we
regard as important and what we don’t find so important. Con-
versely, as has already been suggested, a medium has a way of
constraining our choices; and its influence is probably just as
marked in the reciprocity between a designer and a drawing in
process as it is when the final rendering is given a more public
and sober review. What is especially disturbing, however, is that
this influence may not involve conscious choice at all.

Ernst Mach once said that a mollusk on a rock at sea could have
no knowledge of Euclidean geometry (Ivins 1973, pp. 7-10)—
that is, that truths assumed by Euclid to be self-evident, univer-

Procedural Aspects of Design Thinking 98

 Section

Perspective

Elevation

37
Eight projections of a simple cubic object.

sal, and beyond experience could in fact be apprehended only

by certain kinds of beings. The mollusk, without fingers or simi-
lar movable tactile devices and deprived of, say, a piece of
straight timber, could hardly be expected to observe that parallel
lines don’t converge! It is quite clear from the history of architec-
ture, and the history of ideas for that matter, that developments
in representational techniques—such as perspective drawings
and the systems of geometry of Pascal and Descartes—have had
a profound effect (Ivins 1973, Perez-Gomez 1983). Such develop-
ments have altered what we can represent, see, and therefore
understand and imagine.

Similar observations can be made in the realm of urban design

and planning. Let us take, for instance, recent technical develop-
ments in the use of statistical models for predicting the likely
choice behavior of individuals in response to a policy, such as
increasing the price of gasoline. Here, a lengthy mathematical
expression of terms, usually linear, is constructed on the basis
of data analyses; each term represents some variable aspect of
choice behavior. For example, they may represent the various as-

Procedural Aspects of Design Thinking 99

 pects of a decision about how to make a journey to work, as de-
scribed in the previous chapter, or the aspects of the choice of
a suburban house. Invariably, the validity and very possibility of
such mathematical expressions are subject to the availability
of data, the kinds of inferences allowed by the particular method
of statistical analysis, and even the assumption of independence
among variables that a linear expression entails. Consequently,
certain variables, or aspects of choice behavior, may be more
likely to be incorporated than others about which only qualita-
tive speculations can be made. The upshot is that the proposi-
tion and evaluation of policy alternatives can be determined to a
significant degree by the technique, or medium, of expression
and representation. Furthermore, in the case just outlined, there
may also be a tendency to incorporate variables that are statisti-
cally relevant, although no really satisfactory nonstatistical ex-
planation can be provided (Ingram 1979). This is tantamount to
analog takeover, in that the use of the medium in its own terms
far outstrips the message.

As with drawing and other representational media in architec-

ture, unquestionably great strides have been made in urban de-
sign and planning circles toward the perfection of technical
precision and flexibility. Nevertheless, it is apparent that the
planner, in the end, sees and understands only those things for
which he or she can provide expression.

@ Transformation of a Problem Space

This discussion has focused so far on the definition and illustra-
tion of organizing principles and other heuristic devices. We
have also scrutinized alternative logical structures for the gen-
eral kinds of argument employed when using these devices and
have noted certain observable phenomena in their application,
such as self-referential qualities, end justification, and constancy
of appreciation. Finally, we have explored the contours of the
apparent limits imposed by the media through which heuristics
are applied. It is now time to pursue the matter of how such
principles are chosen and how switching from one kind of prin-
ciple to another takes place during a problem-solving process. It
is clear that for most design problems of any complexity the
thinking comes within the ambit of different rules at different
times.

Logic and the Choice of Heuristic Rules and Constraints

Often, the choice of rule and the consequent constraints that are
applied in organizing problem-solving activity are regarded as
private matters and receive little attention. In other instances,

Procedural Aspects of Design Thinking 100

 such rules and constraints are assumed to derive from the prob-
lem at hand and are tacitly accepted as such. Nevertheless, it
should be apparent from earlier discussions that several general
logical forms of inquiry and argument can be seen at work in the
choice or devising of a particular rule, and therefore the method
by which constraints are introduced into a problem space. This
section will outline three species of logic—deduction, induction,
and abduction—and the types of design situation to which each
is most likely to apply.

In the application of a rule of the form ‘‘If problem X is encoun-

tered under conditions Z, then make use of action Y,” deduction
is at work when action Y belongs to X and includes conditions Z.
Here the relationship between problem X and action Y acts as
a rule, while conditions Z bound the problem by providing cer-
tain specific parameters to be addressed through action Y. For
example, a problem of building a structure with a minimum
amount of deflection in its members would probably be ad-
dressed through an empirical determination of the size and com-
position of the members, using some general mathematical
expression about loads, spans, and so on, subject to specific
requirements for the problem at hand. The same might be said
for the application of other well-known environmental relations
that address specific classes of problems while admitting varia-
tion in the conditions over which the relation can obtain. Logical
deduction appears to be most valid in well-defined problem situ-
ations, where problem X is taken as a bona fide given, with a
clear expression of ends, and where one deduces the appropri-
ate means by using available rules and principles.

It is apparent from another form of argument presented earlier in

connection with heuristic reasoning that induction, too, is a
plausible direction of inquiry. For example, the expression ‘‘If
conditions Z are encountered, then problem X becomes defined’”’
can be seen to lend itself to an inductive reasoning process,
where one moves from the particularities of a situation to a more
comprehensive conclusion. During architectural design this is
often the case, as when the observation of particular user needs
precipitates formulation of a more general problem and the ac-
tions that stem from it. In the activity analysis described by
Sanoff, for example, building users formulate their own design
problems with the help of game-playing techniques. During the
course of attending to detailed arrangements, the users and de-
signers often perceive larger issues (Sanoff 1977, ch. 4).

Procedural Aspects of Design Thinking 101

 A third form of logical inquiry and argumentation is to be seen
at work in the cases of end justification described in a previous
section. As we have seen, in such cases the action and the prob-
lem definition become intertwined. To all intents and purposes,
the problem becomes defined simultaneously with the proposal
of a solution. It is adduced, in much the same manner as a law is
adduced in the service of explaining natural phenomena (Hem-
pel and Oppenheim 1948). This is clearly not a case of deduc-
tion, as there are not sufficient grounds for the erection of a
deductive logical system. Nor is it a case of induction, as the
reasoning process does not necessarily involve systematic ar-
gumentation from particulars to some generality. Rather, it is an
appropriation from outside the immediate context of the prob-
lem space, used for its promise of providing a higher level of
organization. For instance, a designer at the outset of tackling a
problem in housing may decide to make use of a particular type
of configuration. Furthermore, that type becomes the model
through which the problem is essentially understood and
construed. .

This form of inquiry is closely akin to what Peirce refers to, in his
treatment of the logic of scientific inquiry, as “abduction” (Fann
1970). He describes the principle as follows: ‘‘A surprising fact,
C, is observed. But if a proposition, A, were true, C would be a
matter of course. Hence, there is reason to suspect that A is
true” (Peirce 1965, p. 374). He goes on to distinguish abduction
from deduction and induction on the basis of the fact that ab-
duction includes the case where A and C are distinct from one
another and only become related through the existence of some
appropriated scheme, or “‘view of the world,” that has meaning
for both A and C (Peirce 1965, p. 374f.; see figure 38).

A simple abstract example will serve to illustrate this mode of

reasoning (Handa 1983). Suppose we wish to relate two proposi-
tions, A and C, that appear to belong to two different and dis-
tinct realms, X and Y respectively. Here deduction or induction
by means of a third linking proposition B must be ruled out, as B
must belong to either realm X or realm Y, and both are distinct.
In order to relate A and C, we must extend the two distinct
realms and appropriate an alternative view in which they are no
longer distinct, so that B belongs to both the X part and the Y
part. For instance, let us say that A, B, and C are two-dimen-
sional areas on a plane surface (see figure 39). If A is a subset of
X and C is a subset of Y, then X and Y are different planes float-
ing in three-dimensional space, so to speak, such that A C B

Procedural Aspects of Design Thinking 102

 Premise General case Special case

Deduction

®M) (©,) @
Induction
NGAG Aye ANC #8

Abduction

38
A comparison of deduction, induction, and
abduction.

and C C B cannot exist on a specific two-dimensional plane. If

we allow inclusion of a curved three-dimensional space, how-
ever, another area, shown by B, may exist.

Abduction (or adduction) is not a random process. The appro-

priation that is made must show promise in facilitating problem-
solving activity. Returning to our example, presumably it would
be possible to provide a three-dimensional world D that linked A
and C but had no meaning for our investigation. In that world the
means B for providing a satisfactory relationship between A and
C would have no status.

As we have seen from the case studies and other instances, this
mode of inquiry is very common in design. We often employ
heuristics that allow us to import autonomous constraints into
our problem spaces in order to facilitate further activity. In fact,
in the case of ill-defined and wicked problems abduction is the
rule rather than the exception.

Switching of Rules and Constraints

It becomes evident when one examines the structure of problem
solving in architectural design that the distinction of problem
definition, solution generation, and solution evaluation as inde-
pendent stages does not entirely obtain. Often, the provisional
rules that are employed at once define the problem, incorporate

Procedural Aspects of Design Thinking 103

39
Concepts A and C and their worlds X, Y, B
and D.

solutions, and prescribe means for performing evaluations. To

put it another way, success in applying a rule, particularly when
measured in terms of the solution properties of the rule, is
largely guaranteed. For instance, once a housing problem is
construed as one of elaborating a particular typology, then
evaluation becomes similarly directed and the likelihood of
reaching a solution is relatively high. Nevertheless, in actual
practice, judgments are made outside this self-referential
framework. Particular approaches and organizing principles are
abandoned and backtracking does occur. Clearly, for these
changes in direction to take place, some form of evaluation must
occur outside the immediate context of exercising a particular
organizing principle and therefore beyond the realm of a strict
constancy of appreciation. Setting aside criticism from some
other agent, what might be the mechanism for such switching?
How does one organizing principle relinquish control in favor of
another?

In response to these questions several models might be pro-

posed.'? Let us first consider the case of a (reasonably) well-
defined problem, in which an organizing principle is selected
and applied under conditions of constancy of appreciation, a
solution is proposed, and then the latter is evaluated in the light
of known solution properties. Evaluation reveals discrepancies,
if any, between characteristics of the proposed solution and the
properties required for solution. This information forms the basis
for a means-ends analysis by which another rule is selected or
formulated, and so the problem-solving activity proceeds. Even
under the circumstances of a relatively ill-defined problem, if
certain properties of the required solution are known or are as-

Procedural Aspects of Design Thinking 104

 sumed to hold, a similar process can take place and a switch be
made from one heuristic to another.

In a more extreme case of an ill-defined problem, however,

where the required properties of a solution are undefined, ex-
tremely uncertain, or so vague that evaluation cannot adequately
be sustained, the model begins to break down; the process be-
comes a matter of trial-and-error. The ‘dialogue’ between prob-
lem solver and problem situation (Sch6n 1983, ch. 3) cannot be
productively continued, as evaluation can only be performed
self-referentially—within the frame of reference presented by the
rule itself—and therefore no discrepancies are likely to be
found. Strictly speaking, problem solving cannot continue other
than by simply starting again.

This extreme case becomes less problematic if we admit ‘‘selec-

tive inattention” in the constancy of appreciation that takes
place during the dialogue between a problem solver and a prob-
lem situation. Selective inattention involves calling to mind other
Organizing principles in the process of applying one in particu-
lar. It is a speculative aspect of the dialogue, by means of which
a note is made to oneself of the kind, ‘Gee! | recognize a partic-
ular problem here because of other criteria that | have so far
ignored, but | shall press on for now!” In other words, for an
extreme case where conventional means-ends analysis or like
strategies for making a transition must be eliminated, the switch-
ing model might be expressed by the recognition of a new prin-
ciple that is related in some way to the old.

With this formulation as a plausible explanation of the lines

along which switching occurs, the issue now becomes what
might determine the selection of successive organizing princi-
ples. What lies behind selectivity of inattention? Is it a case of
parallel streams of argument, vying with each other in some
way? Or is it just a matter of random association?

To the extent that it can be observed, the selection of successive

organizing principles may appear in a given instance to be quite
idiosyncratic. It is unlikely to be random, however. For if it were,
convergence on a solution of any kind would not happen with
the consistency that is apparent from observation. As we saw in
the case studies in chapter 1, the designers proceeded more or
less directly toward solutions after some adjustment and
reorientation.
The parallel-stream concept might appear to be supported by
frequent remarks on the part of problem solvers in action, of the

Procedural Aspects of Design Thinking 105

 type, ‘‘Now, | can see two possible alternatives,” or “I can think
of several ways to go.’’ Yet under the constancy of appreciation
that is required when one pursues a particular strategy and line
of reasoning, even with sufficient selective inattention to note
other possibilities, ‘‘parallel processing” is probably not occur-
ring. Instead, problem-solving behavior would seem to be serial,
moving in sequence from one line of argument to another, with a
second line of argument perhaps noted during pursuit of the
first. This claim is supported by substantial experimental evi-
dence. In effect, observations on the part of problem solvers
such as ‘‘I can see two alternatives” are better explained as tak-
ing place at the moment of switching organizing principles or of
considering such an action, than as signifying the pursuit of two
directions at once.

Another rationale for switching from one principle to another

might be found within the information contained in an initial line
or argument, without resort to selective inattention or some
other similar mechanism. By simply adopting the opposite, or
“not,” properties to those rendered by one (failing) line of rea-
soning, the solver could formulate a plausible alternative strat-
egy. For instance, when difficulties are encountered with the use
of symmetry in making a composition, then we might change to
the use of asymmetry. In other words, the very proposition of
certain specific properties through an organizing principle may
imply the meaningful existence of opposite properties within the
context of essentially the same rule. This strictly dichotomous
mode does not, however, account for the attention that is seen
to be paid toward other, nonoppositional properties, even during
moments of applying a particular rule. We have seen an instance
of this in the designer in the first chapter who shifted from an
organizing principle about engineering structure to one about
circulation.

Yet another explanation for the switching that occurs might be

proffered, one that is perhaps closer to the mark. If selective in-
attention when applying a particular rule means calling to mind
other procedures, possibilities, or solution images, we may use-
fully regard such organizing principles and heuristic devices as
having properties that allow relations to be struck with other
principles and devices. Here a sufficient overlap of properties
among otherwise distinct organizing principles would allow
such connections to be established in the mind’s eye of a prob-
lem solver. For example, in the second case study described in
chapter 1, the designer began by referring to a specific building
type as a model for the overall organization of a program of ac-

Procedural Aspects of Design Thinking 106

 tivities and for the manner in which they might be architecturally
expressed and resolved. This led, perhaps naturally enough, to
more general speculation about the very idea of combining parts
of a building, thence to consideration of other ways of ad-
dressing the issue of ‘‘combining of parts,” and finally to a
thoroughgoing functional analysis of basic programmatic con-
siderations. What might have appeared to be a fairly radical
jump in procedure—from exploration of the formal architectural
features of a building type to a ‘‘first-principles” examination of
adjacency requirements among the areas of activity to be ac-
commodated in the building—was “in the cards” from the out-
set. In this particular instance the switch was occasioned by a
realization on the part of the designer that the building program
and its parts were too cumbersome to be resolved in the formal
architectural terms inherent in the building type that had been
originally adopted. This realization, in turn, gave rise to a
generalization of the dilemma to one of “parts and their combi-
nation,” and hence to a functional analysis of required spaces
and desirable adjacencies between them. Thus, the perception
of common underlying dimensions of organizing principles may
well account for the lines along which switches from one heu-
ristic to another take place, particularly in cases where the gen-
eral thrust of the original organizing principles seems appropri-
ate but the particular procedure itself is found to be practically
inadequate.

Switching among organizing principles appears to occur not

only when difficulty is encountered but also more casually under
successful pursuit of a particular procedure. Here the purpose,
or result, is the enrichment of a problem space and the supply of
useful additional information. In the protocol analysis just men-
tioned, the designer, having reached the realization of the gen-
eral issue of combining parts, noted that this also meant
something both about the structural support of the building and
about the functional and social purposes to which it would be
put.
@ Order of Application of Rules and Backtracking
When tackling complex problems, designers seem to switch
among heuristic devices as the occasion dictates. Rarely, if ever,
does one organizing principle suffice and provide all the infor-
mation necessary to resolve a problem. It may be seen, however,
from protocol analyses of designers in action, such as those pre-
sented in the first chapter, that the order in which heuristic
devices are applied significantly affects the final outcome of the
proposals that are made. Generally, it is the early attempts at

Procedural Aspects of Design Thinking 107

 organizing the problem space that have a prevailing influence
on subsequent directions, regardless of difficulties that might be
encountered later on (Simon 1970, p. 7).

In light of the foregoing account of problem-solving behavior,

this tendency is quite understandable. The initial structuring of
a problem is the basis on which that problem becomes under-
stood, and thus it is the vantage point from which fruitful
avenues toward solution are sought. Experimental evidence
suggests that the guidance provided by early ordering of a
problem space is relatively long lasting and difficult to abandon
(Simon 1970). This phenomenon was particularly noticeable in
our third case study, as the persistent recurrence of the two
major organizational themes.

Subsequent lines of inquiry are more likely to be abandoned as

unfruitful (that is, not relevant) or intractable. Usually the ac-
knowledgment of their failure triggers a backward pass through
the problem space to a more propitious orientation with a higher
perceived chance of success. This is the phenomenon known as
backtracking, with which we have already become somewhat ac-
quainted (Newell, Shaw, and Simon 1967, p. 72).

Backtracking is most clearly illustrated in relation to a decision

tree, as shown in figure 40. (In most reasonably complex design
problems, of course, a complete representation of the problem
space as a decision tree would be very extensive and well be-
yond casual enumeration of the possibilities by the problem
solver, to whom many of the branches and nodes might be invis-
ible.) Here, the initial pass through the problem space is given by
the link-node sequence aj, b;, Co, d3, e;, and difficulty is encoun-
tered at node e;. Backtracking occurs to the point cz before a
forward pass is made through nodes d, and és.

The essential difference between backtracking and the kind of

switching discussed in the preceding section is one of direc-
tion: in switching there is forward motion through the problem
space during successive attempts to restructure the space to-
ward a solution, whereas in backtracking—which may be re-
garded as a particular case of switching—there is a retreat to an
earlier understanding or construal of the problem. Backtracking
may not involve switching entirely from one organizational de-
vice to another. Instead, a particular line of reasoning may have
choice points embedded within it, among which there might be
movement in search of better opportunities, but without neces-
sarily suspending the overriding organizational concept that is
being applied. Indeed, backtracking rarely involves a radical re-

Procedural Aspects of Design Thinking 108

 40
A diagram illustrating backtracking through a
problem space.

thinking of a problem, other than in the exceptional case where

difficulties are encountered early on in the process, say at
choice point b, in figure 40. Clearly, here reappraisal will lead to
a different trajectory through the problem space. Another inter-
pretation of this kind of behavior, however, is that it is quite
characteristic of an initial reckoning with a problem, when vari-
ous points of attack, each valued more or less equally, are tested
before a commitment is made to sustain one in particular. Early
skirmishes with a problem during which various lines of attack
are very provisionally assayed are quite familiar among
designers.

@ A Concept of Style
Finally, as noted by Simon and others, this kind of theoretical
appraisal of design offers the prospect of a concept of style (Si-
mon 1970, pp. 9-10; Gombrich 1965). It is a concept based not
on the classification of various physical features of architecture
and urban design but on the problem-solving process itself.

We have seen that the final outcome of a design process is

strongly determined by at least three aspects of that process: the
subject matter of the organizing principles that are adopted, the
manner in which these principles are interpreted and reinter-
preted in the context of the problem at hand, and the sequence
of applying such organizing principles. Consistency in style

Procedural Aspects of Design Thinking 109

 among the output of designers can thus be understood as a
habitual way of doing things, of solving problems. The tendency
for this consistency to be most pronounced during particular
times in a designer’s career, say toward the end of ‘‘stylistic”’
episodes, is also understandable. A fluency in a particular way
of designing, and the consistency that comes with it, can only be
reached through experience and constant development. It also
seems to coincide with conditions under which Merleau-Ponty’s
“sedimented principles’ lead to enhanced know-how. At such
moments the potential of certain preferred organizing principles
and enabling prejudices has been more fully realized, and they
become something of a personal form of architectural expres-
sion. For example, Stirling, like many other great architects, has
developed a distinctive repertoire of techniques for dealing with,
say, interior spaces. In his work we see the use of ‘‘porthole”’
windows to provide visual access without overly interrupting
wall planes, the use of “‘pullman-car’”’ cone lighting, and so on.
By extension, style in the broader community sense may be re-
garded as congruent with the collective adoption of certain or-
ganizing principles and practices—that is, with forms of
heuristic reasoning. When dominant principles and forms no
longer prove productive, they are replaced, and shifts in style—
in habitual ways of doing things on the part of the design com-
munity—can be observed.

RE
iinae
Limitations of a Procedural View

As Bruner notes in a discussion about creative problem solving,

there is a certain ‘‘shrillness to our contemporary concern with
creativity’ (Bruner 1967, p. 26)—a shrillness, or heartfelt need,
that he believes stems from our search for sources of dignity in
what we do. ‘‘In an age where dominant values are pragmatic
and whose achievement is an intricate technological order .. . it
is not sufficient to be merely useful... . Creativity seems to im-
bue acts with dignity” (Bruner 1967, p. 2).

in architectural circles during the 1960s and early 1970s, the

shrillness of the concern for the creative problem-solving pro-
cess that Bruner refers to became very apparent, although to-
ward somewhat different ends. Armed with the emerging
information processing theory of problem solving and earlier
rigid-state models, the design disciplines sought to describe and
exemplify the design process (Thornley 1963, Gregory 1966,
Broadbent and Ward 1969, Jones 1970, Broadbent 1973, Wade

Procedural Aspects of Design Thinking 110

 1977). As Mitchell describes their chronology, these efforts were
less the result of theoretical developments in the field of prob-
lem solving than a reaction to the apparent disintegration of ex-
isting techniques and orthodoxies, coupled with a rise in
pressure for higher degrees of user participation (Mitchell 1970).
In this latter regard, what seemed necessary was a clear and log-
ical procedure for producing designs and plans that could be
understood and participated in by all those involved. With re-
spect to the former point, what seemed necessary was a far
greater understanding of design processes, in order that proce-
dures could be improved. In spite of the very real contributions
that were made, at least to our understanding of these pro-
cesses, in almost all cases the step beyond description into a
normative realm in which process became pursued as an end in
itself resulted in abject failure. Attempts to devise the process
became exercises in inanity when compared to the great subtlety
and profundity of observed problem-solving behavior. This
movement, in the view of many, eventually resulted in a form of
“methodolatory”’ (Colquhoun 1967); it is currently experiencing
a relative hiatus, with the possible exception of developments in
the field of computer-aided design, mentioned earlier.

Quite apart from the prescriptive orientation of the design

methodologists, the program of converting description into ex-
planatory theory about what designers do can be seen to foun-
der in another way. So far, what we have discussed about design
could probably just as well be said, at least in large measure, of
some other realm of inquiry. In fact, most adherents of the infor-
mation processing theory of problem solving would make such a
claim for this kind of speculation. It is intended to be general
and, as such, applicable to many different fields of inquiry. There
is no reason to believe that architects and urban designers at-
tack problems in fundamentally different ways from, say, scien-
tists and mathematicians, at least at this level of theorizing.
Common kinds of means-ends analysis, heuristic search, prob-
lem-space planning, and other procedures can be seen at work
across disciplinary boundaries (Perkins 1981, Ghiselin 1952,
Polya 1954, Simon 1973). The fact remains, however, that we
don’t have much difficulty distinguishing what architects do
from the activities of psychologists or doctors, and have only a
little more difficulty with distinctions between, say, architects
and painters or between architects and planners. Similarly, how
these various fields actually conduct their inquiries may be less
well known, but discrimination is still possible. It would seem,
therefore, that such distinctions can be drawn, but as much

Procedural Aspects of Design Thinking 111

 through the subject matter embodied in the principles of prac-
tice as through the structure of inquiry. Furthermore, these are
principles that derive not only from the kinds of problems being
worked on at the time but also from the frames of reference,
attitudes, and perspectives that are more generally adopted. In
short, they come from another realm of theoretical speculation.

At this juncture we might say that design, like other disciplines,

involves a kind of procedural knowledge—that is, both tactical
understanding and know-how—and a kind of substantive knowl-
edge outside the procedures themselves (Sch6n 1983, Ryle
1949). Clearly the two kinds are related; indeed, they are inextric-
ably intertwined. A carpenter, for instance, not only has a
“knowing-how” understanding of procedures such as sawing,
nailing, and routing but also has a ‘‘knowing-that’”’ understand-
ing of the circumstances under which such procedures should
be applied. Through the reflection of the moment, that know-
how and knowledge of criteria are applied at once, constituting
a judgmental action (Schon 1983, pp. 50-69).

With this definition of its constitutive elements, we are brought

more squarely on the question of where next to turn in rendering
an account of practical inquiry in architecture and urban design.
Both an explicit and an implicit part of our discussion so far has
been the idea that structural aspects of a problem-solving pro-
cess, of which we have enumerated nine or ten, significantly de-
termine the outcome. Let us assume, however, that we are given
an elaborate protocol analysis of several problem solvers in ac-
tion, each in the same problem situation. From what has been
said, how might we go about commenting upon the relative ap-
propriateness of each outcome? In other words, how might we
address the issue of adjudicating merit?

First, we could use the protocol analyses, in conjunction with

the structural considerations that have been discussed, in order
to comment upon the comparative technical skill and logical
qualities with which the problem solvers have addressed the
problem situation. With very little extrapolation we should be
able to discern whether the self-referential qualities of various
organizing devices have been fully realized. We could make
comment on the relative efficiency of one approach over an-
other, by reference to episodes of backtracking and other kinds
of switches among the principles employed. We could also prob-
ably detect patterns of more or less logical reasoning and dis-
cuss the extent to which this appears to have influenced
outcomes.

Procedural Aspects of Design Thinking 112

 Second, we could measure the extent to which each proposed
solution appears to satisfy the solution properties of the problem
as given. Here, connections might be drawn between the various
aspects of the problem as stated and the manner in which they
were addressed in the proposed solutions. From this a critical
evaluation could be rendered with respect to the comprehen-
siveness and logical precision of the solution.

Third, we could compare the manner in which the problem has

been construed by each problem solver with the manner in
which it was given. We might ask whether the proposed solution
seems overly end-justified, so that it is a response but not neces-
sarily to the stated problem. On the other hand, we might be in a
position to discriminate instances in which the constraints that
were autonomously supplied to the problem space were never
fully realized and exploited, thus raising questions of redun-
dancy and superficial enrichment of the problem space.

Beyond these general points, however, little can be said without

venturing into comparisons of the appropriateness of the or-
ganizing principles or heuristics themselves. This would become
particularly apparent in the case of those principles that are in-
dependently supplied to the situation, where differences among
problem solvers are apt to be greater than in instances of stricter
problem definition. To achieve an adjudication of merit that tran-
scends personal likes and dislikes would also seem to require
inspection, discussion, and debate about the perspectives
through which organizing principles and constraints are sup-
plied. Otherwise, important issues of applicability, appropri-
ateness, opportunity, and real enhancement of a problem cannot
very well be broached. Clearly this is in large measure a norma-
tive terrain concerned with values and aspirations. Therefore it is
to the normative perspectives that supply and shape organizing
principles that we should turn in order to develop a more thor-
oughgoing account of practical inquiry in architecture and ur-
ban design. This will be the task of the following chapters.

Procedural Aspects of Design Thinking 113

 Chapter
Normative Positions That Guide
Design Thinking
 For architects and urban designers, a major source for the per-
spectives that guide the choice of organizing principles and con-
straints is the realm of theoretical discourse. Theory is assumed
to be about general principles with applicability beyond specific
cases, and, whether it comes by way of systematic speculation
and codification or by way of more indirect experience, to be
well substantiated. To the extent that it has a community of sub-
scribers, theory represents a corpus of principles that are agreed
upon and therefore worthy of emulation.

Normative Positions

A quick perusal of published compendia of various theoretical

positions in architecture will attest a rhetoric that is usually cen-
sorious and acerbic, and ineluctably concerned with ‘‘what
ought to be” (Conrads 1964). Such writings contain essential
statements of values and thus of priorities among concerns per-
taining to architecture. The thrust of the messages is about the
location and merit of norms to be categorically valued and as-
pired toward in the practice of architectural design. Further-
more, most positions can be said to address in some way the
ontological question ‘‘What is architecture versus something
else?’’— or, “What constitutes proper architecture?”’

Sir Nikolaus Pevsner’s response to this question is the cele-

brated statement in the introduction to An Outline of European
Architecture: ‘‘A bicycle shed is a building; Lincoln Cathedral is
a piece of Architecture.” A rationale follows: ‘“‘Nearly everything
that encloses space on a scale sufficient for a human being to
move in is a building. The term architecture only applies to
buildings with a view to aesthetic appeal’ (Pevsner 1961, p. 1).

This definition has its limitations, of course. In it architecture is

portrayed as a subset of building, and therefore built objects that
don’t enclose space sufficient for human activity—facades,
some bridges, and so on—are erased. More important, aesthetic
appeal and not, say, utilitarian considerations is the essential
and sufficient criterion for architecture. The point here, however,
is not to take issue with Pevsner’s definition, but rather to indi-
cate the kind of direction we shall be looking in for a reasonably
consistent idea of theory among competing normative positions.

In this examination we shall follow a similar method to that pur-

sued with regard to problem-solving behavior. Classes of posi-
tions rather than individual positions will be defined, and their
 implications for design assessed. These classes of theoretical
positions will be examined for their qualities of logical coher-
ence and for the substantiation on which they base their claims
to legitimacy.

SES
a TT eI |
Surface Features and Broad Inclinations

Statements of normative positions can generally be found to in-

clude the following elements: (1) the location and identification
of a problem or of pertinent issues under contention, (2) an un-
favorable assessment of prevailing practice and an enumeration
of untapped opportunities, and (3) a counterproposal with its
rationale. Further, the proposal and rationale are typically
founded in the distinctions made between the prevailing prac-
tices and the latent opportunities. For example, Hannes Meyer,
addressing the Bauhaus in February 1928, outlined his thesis on
“building” in the following manner:
all things in this world are a product of the formula: (function
times economy)
all these things are, therefore, not works of art:
all art is composition, hence unsuited to achieve goals.
all life is function and therefore unartistic.
the idea of the “‘composition of a harbor’ is hilarious!
but how is a town plan designed? or a plan of a dwelling? com-

(Meyer 1928, Wingler 1969, p. 153)

This stanza contains a critique of what was perceived to be pre-

vailing practice—namely, architecture as art, art as impractical
and not of this world. It also contains statements concerning the
opportunities for otherwise addressing the issue of building and
architecture—that is, by way of function. Furthermore, the state-
ments locate the problem as being a fundamental question
about how to design and build.

Meyer goes on to elaborate a counterproposal by presenting

building as a ‘‘biological process”’ and calling for the organiza-
tion of new building materials “into a constructive whole based
on economic principles.’’ Completing the biological analogy, he
asserts, ‘‘the individual shape, the body of the structure, .. .
evolve by themselves and are determined by life’ (Meyer 1928;
see figure 41). His position is presented in an almost syllogistic
form. The argument links the location of the problem, through a
critical assessment of existing approaches and a survey of other

Normative Positions That Guide Design Thinking 116

 J,y,
Wi4

41
Project for the Peters School (Petersschule) in
3asle, by Hannes Meyer (1926).

opportunities, to a counterproposal. Whether or not we agree

with it, the thesis has a compelling logical structure.

In Le Corbusier’s ‘Five Points towards a New Architecture,” to

use another well-known example, the line of argument is not
quite so clearly set out. Nevertheless, the opening paragraph
provides a summary of its general tenor, as well as an indication
of the perceived problem. ‘‘The following points,’ announces Le
Corbusier, ‘‘in no way relate to aesthetic fantasies or a striving
for fashionable effects, but concern architectural facts that imply
an entirely new kind of building’ (Le Corbusier and Jeanneret
1926; Conrads 1964, p. 99). Here, a negative judgment is implicit
in the phrases “‘aesthetic fantasies” and ‘‘fashionable effects.”
Elaboration of the five points follows, largely aimed at justifying
their autonomy and viability on technological and functional
grounds. The five points are (1) supports, (2) roof gardens,
(3) free design of the ground plane, (4) horizontal windows, and
(5) free design of the facade; held to be the necessary ingredients
of ‘‘proper’”’ architecture, they are presented as inherently new
opportunities for building and ones that are more intrinsically
appropriate to the conditions of the time (see figure 42).

Normative Positions That Guide Design Thinking 117

42
The Villa Stein—De Monzie (Maison Stein) in
Garches, by Le Corbusier and Pierre Jeanneret
(1927).

Le Corbusier concludes the argument by asserting that “‘the

essential points .. . represent a fundamentally new aesthetic.
Nothing is left to us of the architecture of past epochs” (Conrads
1964, p. 100). Although he omits several steps linking the
counterproposal of the five points, and their rationale, to the per-
ceived problem of socially disconnected, historicist ‘aesthetic
fantasies,” the syllogism can be readily filled in by implication.

A more vexing subject for this kind of analysis and framework is

the credo or the aphorism, sometimes drawn from architecture’s
oral tradition. For example, there are Van der Velde’s ‘‘Credo”
(Van der Velde 1907), Kahn’s poetic invocation “Order is...”
(Kahn 1961), and Mies van der Rohe’s oft-misquoted ‘“‘beinahe
nichts” —‘‘almost nothing” (Mies van der Rohe 1923).' Despite
their idiosyncratic use of language, the same kind of argumenta-
tion can still be discovered in these statements, if one is willing
to place them in context and engage in some interpretation.
Their effectual status as expressions of normative positions is
attested by the reverberations they have caused in architectural
circles.

Normative Positions That Guide Design Thinking 118

 On a more prosaic level, a considerable body of commentary has
been developed around the central question posed by ‘‘form fol-
lows function.’’? This commentary moves beyond the emblem-
atic qualities of the aphorism and stands at the center of a
long-standing debate regarding these two architectural
categories.

The import of vague or brief statements can be just as rigorous

and compelling as that of more complete dissertations. For our
purposes, therefore, any serious statement of a normative posi-
tion, regardless of its surface qualities, should be admitted for
further scrutiny. It will be useful, however, to differentiate among
these positions according to one of their surface qualities: their
location on a continuum that ranges from what will be called,
after Attoe, doctrinaire positions at one extreme to what will be
called categorical systems at the other (Attoe 1978).
@ Doctrinaire Positions versus Categorical Systems
In the extreme case doctrinaire positions will coincide with the
assertion of a singular point of view, attitude, and approach that
is of primary if not sole importance for defining “‘proper”’ archi-
tecture and for guiding production. Such positions may take
several forms, as can be seen from the foregoing discussion.
They may involve a general invocation of a guiding principle,
aimed at influencing the attitudes of designers. For example,
there are the positions closely aligned with certain construals of
aphorisms such as “‘less is more,” ‘‘form follows function,” and
so on. They may also involve a prescription of those architec-
tural elements that should be given primacy. Le Corbusier’s
“Five Points” exhibits this characteristic, as does Hannes
Meyer’s proposal in its entirety.

By contrast, categorical systems may appear on the surface to

be less definitional and prescriptive regarding their requirements
for proper architecture. Nevertheless, they are frameworks nec-
essary for answering what proper architecture is, versus some-
thing else. They are ostensibly less singular in viewpoint than
doctrinaire positions. They also have the distinguishing feature
of possessing some explicit external model that elaborates, con-
nects, and sustains norms and categories for distinguishing
what counts from what does not count. The model is external in
the sense that it has a coherent existence ouside the environ-
ment under immediate scrutiny. In practical inquiry the model
provides the basis for assessment and guidance (Norberg-
Schulz 1965).

Normative Positions That Guide Design Thinking 119

 Categorical systems have a long and noble tradition. They can
be traced through the theoretical proposals of such luminaries
as Vitruvius and Alberti to the more contemporary efforts of
scholars such as Norberg-Schulz (Vitruvius 1940, Alberti 1955
[1726], Norberg-Schulz 1965). Even if these works do not entirely
live up to the foregoing definition, they certainly seem to lean in
a systematic rather than doctrinaire direction.

The major features of a categorical system are illustrated by the

early theory construction of Norberg-Schulz—specifically, by his
“integrated theory of architecture,” as propounded in /ntentions
in Architecture. An ‘‘architectural totality,’’ as he calls it, is com-
pletely described by definitions in three categories, or dimen-
sions. These are the ‘‘building task,” ‘“‘form,’’ and “‘technics.” A
particular architectural totality at time t is regarded as an “‘ar-
chitectural system,” its style corresponding to the mode of for-
mal organization. Independence is claimed for these broad
categorical distinctions: as Norberg-Schulz puts it, ‘‘The totality
of a building task is realized technically within a style’ (Norberg-
Schulz 1965, p. 87). The three dimensions, then, serve as useful
abstractions for thinking about architecture and for distin-
guishing it from something else.

The definition of the building task roughly follows Paulsson’s

distinction between physical and symbolic aspects of the built
environment. It is assumed that the building task is equivalent to
the provision of shelter and physical control on the one hand
and the provision of a “‘frame for actions and social structures”
on the other (Norberg-Schulz 1965, pp. 109-130).

The dimension of form is essentially concerned with the depic-

tion of “elements” and ‘relations’ (Norberg-Schulz 1965, pp.
131-160). Here, elements are the characteristic units that make
up an architectural form. Norberg-Schulz tends to regard them
in terms of ‘‘space cells,” “‘mass forms,” and ‘boundary sur-
faces” (p. 97).° Relations, on the other hand, are tridimensional,
or spatial, and represent a “‘lawful” way of distributing elements.
Thus, “formal structures” and ‘‘style’’ denote various types or
groupings of elements and relations, as well as the rules and
lawful procedures. In other words, formal arrangements consti-
tute a formal language, out of which emerges the style of the
architectural production.

The dimension of technics is concerned with how technical

components are made from materials and organized into ‘‘tech-
nical systems”’ (p. 161f.). Furthermore, the ‘‘capacity” of such
systems is open to scrutiny—that is, the ability, technically

Normative Positions That Guide Design Thinking 120

 speaking, to realize a building task. A technical system is not a
symbol system; the relationship between building task, form,
and technics is that the form always mediates between the task
and its technical realization, or the means that are brought to
bear.

What Norberg-Schulz presents us with is an almost Vitruvian

tripartite schema of related yet independent dimensions and
categories of proper architecture. The framework is ostensibly
analytical and concerned with the definition of appropriate
categories and their relations; it is, then, a categorical system.
It is ostensibly analytical because in the selection of sub-
categories, such as formal elements, the definitions border on
prescription of certain features at the expense of others. For
example, mass forms are accorded a relatively special status
(p. 133f.). Nevertheless, the scheme is a far cry from the defini-
tive, prescriptive pronouncements of true doctrinaire positions.

The differences between doctrinaire positions and categorical

systems are by now, it is hoped, clear. The principal dimensions
along which distinctions can be made are (1) the singularity of
viewpoint; (2) the explicit nature of the external model of norms
at work in promulgating the position; (3) the systematic level of
explication, a corollary of (2); and (4) the scope of the rationale
provided for the theoretical propositions. In connection with this
last point, it should be noted that Norberg-Schulz provides ex-
tensive and well-argued support for his categorical system,
building on prior works of a similar kind. This is particularly evi-
dent in earlier sections of the work, where background material
is presented and a conceptual framework is established.

All positions, of course, can be seen as ideological or biased to

some degree. They clearly favor one set of circumstances over
others. If they are to be coherent, all must contain some line of
argument. And if they are to hold designers’ attention, all are
required to sustain claims of legitimacy and centrality. This
statement of the obvious has the intended function of holding at
bay the probable judgment that categorical systems are prefera-
ble to doctrinaire positions on the grounds that the former are
more overtly comprehensive and well organized. For, as the ex-
amples of Meyer, Le Corbusier, and others have shown, a sense
of explicitly defined comprehensiveness and organization is not
necessarily the sine qua non for rendering such judgments.

@ Normative Positions and Architectural Categories

The assertion that normative positions, as so far described, pro-
vide much of the basis for architectural design requires little

Normative Positions That Guide Design Thinking 121

 defense. Leading practitioners, educators, and critics each
espouse normative positions of some sort. To deny the connec-
tion would be perverse in the extreme, as it would amount to
saying that what these people profess and what they do in ar-
chitecture are entirely unrelated.

Even theoretical positions that may appear to be otherwise in-

clined are really of the same stripe. For example, Hillier, Mus-
grove, and O’Sullivan’s propositions revolving around a building
acting as a ‘climate modifier,” a ‘‘behavior modifier,” a ‘‘re-
source modifier,’ and so on, may incorporate different kinds of
norms and architectural categories from those discussed, but
they are norms and categories nonetheless (Hillier, Musgrove,
and O’Sullivan 1972). Alexander’s opus, ranging from Notes on
the Synthesis of Form to The Linz Cafe, although it may take on
the guise of ‘scientific’ explanatory theory, fundamentally es-
pouses a consistent set of norms and categories (Alexander
1964, 1981). They are norms founded in a strong sense of behav-
ioral determinism and a popular or consensus view of what is
proper. Throughout Alexander’s work there is also a very defi-
nite orientation toward practical guidance and a strong convic-
tion of the possibility of defining immutable and foundational
categories.

Even ostensibly mundane building codes and occupancy stan-

dards must be included. Such texts are also normative and set
squarely in the realm of practice. Clearly, examples other than
those used might have been discussed in a similar manner; the
range of surface features and the essential nature of the inherent
structure of argumentation could have been shown to conform
sufficiently to the scheme that has been advanced.

[sas Simoneau eo aig reece ie eee

Further Differentiating Features

Beyond the doctrinaire or categorical inclinations and other sur-

face features of normative architectural positions, a number of
other differentiating characteristics can be observed. For in-
stance, there is the general orientation, or program of investiga-
tion. As we saw in the last chapter, Anderson describes the early
work of Le Corbusier in terms of several instrumental ideas, or
“programmes,” in which the implications of various related de-
sign concepts were explored and refined over time. There is also
the matter of preferred spatial concepts, to the extent that they
can be discerned independent of orientation. For example, the
use of a formal type as a guiding idea, or ideal, may involve a

Normative Positions That Guide Design Thinking 122

 quite different spatial concept from one predicated on a rela-
tional model of functional requirements. In addition, there is the
degree of emphasis accorded to the context outside matters of
purely architectural expression (such as the long-term use and
institutional sponsorship of a building). Finally, most well-devel-
oped architectural positions exhibit a multiplicity of these char-
acteristics. Designers are at once concerned with exploring
influential programs, favoring particular spatial concepts in the
expression of their ideas, and addressing the general social con-
text and the object quality of the work. This last aspect will be
developed further in the next chapter.

@ An Analytical Framework
For the purpose of discussing these characteristics of architec-
tural positions, let us imagine a simple taxonomic construct that
allows aspects of various candidate positions to be commonly
portrayed. The aspects in question are production, architectural
devices, and orientation. Production shall refer to visible works
under some commonly expressed or descriptive term that iden-
tifies the kind of architectural output of the position—in other
words, a ‘label,’ stylistic or otherwise, that is or might be used
to identify the group of practitioners and theorists (‘‘Brutalists,”’
for example). Architectural devices will refer to kinds of ar-
chitectonic elements favored or prescribed by the position, as
well as the use of other less tangible leitmotifs that operationally
describe the position’s production. For example, Le Corbusier’s
five points are clearly architectonic elements or devices, and
ones that are just as clearly prescribed for the production of
proper architecture. Finally, orientation will cover the critical
stance and larger (social) purpose of the position, or one that
might legitimately be imputed to that position.

In addition to incorporating these three types of information, let

us further enrich the construct by regarding it roughly as a line
of argument in which statements about orientation support pre-
scriptive devices, and so on:

Orientation — Architectural Devices — Production.

Generally, among the practitioners who may be categorized as’

more or less belonging to a particular position, a reasonable
amount of agreement is likely to be found on matters of orienta-
tion. Considerable variation, however, is likely to be seen in the
choice of architectural devices, and a very real contrast is often
evident in the conformation of the actual work. With these obser-
vations in mind, we are now in a position to fill in the analytical

Normative Positions That Guide Design Thinking 123

 construct with specific information from some normative
positions.

@ Application of the Framework

Four positions, or classes of positions, have been chosen, but
only for purposes of illustration. This is certainly not an attempt
to conduct a comprehensive survey, or a survey of any kind for
that matter. The choice of positions reflects four strains of ar-
chitectural thinking that are highly visible in contemporary
theory and practice; but they have existed in the past, too, and
will no doubt exist in the future. Further, when examining each
position, it is not proposed to conduct a critique of specific ex-
amples, although how at least one kind of critique might be
performed will probably be clear from the application of the
analytical framework just described. Finally, a certain amount of
caricature and overlap is probably unavoidable in this kind of
contracted analysis. Nevertheless, a sense of the broad orienta-
tions of the positions in question can be maintained in a reason-
ably defensible fashion, or at least the strains can be seen to
appear and sometimes commingle in much of contemporary
production.

The four positions that will be briefly examined are as follows.

First, there is a functionalist position, distinguished by an em-
phasis on the accommodation ot activities and the influence of
building technology. Second, there is a populist position, char-
acterized by an acknowledgment and interpretation of contem-
porary commonplace building practices and user preferences.
Third, there is a conventionist position, using an architecture of
largely historical reference; and finally, a formalist position, us-
ing an architecture of formal possibilities for their own sake.

The orientation of the functionalist position may be described in

the following way. Architecture is a matter of accommodating
the functions that are prescribed for it and of functioning in a
manner that is consistent with its material composition and con-
struction. “Form follows function” is the guiding doctrine on
both matters of use and building technology. Architecture must
be made of the ‘‘right stuff’ and in the “right way” without
superfluous ornament or artifice. It must also work efficiently, in
the sense of directly conforming to some ideal of use require-
ments; it must, as Le Corbusier put it, be a “‘machine to live in,”
with all the parsimonious aspects of machine design (Le Corbu-
sier 1959, p. 263f.; 1967, p. 204).

The specific architectural devices of the position are numerous

and varied. Often, however, explicit expression of a building’s

Normative Positions That Guide Design Thinking 124

 essential structure and process of fabrication are hallmarks.
Spatial organization invariably grows out of the program of uses,
in a straightforward manner (Herdeg 1983). Ulrich Franzen’s Al-
ley Theater of 1965, for instance, presents use with a diagram of
functional organization, boldly clad in ‘‘off-form’ concrete. The
outlines of both major spaces and circulation elements (such as
stairways) are clearly visible (see figure 43). Frequently there is
an underlying concern with standardization and systematic or-
ganization on a number of counts. First, a building is seen to be
made up of an assemblage of parts, each under the discipline of
rational (that is, functionally systematic) organizing principles.
Consequently, the parts themselves tend to become standard-
ized, their variation minimized, and their assemblage conceived
as a “‘building system” (Dietz and Cutler 1971). Second, ‘‘use
standards,” representing so-called optimal spatial configura-
tions for accommodating human activities, are employed. Fre-
quently these standards are ergonometrically determined, and in
some settings (Such as mass housing) there is an effort to hold
to minimum standards.* Circulation within a building or building
complex is often seen as a matter of maximizing the proximity of
use functions and minimizing travel distances. Third, there is a
concern for a clear hierarchy of uses and elements of buildings.
Consequently, categorical distinctions are made between promi-
nent and subordinate areas of activity and between the modes in
which they are expressed.

Functionalist production is perhaps epitomized by the so-called

International Style, with its ubiquitous array of steel, concrete,
and glass commercial buildings, each consistent in basic format,
regardless of location and resulting differences in cultural set-
ting. The position can also be seen at work in the spatial organi-
zation of many postwar urban and suburban areas, with their
adherence to concepts of transportation efficiency and an eco-
nomically determined distribution of land uses.

The orientation of the populist position might be characterized

thus: Architecture has drifted too far from popular conceptions,
needs, and aspirations with regard to building. Therefore, an in-
clusive interpretation of the prevailing sociocultural climate, and
especially its commonplace physical and symbolic qualities, is
required as a source for architectural expression. Very much at
issue are understanding the popular consensus in matters of
building and the use of popular preferences as a legitimate
source of guidance for architectural production.

Normative Positions That Guide Design Thinking 125

43
Plan and general view of the Alley Theater in
Houston, by Ulrich Franzen (1965-1968).

Normative Positions That Guide Design Thinking 126

 The architectural devices of the position can be seen to range
widely in the use and rearticulation of commonplace elements of
the contemporary built landscape (see figure 44). For example,
the architecture of the commercial strip and the franchise are
elevated, by some, to the level of serious discussion and devel-
opment (Venturi and Scott Brown 1968, 1971; Venturi, Scott
Brown, and Izenour 1972). They become a basis for appropriat-
ing formal organizing principles. Similarly the eclecticism with
which otherwise functionally rational suburban apartment and
housing complexes are adorned is given serious attention. In
short, the symbolic qualities of popular forms of architectural
expression are acknowledged and interpreted (Rowe 1972).
There is another strain, or orientation, at work within this gen-
eral position, although relatively separate from a direct concern
with formal devices of the above kind. Here user needs and pref-
erences are often regarded as both necessary and largely suffi-
cient determinants of form. In the work of Alexander and his
followers, for example, on reusable ‘‘patterns”’ of building strong
populist arguments are made, alongside the practice of a fairly
high degree of behavioral determinism. More broadly, the rela-
tionship between behavioral and building environments and the
effect the latter has on the former are given undeniable status
(Alexander et al. 1977).

It might be argued that this is the same concern that we saw in

the functionalist position’s expression of use program. If any-
thing, however, the populist position represents a reaction to
that kind of determinism as being biased and preordained by
“elite” institutions, while the actual users have little or nothing
to say. By contrast, the programming of facilities is aimed at re-
vealing those hidden biases and at democratically satisfying the
environmental requirements of a building or building complex.
Consequently, the procedural devices of data collection and
analysis of user needs take on a dominant role.

The work of Venturi and Rauch is probably most directly associ-

ated with the appropriation of popular signs and symbols (Scott
Brown 1977). In a recent residential development in the Houston
area, a skillful layout of housing units was adorned with ‘‘classi-
cal elements” in the form of a direct appliqué to the building
facades (figure 45; Kaliski 1983). The message was clear. These
are the residences of people of substance and status. As a news-
paper advertisement for the project put it, “Park Regency, you've
got great deals and lotsa class.’’ In the end the devices have
much in common with surrounding signs and symbols, while at
the same time clarifying the general proposition of such symbol-

Normative Positions That Guide Design Thinking 127

 44
Signs on the commercial strip (Houston).

Normative Positions That Guide Design Thinking 128

 SE

45
The Park Regency Terrace residences in Houston, by
Venturi, Rauch, and Scott Brown (1981-1983).

ism. After all, this is an area in which apartment and condo-

minium complexes, constructed on relatively low budgets,
strive through decoration and other surface accessories to set
the “right tone.” Ersatz Regency competes with mock Tudor, as
the marketplace takes a distinct turn toward Vanity Fair. The
Venturis’ unabashed use of applied decoration, which borders
on caricature, makes the symbolic statement but at the same
time almost disengages the signs themselves from the buildings.

Closely akin to the populist orientation is that of conventionism,

expressed in the following statements. First, architecture must
engage and be appealing to its audience—both its users and
those who only look at it. (There is a concomitant criticism in-
volved here, that such engagement is somehow lacking in the
contemporary setting.) Second, certain widely held conventions
have developed in architecture over the centuries. These con-
ventions rather literally frame sociocultural expectations regard-
ing the selection and conformation of architectonic elements.
Therefore, the body of historical references, conventions, de-
vices, and types is an essential source for appealing architec-
ture. At issue within the position are matters of selection and
reinterpretation of elements.

Normative Positions That Guide Design Thinking 129

 Typical architectural devices of the position include the use of
building types, as much for their iconography as for their overall
organizational principles. But perhaps most important, a
reliance is placed on historical and historicist references,
conventions, and rule structures for the palette of architectural
expression. For example, classical orders and iconography are
often used as a point of departure for architectural renderings.
In various hands these renderings may be quite literal or quite
abstract. In the work of Thomas Gordon Smith, for instance
(House in Mathew Street, 1978), the references are quite literal,
whereas those in the recent work of Stern are less so (Jencks
1980).
Much of the work presented in Jencks’s recent Post-Modern
Classicism, among other texts, serves to illustrate this posi-
tion (Jencks 1980, Portoghesi 1982). For instance, the Laurentian
House, in Livermore, California, by Thomas Gordon Smith,
clearly draws upon historical references for both its figurative
expression and its organization. The Municipal Control Building
for Quail Valley in Missouri City, by Taft Architects (shown in
figure 46), evokes a strong memory of similar buildings built dur-
ing the 1920s and 1930s (Papademetriou 1980). Finally, Moore’s
celebrated scheme for the Piazza d’ltalia in New Orleans (1978—
1979), commissioned in celebration of the Italian community in
that city, makes direct figural and formal references to Italian
classicism and the tradition of fountain design and plaza mak-
ing (figure 47). More specifically, as Jencks notes in his photo-
graphic juxtaposition of images, the likeness to Salvi’s Trevi
Fountain in Rome (c. 1732) is unmistakable (Jencks 1980, p. 21).

As for the formalist position, its orientation might be summed up

in this way. Architecture is largely an autonomous realm of ex-
pression. It has been and will continue to be so. Therefore, the
program for development in architecture must be located strictly
in the realm of architectural possibilities. Here we find an over-
riding concern with autonomous “‘languages’”’ of architecture,
their compositional qualities, and their inherent and possible
meanings. At issue are formal and figural possibilities made leg-
ible by the use of conventions that are agreed upon and promul-
gated by a community of peers.°

This description would seem to fit the conventionist position as

well. After all, the use of recognizable architectural conventions
is a form of expression that is essentially autonomous to ar-
chitecture. The important distinction here is the insistence on
exploring new formal and figural possibilities, and on reinter-

Normative Positions That Guide Design Thinking 130

46
The Municipal Control Building of the Quail Valley
utility district in Missouri City, by Taft Architects
(1978-1980).

preting traditional architectural conventions, which goes well

beyond borrowing. For this reason production is apt to incorpo-
rate much more from the private languages of architectural
expression.

The architectural devices of formalism, as one might imagine,

are apt to be quite diverse. They include exploitation and trans-
formation of the formal qualities of regular solids, as seen in
Eisenman’s ‘“‘Houses,”’ with all their linguistic, syntactic meta-
references (see figure 48). By contrast, in the later work of
Graves one sees figural explorations and inventions, based upon
“the myths and rituals of a society” (Graves 1984). Formalist de-
vices might also include Hollein’s more geomorphic construc-
tions, about which he observes, “‘The shape of a building does
not evolve out of the material conditions of a purpose....A
building is itself... . Architecture is purposeless [and] what we
build will find its utilization’ (Conrads 1964, p. 181). Finally, the
rationalist doctrines of Rossi, Krier, Huet, and others attach
great importance to the reinterpretation of artifacts within the
urban landscape such as monuments, streets, squares, and
quarters (Rossi 1982, Krier 1978, Huet 1978).

Other positions might also have been discussed. For instance,

use of the ‘‘vernacular’’ and the ‘‘situational context” are both
common sources of organizing principles in contemporary ar-

Normative Positions That Guide Design Thinking 131

47
Piazza d'Italia in New Orleans, by Charles Moore,
Perez Associates, and Ron Filson (1978-1979).

Normative Positions That Guide Design Thinking 132

48
House 11a in Palo Alto, by Peter Eisenman (1979).

Normative Positions That Guide Design Thinking 133

 chitectural production.® In our scheme both might be regarded
as combining elements of the populist and conventionist doc-
trines. Vernacular and contextual urban architectural expres-
sions are certainly populist with respect to their incorporation
and celebration of commonplace elements, and they can also be
said to involve rule structures of a conventionist kind that lend
them coherence and meaning. They do have an independent as-
pect, however, particularly as expressions of the phenomenon of
place, or genius loci (Norberg-Schulz 1980a). In this connection,
Frampton views contemporary works of Siza, Utzon, Barragan,
and others as deriving their importance and special meaning
from an acute sense of place, sociocultural milieu, and local
building practices; his concept of regionalism defines architec-
ture largely from this standpoint (Frampton 1982, 1983).

[gta peewee nee ay ten ice AN ene aloes A Redo ee

Problems of Substantiation

With this general idea of theory in mind, we must now pursue

more fundamental questions concerning the substantiation and
legitimation of normative positions in architecture. This analysis
will begin with questions of logical coherence and will proceed
to a discussion of two alternative frames of reference from which
interpretations and evaluations of normative positions might be
rendered. The first of these frames of reference assumes that
claims to legitimacy are predicated on fundamental conceptions
of the human condition and must be judged in terms of the valid-
ity of those conceptions; the second considers architectural po-
sitions from the point of view of the relative appropriateness of
the means employed, under the assumption that the ends are
more or less commonly agreed upon. This discussion will be
followed by an attempt to address the problem of assigning
priorities among architectural principles, once those principles
have been judged to have merit.

@ Logical Coherence
Let us begin examining the issue of logical coherence from the
vantage point of an abstract model of the basic structure of the
theoretical positions, as identified during the earlier discussion
of surface features. This model may be described as a critique of
existing practices (C), implying the location of a problem and
alternative opportunities (P), implying counterpropositions and
their rationale (CPR). Symbolically the relationship may be ex-
pressed as C — P and P — CPR. The model should be regarded
less as a three-step process for theory construction than as a

Normative Positions That Guide Design Thinking 134

 constellation of statements with logical connections. It is also an
abstract representation of how architectural positions might be
seen to change and shift over time, building upon prior positions
as points of departure. Further, it might be seen in a more tem-
porary light, as a model for the selection of an organizing princi-
ple during a particular design project. Thus it may also be
written more simply in the form of a single relationship: location
of a problem (P), implying a proposition (Pr), or, P— Pr.

Clearly in this reformulation an assumption has been made that

the problem definition P and the proposition Pr are clear de-
clarative statements of the kind we saw earlier in Le Corbusier's
and Meyer's proposals, or that they can be faithfully contracted
to form such statements. The earlier subdivisions of critique,
identification of alternative opportunities, and rationale for the
propositions can be understood as auxiliary statements involv-
ing modes of argumentation linking the problem statement P
with the proposition Pr. For instance, the critique may be made
in the light of the potential offered by alternative opportunities
(for example, technology) and may develop a rationale that is
contingent upon these opportunities (such as, that new tech-
nologies should be embraced because of their efficiency). This
reformulation also squares better with the idea of an external
model mentioned earlier in connection with categorical systems
and the work of Norberg-Schulz. In other words, the relation be-
tween P and Pr is defined and orchestrated by the external
model M containing information about what constitutes proper
architecture and the underlying rationale for such an assertion.

In more extensive form, the idealized structure of a position can

be expressed in the following manner. If problem P is located,
then proposition Pr follows. Furthermore, if P is true, then Pr
also holds, subject to auxiliary arguments. That is, in the pres-
ence of the external model M (for instance, a matter of ‘‘tech-
nics’), the problem P is located (for instance, ignorance of
modern technology), and if M is true, then P also holds. In addi-
tion, in the presence of the same external model M the proposi-
tion or counterproposal Pr (say, the use of modern technology)
also holds, and again, if M is true, then so is Pr. In summary, P >
Pr subject to M— P and M— Pr. Thus, the claim for legitimacy,
or the process of substantiation, rests on the extent to which the
external model or its implied surrogate can be sustained and
proved for both the problem statement P and the proposition Pr.
Moreover, if we allow that M is in fact robust enough to allow
explication of P and Pr in complementary fashion so that they
are ‘‘connected,”’ at least in rough content (that is, provisionally

Normative Positions That Guide Design Thinking 135

 P — Pr), then the strategies or types of argumentation at work in
making the connection become important. Otherwise Pr would
be entirely independent, an issue we shall take up a little later
on.
For the sake of further clarity in exposition, let us also assume
that Prp constitutes a set of principles under contention in the
perception of the extant problem P (that is, P— Prp). As we saw
earlier, both Le Corbusier and Meyer were essentially arguing
against a set of existing principles and practices when they pro-
mulgated their theoretical positions.

General Strategies
In principle, at least three kinds of strategies can be employed in
making the connection between a perceived problem P, or prob-
lematic set of principles Prp, and a proposition or counter-
proposal Pr. First, there is the type of approach that involves
modification of troublesome yet reparable features of an existing
position without entirely undermining that position—that is, Prp
— Pr, where Prp M Pr and M => Prp. In other words, in making
the counterproposal the troublesome subset of principles Prp is
excluded in the presence of the external model M. In practice,
this strategy would correspond to the general adoption of a par-
ticular orthodoxy or architectural position, with the absence of
certain of its features. Here the exclusion of features may arise
not only through disagreement but because they are considered
irrelevant to a particular realm of application or speculation.
A second strategy is that of extending existing positions to take
into account other contingencies or possibilities that have so far
been ignored—that is, Prp > Pr, where Prp U Pr and M = Prp.
In making a new proposition, a set of existing principles is ac-
cepted as appropriate with either an elaboration or the addition
of consistent principles furnished through the external model.
This is a common practice in the normal development of archi-
tectural positions and often coincides with their rise to promi-
nence and acceptance from relatively tentative beginnings. In
effect, the general thrust of a position, articulated by certain key
principles, becomes elaborated and extended in a coherent and
complementary manner. Episodes in the development of ‘‘mod-
ern movements” are cases in point.

Third, there is the strategy of replacing an existing position en-

tirely. This should not be construed as a willful act for the sake
of being contrary. Rather, in the course of the development of
architectural positions the systematic questioning of certain

Normative Positions That Guide Design Thinking 136

 basic premises can result in what turn out to be new positions.
Here several variations can be distinguished.

To begin with, there is what we might call an oppositional or

even an iconoclastic approach. If a position currently at work is
seen to be problematic in some way, then something close to a
diametrically opposed counterproposal is made: Prp —> Pr,
where Pr = Prp. The positive effects of this strategy can be seen
in the use of paradox as a point of departure. In certain regards,
the doctrine Venturi expresses in Complexity and Contradiction
in Architecture can be seen to represent such a shift within a
modernist position (Venturi 1966). Here, we are confronted with
a well-argued case for replacing a generally accepted ational
and systematic ordering of architectural elements and design
principles with another view that stands many of the accepted
principles of modernism on their heads. In addition to the use of
historical references, the work calls for celebration of certain
ambiguities in architectural expression in contradistinction to
their unambiguous resolution. To quote from Venturi’s introduc-
tion, “Everything is said in the context of architecture and
consequently certain targets are attacked—in general, the
limitations of orthodox modern architecture and city planning, in
particular, the platitudinous architects who invoke integrity,
technology, or electronic programming as ends in architecture
... and suppress those complexities and contradictions inherent
in art and experience” (Venturi 1966, p. 21).

There is also what might be called a comparative approach,

whereby new propositions are based on an enumeration of and
comparison among alternative principles of the same general
kind as the extant set Prp. Symbolically

Prp — Pr, where

M— (Perio een, rand
\(Pr;, Pro, ..., Prp)| > Pr, as well as
(Pry, Pro, ..., Pp) > Prp.
Here the concept of ‘principles of the same general kind” may
be subject to several constructions. It may suggest that
there is a difference with regard to means, while the principles
involved follow the general direction of the extant set Prp. For
example, although the broad goals of an architectural position
may be maintained, the strategies for achieving those goals are
replaced by others deemed superior. Or it may mean that the
issues of central concern in an architectural position are
accepted, but entirely different propositions about those issues
are offered. In the ‘‘form follows function” debate, for example,

Normative Positions That Guide Design Thinking 137

 the aphorism, and all that it intends, might be reversed so that
“function follows form,” thus shifting the emphasis of architec-
tural production, although still around the same central issue.

It could be argued that this last example is more closely aligned

with the oppositional approach described earlier, or, conversely,
that Venturi’s argument in Complexity and Contradiction is a
case of the comparative approach. Such distinctions are largely
a matter of degree. It should also be noted that it is possible to
combine the modification and extension of an existing position;
indeed, such a combination frequently occurs. If anything, Ven-
turi’s position most closely fits the latter mold in that it broadens
the realm of modern architectural concerns. Clearly, in practice
less-than-perfect examples of these ‘‘in principle” strategies are
likely to be found.
Finally, there are what we might call independent strategies, in
which a set of principles constituting an architectural position
emerges largely independently of any perception of an initial set
Prp. \n fact, the position might arise without the explicit exis-
tence of some set of norms that poses difficulties. This is clearly
a case of manufacturing a position out of whole cloth—of pro-
posing a direction that is both complementary to and indepen-
dent of existing positions. The work of Cedric Price in the days
of the ‘‘Potteries Thinkbelt’”’ (Price 1965, 1966a, 1966b) and other
similar excursions into the realm of infrastructural reuse and
interpretation by the likes of Archigram represent this kind of
approach (figure 49).

Pitfalls
In all cases save the last, the strategies have a distinct referential
character. They in some way involve a prior set of propositions
and positions, whether in the form of acceptance, indifference,
or rejection. This can pose a severe problem for sustaining
claims of legitimacy, on at least two counts. First, the interde-
pendence between the proposed position and prior positions
may be so overdrawn as to amount to a circularity of argument.
It is a circularity that may increase understanding but that makes
the proposal neither more nor less legitimate than its predeces-
sor. For example, in the oppositional strategy outlined earlier,
the prior proposal Prp and the counterproposal Prp, logically
speaking, have the same status. They rest on the same founda-
tion, even if one is the ‘‘not” side of the other. Thus, ‘“‘less is
more’ and “‘less is a bore” have equal force, more or less, de-
pending on who's using them (Norberg-Schulz 1968). Further

Normative Positions That Guide Design Thinking 138

 Slee
-~ Bee

5 oa aaeamreson ewe ge oN ie a

49
Walking City, by Ron J. Herron (1964).

substantiation of one at the expense of the other is lacking at the

level implied by the rhetoric.

Other, less clear-cut cases may suffer from the same kind of im-
pediment, albeit at lower intensities. Definition of a problem
solely by way of an alternative position may raise questions
regarding the substantiation of the problem statement in the
first place. Hannes Meyer’s thesis on building can be seen to
come perilously close to this position, through the strictly oppo-
sitional use of “function” and ‘‘aesthetics” in relationship to
‘life’ and the ‘“‘process of building.” It is tantamount to an
argument by assertion.

The second kind of difficulty arises when the position being pro-
posed uses prior propositions as building blocks and tacitly or
uncritically assumes their legitimacy. Clearly, the strategies of
extension and even replacement have this characteristic.
Whether this is done unwittingly or not is beside the point. For
example, if one were to develop a theoretical position in design
tacitly using the notion of ‘type,’ one might run the risk of sub-
verting the proposal by claiming a relevance that cannot, in fact,
be supported. Similarly, embracing ‘‘traditional’ and historical
principles per se, without regard to their sociocultural import in
a contemporary setting, can also be disastrous.

We have seen that normative positions consist of finite sets of

categories that are, or can be, used to distinguish one position
from another. They are at once the core of the definition of what
proper architecture is, versus what it is not, and the principles

Normative Positions That Guide Design Thinking 139

 that provide guidance for practical inquiry. It would seem, there-
fore, that adjudication of the claims of competing positions, un-
less it can be clearly decided on the basis of logical coherence
within the realm of accepted architectural categories, should
proceed in another direction.

@ Architectural Positions as Doctrines about Ends

But what is this direction? How might we further substantiate
claims for legitimacy for the architectural categories presented
by competing or coexisting positions? Here, it is apparent that
we need to move outside the realm in which architecture can
claim exclusive knowledge or expertise. Substantiation for cate-
gories, and hence positions, must come through linking the sub-
ject of speculation, namely the concept of architecture, toa
broader frame of reference. Furthermore, the frame of reference
must not only represent a touchstone for architectural experi-
ence but also have certain inert or stable qualities; otherwise,
the dilemma of making distinctions among positions will remain
unresolvable. Such a link is clearly assumed in statements like
the following, by Vincent Scully: “Architecture, like all art, ...
has revealed some of the basic truths of the human condition
and, again like all art, has played a part in changing and reform-
ing that condition itself’ (Scully 1979, p. 10).

A useful way of illustrating and exploring this point is by inquir-

ing what is held to be the basis for the claim to legitimacy of
each of the architectural positions outlined earlier in the chap-
ter. Briefly, the functionalist position would seem to be founded
on the belief that man’s choice behavior can be analytically typi-
fied and idealized and that an unadorned, efficient reckoning
with building can provide an appropriate form of architectural
expression. There is no need for self-conscious references to be
made to the past; the present circumstances are what count. A
‘science of man” is possible and, in combination with technical
achievements from natural science, will provide a firm basis for
architecture. By contrast, the apparent basis of the claim for
legitimacy of the populist position is the existence of commonly
held preferences and needs vis a vis the built environment. Al-
though these preferences and needs are usually latent or uncon-
scious, they can nonetheless be approached, understood, and
appropriated through outward building practices. Substantiation
also seems to hinge on a belief that a consensus struck in the
‘popular marketplace”’ is reasonable and good, at least for the
purpose of guiding architectural production.

Normative Positions That Guide Design Thinking 140

 For the conventionist position the apparent basis of the claim to
legitimacy is that commonly held expectations exist vis a vis the
formal qualities of architecture, as can be demonstrated by the
endurance of past forms of composition, organization, and ex-
pression. Such a claim hinges on the conviction that variations
in the expectations within a sociocultural milieu are continuous
and small. In other words, conditions of the past can be sus-
tained into the future without undue or disabling distortions. In
contrast, claims of legitimacy by the formalist or elitist position
seem to be predicated on the possibility that intersubjective con-
ventions can be broadly developed, based on common human
experience, or more particularly, common human rationality.
Non-alienation from an “autonomous architecture” would,
therefore, seem to hinge on both broad societal acceptance and
inculcation of these conventions.

From this admittedly sketchy account several interesting obser-

vations emerge. In all cases it seems possible to erect an appeal
for substantiation that is based on a conviction involving a
widely held view of man and his world. At least there is the com-
mon belief in the possibility of human engagement with architec-
ture beyond matters of shelter and mundane function. At the
heart of this conviction would seem to be a faith in a common
rationality and a common thread of experience. If this were not
the case, then the idea of a consensus of peers and its broader
influence, implicit in the elitist position, could not have been es-
tablished. The habitual proclivities of laymen would not result in
the necessary foundations for populist programs. And the idea
of aesthetic conventions as social rule structures, needed to ad-
vance the conventionist position, could not be sustained. More-
over, it is a belief about man’s being and aspirations that goes
beyond surface conduct.

Within this general conviction, however, two ideas would seem

to be competing for attention. To use Bernstein’s terms, they are
a ‘foundational’ view of man and his world and a ‘‘relativist”’
view (Bernstein 1983, p. 8f.). According to the foundational view,
man’s experience can be reduced to certain ‘‘common de-
nominators”’ through some ahistorical and aspatial matrix or
framework. Hence, the postulated common human rationality
spoken of earlier can be located and, once located, will remain
invariant. By contrast, the relativist view claims that no matter to
what extreme one moves within a foundationalist program, the
observations one can make about rationality and experience can
never escape the influence of particular traditions, cultural
frameworks, or ‘‘forms of life.”’

Normative Positions That Guide Design Thinking 141

 It might be argued in the case of an elitist position that man’s
basic perceptual equipment, with sufficient guidance, should al-
low engagement with architecture of a foundational kind. It can
also be argued, however, that such engagement must make
use of a certain knowledge about the architectural conventions
that are inevitably associated with a particular tradition or form
of life.
Similarly, a populist position would seem to be quite relativistic
in embracing the prevailing preferences of a specific time and
place, yet the position must also assume that these popular
manifestations are foundational, at least for that specific time
and place. The problem with such a view is that popular mani-
festations need not be foundational, particularly in cases where
the range of available models is narrowed by notions that people
have been taught to have about what is appropriate to their eco-
nomic or social circumstances. This point is central to the posi-
tion of Alexander and his followers, who seek to subvert the
“tyranny of hidden programs” and replace what people think
might be expected of them with what they themselves hold ‘‘near
and dear’ (Silverstein and Jacobson 1978). Yet what is held near
and dear by a group may reveal a narrow preference for certain
environmental elements through an ignorance of other pos-
sibilities. Under these circumstances a popular consensus view
may prevail where better and even more acceptable solutions
could have been found.

The functionalist position is clearly foundational in almost all re-

spects. As pointed out, this inclination can produce an unrelent-
ing sameness in the built environment that disregards special
local circumstances. Hence, the presumed universality of the po-
sition tends to become more and more narrowly defined.

The foundational aspect is apparent in the conventionist posi-

tion’s (implicit) belief that certain forms of architectural expres-
sion found in the historical record are worthy of emulation
because they represent touchstones of human experience with
architecture. Nevertheless, it is also clear that the history of ar-
chitecture is inextricably linked to history in general—that a par-
ticular form of expression is linked to the culture and traditions
of the moment in history when it was devised. It is possible that
that form of architectural expression will take on a meaningful
aspect in the absence of other conditions of shared thought and
experience, but it is hardly self-evident.

Finally, an inherent weakness of the type of theory under discus-

sion is that even though foundations in a larger frame of refer-

Normative Positions That Guide Design Thinking 142

 ence might be provided, discourse often fails to turn sufficiently
in that direction in order to substantiate merit. Without this turn
it is all too evident that the locus of these normative positions in
architecture remains very much at the level of production, ar-
chitectural devices, and general outlines of orientation. What
then appears lacking is a more profound basis from which the
question of the efficacy of the proposed categories might be ap-
proached. Whether this is a serious weakness for design think-
ing is an issue to which we shall return.

@ Architectural Positions as Doctrines about Means

There is another way of looking at architectural positions and
their substantiation that makes many of the same points, but
with some important differences. Instead of regarding architec-
tural positions as arguments with proposals linked solely to
some deep-seated view of man and his world, we can regard
them as essentially teleological doctrines, or doctrines about
means.

Teleological doctrines, after Frankena’s definition, are those in

which the concept of “‘good”’ is defined independently from the
concept of “right” (Frankena 1963, p. 13). To paraphrase Rawls,
the right is then defined as that which maximizes the good. For
example, institutions and acts are right that of the available alter-
natives produce the most good (Rawls 1971, p. 24f.). Or, in our
case, architectural positions are right that maximize their con-
cept of good. Those that are right are also presumably prefera-
ble. The necessity for the independence between good and right
is fairly clear. It avoids the problem of defining good in terms of
what’s right and vice versa.

One might well ask what is the pertinence of seeing architectural

positions in this way. First, in the four cases just described an
inherent concept of good can be seen to be roughly congruent,
or shared. It is a concept of good that says, among other things,
that all architecture must engage its audience. It must foster
understanding and be intelligible. This is clearly not the only
concept necessarily at work, but it is one that would have broad
adherence, even beyond the four positions mentioned.

Second, on the issue of right seen independently of good, four

rather different interpretations, or means toward maximization of
the good, would seem to be at work. Certainly, as we have seen,
the general orientations and architectural devices of the four po-
sitions are distinctive and different. The question then becomes,
Which position might be right, seen in terms of maximizing, for
instance, the concept of engagement? Or, conversely, under

Normative Positions That Guide Design Thinking 143

 what circumstances can an architectural production, in these
same terms of engagement, be called arbitrary?

In Hubbard’s defense of an ‘‘architecture of convention,” which

is close to our conventionist position, he asserts primacy for
“immediate appeal” on the engagement issue (Hubbard 1981).
By this he means the mustering together of those sensations and
concepts that facilitate immediate understanding, comprehen-
sion, liking, and gratification. A proponent of a populist position
would presumably go along with the general thrust of this crite-
rion but would assert primacy for production coming out of the
realm of commonplace artifacts. To the extent that the same
conventions and rules are at work in the locale of interest, the
populist and conventionist positions would seem to be con-
gruent. Both positions would also seem to be at odds with the
elitist position, and proponents of each position tend, in fact, to
be overtly critical of the elitist position for its alleged formal ar-
bitrariness.’ At this point we might be inclined to conclude that
the conventionist and populist positions are right and the elitist
not, in spite of al! three’s shared conviction about what is good.
In asimilar vein, production from the functionalist position
would seem to straddle the fence, appearing to some as what it
is and what it is made from, while appearing to others as aloof
and abstract.

We might, however, supplant the largely consensus-based idea

of “‘appeal’”’ as the criterion for non-arbitrariness with one that is
a little more even handed—in other words, not use an obvious
strength of one position at the expense of others. Here we might
go on to say that a work of architecture is arbitrary when it trans-
gresses the limit of its (architectural) possibilities without also
providing a framework for extending the level of reflection, so
that the transgression might be understood.

Under this criterion the elitist and functionalist positions might

have just as legitimate a claim to being right as the others, at
least to the extent that the framework for extending the level of
reflection is manifest. For example, a work based upon the
geometric possibilities of regular solids may be very abstract but
certainly approachable through the rules of geometry. Adher-
ents of an elitist position might also argue, and do, that that ex-
tension of inherent formal possibilities in architecture is a good
in itself and therefore a social responsibility. It should also be
pointed out that both the populist and the conventionist position
are fundamentally engaged in interpretation, and hence in exten-
sion of this kind.

Normative Positions That Guide Design Thinking 144

 The overall effect of this line of discussion is that it raises not
only the issue of a good itself as requiring substantiation but
also the question of right and proper architecture, given some
reasonably common agreement about good. Without attempting
to resolve this issue among the positions just presented, the dis-
cussion does point up the need for theoretical frameworks that
go beyond a parochial form of discourse about certain catego-
ries, in search of better bases for substantiation.

Besides the admittedly somewhat vague matter of engagement,

we might speak more directly of other qualities of architecture,
such as its ‘‘thingness”’ (Bachelard 1979, pp. 201—206) and tec-
tonic aspect, or its ‘‘placedness’’ as a part of a locale or col-
lective enterprise. As to the former quality, the functionalist
position might rise to the fore with its program and orientation
toward material integrity. In the latter case, the populist position,
at least for the present, could lay a strong claim to our attention;
although if continuity is a valued dimension of the quality of
placedness, then the conventionist and elitist positions could
both press their claims. First and foremost, however, there must
be engagement for us to have knowledge of these qualities. As a
corollary, it is through these kinds of qualities that we are en-
gaged and can therefore participate.

@ Assignment of Priority to Categorical Claims

At least in practice, one of the pressing problems involved with
architectural categories, once they have been defined, is the as-
signment of priorities among the aspects that are to be consid-
ered good and proper. For example, suppose one had to choose
between “‘“‘imageability,’’ functional conformance, and structural
innovation; how might one assign the priorities among these
three categories in designing a building? An essential feature of
theory that purports to guide practical action is the provision of
some basis for assigning such categorical priorities.

Following Rawls’s analysis, we can distinguish five methods or

procedures for assigning priorities (Rawls 1971, pp. 40—45).
First, there is the use of intuition, or more precisely, the balanc-
ing of prima facie principles (implying categories) on the basis of
individual experience, insight, and discretion. This is certainly
not necessarily a systematic approach, but it is also by no means
necessarily irrational. For instance, experience may have shown
us that by first attending to functional conformance we can
then produce a technically sound building with a pleasing
appearance.

Normative Positions That Guide Design Thinking 145

 Second, there is the use of some single-principle doctrine
whereby all candidate considerations are conflated to one over-
riding set of categories or principles. This procedure is explicit
in the so-called doctrinaire positions described earlier, at least to
the extent that they require such considerations to be made. It is
certainly systematic, although it can become troublesome when
the otherwise highly valued principles or goods involved are too
disparate to warrant such aggregation. This may occur, for ex-
ample, under circumstances when a striving toward structural
innovation severely compromises performance with regard to
functional organization.

The third procedure, in principle anyway, is the ordering of a

true multiplicity of principles, or goods, through a categorical
system that explicitly identifies architectural categories and their
most appropriate interrelationships. Embedded in Norberg-
Schulz’s theory, discussed earlier, is an example of this kind of
approach: ‘The totality of a building task is realized technically
within a style’ (Norberg-Schulz 1965, p. 87). In general, however,
explicit guidance as to how to approach the ordering problem is
not fully provided by systems of this type. The preference for
certain forms of expression over others invariably clouds the is-
sue, reducing the ordering to something closer to an intuitive or
single-principle process.

The use of what Rawls calls ‘‘prudential judgment” is a fourth

kind of procedure that is evident in theoretical development of
normative architectural positions (Rawls 1971, p. 44). This is
where decisions about the relative priorities and underlying or-
dering of principles are made from the standpoint of specific
“settings,”’ or contextual viewpoints. For example, in architec-
ture there is a reasonably extensive tradition of adjudication
through the use of concepts such as genius loci, social domain,
and sense of place. This can be contrasted with various brands
of “universalism” that deal with organizing principles and cate-
gories independent of time and place, such as the so-called uni-
versal principles about building technology that are embraced
by the functionalist position. One of the advantages of this type
of procedure is that it can, or should, provide systematic specu-
lation regarding the circumstances under which the particular
concept of setting becomes uncoupled or disengaged from the
principles, categories, and goods at hand. For instance, the cul-
tural setting for a design may be such that imported references
from another tradition are anachronistic, in spite of a designer’s
opinion on the matter. In other words, the presence of a bound-

Normative Positions That Guide Design Thinking 146

 ing Critical perspective is required. In practice, however, such a
perspective may not actually be provided.

The final procedure involves a process Rawls terms ‘‘lexical

ordering” (Rawls 1971, p. 44). This approach entails identifying,
from among all candidates, the principle or concept of good that
would have to be satisfied in the first instance, followed by the
next principle or good for the second instance, and so on.
Rather than being merely an intuitive rank ordering of principles,
it is systematic, at least to the extent that superordinate and sub-
ordinate connections and relations among principles are made
explicit. It certainly moves beyond both the single-principle and
categorically systematic methods, in that more than one kind of
concept of good is involved, and a basis for substantiating an
ordering is explicitly provided beyond an individual designer’s
Opinion.

An example of this approach, albeit outside the realm of archi-

tecture, would be the ordering of the two principles ‘‘equality for
all” and “equal distribution of wealth.” Here it might be argued
that equality is the greater good, for it sets up certain indis-
pensable preconditions or rules about questions of wealth
distribution, and not vice versa (Rawls 1971, p. 41). In the realm
of architecture, it was argued earlier that meeting the criterion
“architecture must engage its audience” was an indispensable
precondition for achieving other goals. Without such engage-
ment there can be no further understanding of a work.

in summary, whether we take an ontologically inclined or a tele-

ologically inclined view of architectural categories and their sub-
stantiation, the underlying principles employed in the ordering
process, once valuations of merit have been made, should be of
interest. There are, in fact, a number of such procedures already
at work in architectural discourse and others that are less widely
used but just as plausible.
@ Theory and the Implied ‘‘Urcorpus”’
Before concluding this discussion of normative theory in archi-
tecture, it will be useful to return to two issues raised in earlier
portions of the account. They are (1) the centrality of a position
to the realm of contemporary architecture and (2) the compre-
hensiveness of a position in the same realm. It is safe to say
that the most compelling positions, at a given moment, are the
ones that seem to address the important issues of that moment.
For example, in early modernism, the opportunities presented
by a new technology and, at least in Europe, a new set of social
and political conditions were important and central issues of the

Normative Positions That Guide Design Thinking 147

 time. We would also probably agree that architectural theory
should atternpt to comprehensively address and define the
realm of architecture. This is a question of both determining
what architecture is, versus what it is not, and filling in between
the bounds thus demarcated.

In examining many normative positions, however, and following

the contours of theoretical discourse more generally, one finds
much that remains unsaid. Certain dimensions in the realm of
architecture are implicitly assumed and embraced by otherwise
distinct architectural positions. That such incompleteness
should be the case is understandable. The focus of good theory
is usually on those aspects of interest or those that command
speculation at a point in time. More mundane aspects are often
only cursorily dealt with, if they are dealt with at all. During these
moments the theorist appears to be drawing on what one might
call, after Levi (1980), an “‘urcorpus”’ of Knowledge that is con-
sidered to be “incorrigible” and infallible.? There are certain as-
pects of architecture, at least at certain moments, about which
there appears to be unanimity. Theory then, it would seem, is
mainly concerned with that which lies outside the urcorpus of
knowledge.

A significant problem can develop over time, however, when

theoretical discourse in the controversial realm outside the ur-
corpus begins to demonstrate considerable variation and differ-
ences in direction—when the locus of debate shifts from one set
of dimensions or issues to others. For instance, it might be ar-
gued that the attack on modernism, on the grounds of its con-
temporary failure in the realm of popular visual appeal, ignores
the sociopolitical conditions of dwelling that were being ad-
dressed at the time of modernism’s evolution. Here the problem
is that such shifts may call into question the very contents of the
urcorpus. For when dimensions of architecture in one position
or at one time are ignored in other positions or at other times,
the qualities of the stand taken by the positions that choose to
ignore those dimensions come into question. The upshot is that
the urcorpus appears to be subject to correction and change.
Thus, failure to define the boundaries of the urcorpus, when en-
gaging in debate outside its precincts, raises questions of am-
biguity in normative positions and weakens their claims to
substantiation. When a challenge is made to a position, retorts
like “Of course we consider those things, but what architecture
is really about is ” become suspect.

Normative Positions That Guide Design Thinking 148

Theory and Practice

If we accept that many otherwise competing, complementary, or

simply different architectural positions embody similar notions
of good, then their competing claims to legitimacy and centrality
might be seen as essentially of the moment. Proponents are
dealing each in his own way with what they see as the valid
hierarchy or priority of needs and aspirations during their time.
They are also each attempting to maximize those goods in the
manner that is most appropriate to this supposed hierarchy and
to the resources available at the time. Substantiation is thus a
question of goods, their priority, and the right means for bring-
ing them about.

On the other hand, there is at least a tacit assumption of inter-

temporal transcendence in most positions. They attempt to
stand for concepts of proper architecture regardless of the mo-
ment. Therefore, adequate substantiation of claims to legitimacy
and centrality must address a twofold issue. First, a position
must show that the guiding idea of good and the ensuing pre-
scription of means is fundamental at the moment and that sub-
stantiation can be sustained between moments. In the light of
earlier discussion, this would mean that the particular version of
an appeal to common rationality and experience will hold in
other momentary circumstances. Second, the theoretical posi-
tion must exhibit a character of completeness by embracing
other fundamental dimensions of architecture that it does not
directly address.

The posing of these two criteria should not be construed as a

call for the convergence of all positions on some “middle
ground.” For that to be practical, or even desirable, we would
have to agree upon an “‘essence of things” that remains fixed, is
incapable of variation, denies other experience, and is ahistoric.
This hardly seems reasonable. Nevertheless, it is a call for as-
suming the responsibility of going beyond the realm of purely
architectural categories and self-referential lines of argument in
making claims to legitimacy.

Finally, one might ask what all this has to do with the guidance
of practical inquiry, design thinking, or professional action. As
we have seen, normative positions of the types examined, as well
as those that are more idiosyncratic and personal, do in fact pro-
vide direction for action. There is an undeniable connection be-
tween design thinking, the artifacts from its practice, and
thecretical pronouncements from the same or a nearby source.

Normative Positions That Guide Design Thinking 149

 Throughout, an appropriate balance must be struck between
aloof abstractions providing for nearly infinite interpretations
and concrete prescriptive principles. Without such a relationship
theoretical positions would be inoperative in the practical realm.
The point is, however, that when there are failures to provide
sufficient substantiation, theoretical positions are vulnerable and
practical inquiry based upon them is impaired. A plethora of po-
sitions arises, all of them making more or less equally valid
claims, and decisions about which of them to look to for guid-
ance become merely a matter of personal taste.

Normative Positions That Guide Design Thinking 150

=

z \

——
——
i =
ra a rs

— —

SS

—_ =
_
 ~ Chapter
Architectural Positions and Their
Realms of Inquiry
 Within the various normative positions that seek to describe
what constitutes proper architecture, two realms of inquiry can
be discerned. Each derives its distinctive characteristics from
both the locus of inquiry and the interpretative frameworks em-
ployed in seeking substantiation and grounding in meaning for
architectural production. These interpretative frameworks, it will
be argued, can have the effect of circumscribing interpretation
within their methodological confines, sometimes with undesir-
able consequences. In the sense that they prescribe the exclu-
sion or inclusion of certain data and kinds of observation, they
are ideological; yet the oppositional relation in which they are
often perceived to stand disguises common purposes and inter-
pretative problems. Furthermore, architectural positions, al-
though they may strongly favor one realm of inquiry, rarely show
sole conformance with either. Rather, each realm provides a
vantage point from which some measure of legitimation may be
secured for a position’s proposals and subsequent production.

Eee ea Ee ea
Two Realms of Inquiry

In the first realm of inquiry architecture is seen in relationship to

a hypothesized society or interpretation of man and his world.'
Here the fundamental concern is with legitimating and validating
architecture with reference to ‘‘nature’”’ or to a set of events or
presumed facts that lie outside architecture itself. Inquiry fre-
quents this outside domain in an effort to erect or borrow the
necessary interpretative scaffolding to explain, compare, and de-
vise architectural design principles within the inside world of ar-
chitectural objects and production.

In the second realm architecture is seen in relationship to itself

and its constituent elements. As Vidler puts it in his exposition
of neorationalism, ‘“Columns, houses, and urban spaces, while
linked in an unbreakable chain of continuity, refer only to their
own nature as architectural elements, and their geometries are
neither naturalistic nor technical but essentially architectural”
(Vidler 1978, p. 31). The locus of inquiry is the architectural ob-
ject—in Vidler’s and the neorationalists’ case, the form of the
city. Such interpretative frameworks see the object world as al-
ready a domain of man’s manifest being and enterprise, and thus
as not requiring outside excursions for the purposes of valida-
tion and legitimation. Eisenman, for example, states that “much
of the ‘modernist’ enterprise had to do with work on the lan-
guage itself,’’ by which he means the language of architectural
 form, ‘‘one that was concerned with its own objecthood” (Eisen-
man 1978, p. 22).

Within these two realms of inquiry, “grounding in meaning” it-

self takes on a number of constructions. It can mean the making
of architecture that is understandable, intelligible, and useful,
and thus congruent with man’s needs and wants. Or it can mean
an architecture that is rhetorical, argumentative, polemical, and
thus capable of providing commentary on man and his world. It
is generally the case that the first realm of inquiry, where archi-
tecture is seen in relationship to a nature outside itself, tends
toward the former sense of ‘‘grounding in meaning,” while the
second view tends toward the latter sense. For instance, the
grounding of architecture in a view of social behavior, institu-
tional procedures, and the like allows discrete needs and wants
to be addressed and a relatively direct connection to be made
back to an architectural object. On the other hand, interplay
within a palette of rhetorical architectural resources, especially
when they exhibit historical continuity, allows for critical strate-
gies of quotation, redefinition, and polemic, the idea being that
architectural elements and fragments from the past, with all their
associated cultural meanings, can be recombined to make new
statements.

fee Sg Saoeensle a pate Ras RS a Gee Sieeen ey ony

Architecture from a Naturalistic
Interpretation of Man and His World

In seeking substantiation from outside the domain of architec-

ture, recent and contemporary architectural and urban design
positions have tended to look to three sources: human behav-
ioral science, production technologies, and—later on—various
fields of environmental management involving both technologi-
cal means and controlling influences on choice behavior. The
kind of interpretation that is involved has been described by
Bernstein as ‘naturalistic’ in the sense that it considers reality
to be adequately described by concepts of “‘natural phenomena”
(Bernstein 1976). It is a ‘‘scientific” interpretation of man and his
world, after the natural sciences. It represents an epistemology
and world view that are very definitely foundational in inclina-
tion, and it generally coincides with a positivist orthodoxy.” Con-
sequently, this interpretative framework tends to recognize two
legitimate models of knowledge—namely, natural science and
formal disciplines such as mathematics and logic.

Architectural Positions and Their Realms of Inquiry 154

 @ Functionalism and the Modern Movement
Amid the postrevolutionary fervor of the early twentieth century,
architects sought to define their role through an avant-garde
stance. They espoused, as Barbara Miller put it, “an archi-
tecture of a new beginning and social order... asearch fora
new ‘spirit’ for anew community” (Lane 1968, p. 60). The mod-
ern movement was more than a transformation of past practices,
such as the historicist styles of Behrens, Bonatz, and other late-
nineteenth-century architects. It was a complete departure, even
a rejection of such earlier orientations.

By now the history of the development of this departure is well

known. Throughout, there was an embracing of the apparent op-
portunities presented by new building materials, new fabrication
technologies, and new models in the form of artifacts of the ma-
chine age, particularly in the areas of flight, navigation, and
overland travel (Le Corbusier 1959). ‘‘The architect is the
organizer of the building sciences,’ asserted Hannes Meyer
(Schnaidt 1965, p. 31).

The conception of a new social order became manifested in the

stress that was placed, in housing for instance, on the individual
unit—on the manipulation of the unit in direct conformance with
the functions it was to perform and on the equitable apportion-
ment of exterior site amenity among units. In public housing proj-
ects, or Sied/ungen—such as those at Weissenhof, Magdeburg,
Dessau, and Berlin, by the architects of Der Ring, the Bauhaus,
and CIAM (see figure 50)—the individual units, sometimes de-
tached, were carefully designed to accommodate the functions
of day-to-day life, and each had a more or less equal share ina
hierarchy of public and private spaces. Aggregation of units
maximized benign solar orientation, and a relative independence
from the ground plane was provided through the use of pilotes
and other devices. This strategy again effectively preserved, if by
a sort of denial, an equitable distribution among unit dwellers of
amenity and opportunity with respect to the street and outdoor
recreation areas.

As others have eloquently pointed out, what seems to underlie

the distinctions that were made in these schemes is a kind of
“naive functionalism”’ (Rossi 1982, pp. 46—48), in which a close
and inevitably static fit between function and form was striven
toward along the lines of a functional determination involving
discrete classifications of working, dwelling, and playing activi-
ties. Laudable though the notions of equitable access to accom-
modations, sunlight, and the public domain may have been, in

Architectural Positions and Their Realms of Inquiry 155

 ee

‘as y wee
ann ae
ren en i
* Sn eee ot Sim wee
; Pi | we a
ser. i ad a ae

50
Apartment buildings in Siemensstadt, Berlin, by
Walter Gropius (1929).

practice many projects, especialiy in the hands of lesser archi-

tects, exhibited a tendency toward homogeneity, standardiza-
tion, and even monotony. This tendency may have been inevitable
under the pressures of equitable distribution of spaces;
a highly deterministic, and in many senses narrow, categoriza-
tion of function; and moves toward standardization of fabrica-
tion systems and materials for their promise of reduced costs.

These shortcomings were recognized early on. Emil Utitz and

Schultze-Naumberg, among others, began in the period 1923-
1926 to voice criticism of the ‘‘new architecture’ (Utitz 1923, p.
144). In fact, the appellation “‘functionalism” (Zweckmassigkeit)®
was coined to describe an architecture of engineering and plan-
ning that, while workable, economical, and efficient, never-
theless lacked qualities of ‘“‘coziness, individuality, warmth and a
general rejoicing in the act of dwelling” (Lane 1968, p. 131).
Criticism was also leveled at the universality of the ‘‘new style,”’
which seemed to be built for anyone, anywhere. This visual char-
acteristic is attributable to ideas taken from the three sources

Architectural Positions and Their Realms of Inquiry 156

 mentioned earlier—a narrow interpretation of function, equality
of access to site amenities, and the use of modern production
technology. Propositions about choice behavior, conflated with
ideas of “typical” and ‘rational’ man that were becoming preva-
lent at the time in other spheres, such as the behavioral sci-
ences, were particularly influential.

In one of those disastrous moments of collision with the forces

of history, the new architecture, which the critiques of the early
1920s had identified as a ‘‘style of labor and social utility’ having
a direct association with proletarian movements and Bolshe-
vism, became the target of the political right in the 1930s. The
fate of the Bauhaus at the hands of the National Socialist gov-
ernment of Germany is well known (Lane 1968, p. 147f.).

Of course Gropius, Taut, and others did not see their architec-
ture as a mere product of function and technological opportu-
nity. To them it represented the possibility of a new aesthetic in
which an abstract geometric order would emerge, symbolizing a
new practical realism amid the chaotic ferment of modern life.
“The old forms are in ruins,’ wrote Gropius; ‘“‘we float in space
and cannot yet perceive the new order. .. . What is architecture?
The crystalline expression of man’s noblest thoughts” (Conrads
1964, p. 46). Again to quote Gropius, ‘‘The ultimate aim of all
visual arts is the complete building” (Wingler 1969, p. 31). Here
we begin to step across the boundary between our two general
views of architecture. The modernists were clearly expressing a
concern for architecture in itself. Yet theirs was not a strategy
that operated entirely within the rhetorical resources of architec-
ture and its constituent elements. It was an effort to devise a new
aesthetic order without reference to past orthodoxies and with
an express interest in the materials, techniques, concepts of be-
havior and ethics, and political developments of the time. A
paramount principle of the modern movement was the prohibi-
tion against all direct stylistic references. This point will be taken
up later, however, in the discussion of architectural language.

The spirit of ‘form follows function” and the preoccupation with

new building technologies were, of course, to live on and thrive
beyond the early modern period. The CIAM (Congress Inter-
nationale d’Architecture Moderne) was to give way to, or be-
come identified with, the so-called international style (figure 51).
In the postwar years a reappraisal and considerable reaffirma-
tion of earlier positions was to be found in the pronouncements,
proposals, and projects of Team Ten (Smithson 1968). As Colin
Rowe and Fred Koetter were to write much later in something of

Architectural Positions and Their Realms of Inquiry 157

Buildings in the central area of Brasilia (Plaza of the
Three Powers), by Oscar Niemeyer (1960).

an epitaph, ‘‘If the combination of fantasies about science—with

its objectivity—comprised one of the most appealing and
pathetic of late-nineteenth-century doctrines, then the decisive
twentieth-century embodiment of these themes in the form of
building could not fail to stimulate; and, the more it excited the
imagination, the more the conception of a scientific, progressive
and historically relevant architecture could only serve as a focus
for a still further concentration of fantasy’ (Rowe and Koeiter
1978, p. 4).

Planning Orthodoxy and the Form of Cities

In other realms concerned with the building of environments,
orthodox approaches of a positivist stamp emerged with similar
traits to those of the modern movement in architecture. Indeed,
when we consider the extensive influence of town and regional
planning practices and resulting policy and legal instruments
such as zoning, systems of economic incentives and disincen-
tives, and similar land-use and building control mechanisms, the
issue of managing and shaping buildings, towns, and cities ap-
pears inextricably engaged with architecture. To be more accu-
rate, modern forms of managerial intervention and development

Architectural Positions and Their Realms of Inquiry 158

 predate the architectural movements of the early twentieth cen-
tury; and to the extent that a reciprocal relationship existed be-
tween planning and architecture, the direction of influence was
more pronounced from the former to the latter (Scott 1969, pp.
1—46). Urbanism was confidently viewed as essentially a case of
functional description, analysis, mediation, and control, all pred-
icated on good ‘‘scientifically based’’ methods of interpreting
the world (Mies van der Rohe 1923, p. 89).

From relatively early in this century, at least within the Anglo-

American tradition of planning, the promise of an emphatic
knowledge of human behavior—a science of human behavior—
fueled attempts to engage in what later became decried as ‘‘so-
cial engineering” (Scott 1969, ch. 3; Krueckeberg 1983; Clawson
and Hall 1973). The terrain of cities was subdivided along the
lines of distinct and discrete patterns of use, with very little op-
portunity for mixing (figure 52). Urban land and buildings be-
came highly valuable commodities, owing to the placement of
transportation systems and their associated gradients of accessi-
bility, and to the separation and concentration of functions. Af-
ter all, the home environment should be just that, a residential
sector with support from services such as schooling, shopping,
and recreation, while places of work should be aggregated and
serviced with their appropriate supporting functions. Such plans
were deemed efficient. Intervening corridors provided access
and with it better and worse opportunities for diminishing times
and other inconveniences of travel. The results of these exer-
cises provided a certain predictable and manipulable order,
setting the stage for further episodes of development and
redevelopment.

The appropriate accommodation and celebration of special

uses, such as monuments, became something of a problem amid
the general tendency toward standardization and homogeneity
of uses. Rossi, for one, eloquently criticizes this aspect of mod-
ern cities (Rossi 1982, p. 55f.). More often than not, the problem
was addressed through strategies like ‘‘connectivity,” in which
certain routes that already existed for other functional reasons
were accentuated in ways that allowed for the display of special
uses. This strategy did not attempt to break in any way with the
entrenched linear continuity of land use and movement patterns
(Crane 1964). Finally, certain general norms of behavior and as-
pirations were adduced, with the consequence that all-inclusive
social contracts were struck (zoning ordinances, for instance).
The result was that people on one side of a city with no real
relationship to those on another found themselves responsible

Architectural Positions and Their Realms of Inquiry 159

52
Suburban tract housing in the United States.

Architectural Positions and Their Realms of Inquiry 160

 for each other. By the 1950s the modern city had come very
much into being.

Behind the scenes the influence of ‘‘scientifically based”’

methods of interpreting man and his world can be seen at work
in at least two ways. First, there is the general practice of for-
mulating problems concerned with the built environment in
terms of the empirical orthodoxy’s accepted procedures. Here
the results of what count as valid and reliable exercises in mea-
surement and analysis tend to determine which problems are
identified and how they are understood.

The second manifestation of this orthodoxy’s influence may be

found in the widespread practice of ‘‘mapping”’ scientific explan-
atory theory about the relationship between man and his world
into “‘instrumental action.” For example, empirically based ex-
planations of the trip-making behavior of urban residents often
become the bases on which strategies to solve various transpor-
tation problems are developed. Similarly, an inference about
environmentally defined patterns of behavior may be used
unquestioningly as a basis for architectural design.

Generally, instrumental action means either a form of social con-

tract that is used to regulate the conduct of human affairs as
they relate to the built environment, or a physical intervention in
the urban landscape that is proposed for similar purposes. Map-
ping is the process by which information from the formulation of
a problem serves as a basis for the development of a social con-
tract, designed intervention, or other instrumental action. For in-
stance, a housing problem might be understood as such through
the systematic gathering of empirical data, followed by the pro-
mulgation of some policy instrument (Such as economic incen-
tives) or a specific proposal about buildings that has meaning
within the particular problem formulation.

Both influences are evident in the typical technical parapher-

nalia of urban design and planning—various schemes and
methods for performing quantitative analyses of empirical data,
computer models of a statistical or logical variety that are in-
tended to mirror the subject matter of interest, “‘quantitative indi-
ces’’ of performance that link some measurable part of the world
to some abstract dimension of quality (Dickey and Watts 1978,
Catanese 1972). Witness also the technical undertakings of gov-
ernment agencies, either in-house or through sponsorship, that
support legislative activities directly concerned with building
and urban issues.

Architectural Positions and Their Realms of Inquiry 161

 @ Systems Approaches to Interpretation
At the heart of these ‘‘scientific’’ methods for interpreting man
and his world for design purposes are empirical systems ap-
proaches—more particularly, the concept of model and the ac-
tivity of operational modeling (Chadwick 1972, La Patra 1973,
Handler 1970, Ferguson 1975). For example, there is the com-
mon use of behavioral models, of one sort or another, that
purport to characterize the habitual or special behavior of
individuals and groups contingent upon a variety of environmen-
tal circumstances. In general, for the ‘‘social bases’ of design
contemporary practitioners place a heavy reliance on interpreta-
tive frameworks that come directly out of social sciences such as
anthropology and sociology.

In the earlier work of the current era by researchers like Hall and
Sommer we see ‘‘territoriality” introduced, or reintroduced, with
its concomitant concepts of ‘personal distance,” the effects of
crowding on social behavior, and the perception of space (Hall
1969, Sommer 1969). Cross-cultural comparisons were included
to reinforce the contingent and relative importance of these di-
mensions of man-environment relations, and detailed case stud-
ies were conducted around specific activities such as learning
and recreation. Barker, in an influential work dealing with ap-
proaches for studying environments of human behavior,
broadened the methodological range in the direction of ecolog-
ical concepts for defining fundamental ‘environmental units”
and their structural interrelationships (Barker 1968; Proshansky,
Ittelson, and Riulin 1970). Lynch’s early studies on the image of
the city also gave considerable weight to empirically determined
concepts of spatial perception, and the idea of “‘cognitive map-
ping” was to become highly influential in the theory and practice
of urban design (Lynch 1960).

Subsequent investigations and proposals of design criteria by

Rapaport, Craik, and Zube, as well as Moore and Golledge, elab-
orated on these themes, extended the range of methodological
possibilities, and reiterated the importance of a detailed explicit
understanding—a model—of man-environment relations
(Rapaport 1969, 1977; Craik and Zube 1976; Moore and Golledge
1976). At the same time, texts appeared that brought these ideas
and empirical findings to bear more directly on the design pro-
cess. The anthology of Lang and colleagues was aimed at the
practicing professional as much as the researcher and provided
both prescriptive advice and conceptual frameworks for ad-
dressing the design of various settings (Lang, Burnette, Moleski,
and Vachon 1974). Later works by Sanoff and Preiser were even

Architectural Positions and Their Realms of Inquiry 162

 more specific and specialized in the area of facility programming
(Sanoff 1977, Preiser 1978).

The Concept of Model

A review of literature dealing with modeling reveals its kinship
with the logical-empirical approach and the hypothetical-deduc-
tive system of theory construction (Krimerman 1969). Under this
system scientific explanation has two major components, an ex-
planandum describing the phenomenon to be explained and an
explanans containing statements of antecedent conditions and
of the all-important general laws that account for the phenome-
non. For example, if a vessel containing a fluid explodes when
heated, the event can be accounted for by physical conditions
surrounding the event and by laws, such as that pressure is di-
rectly proportional to temperature (Hempel and Oppenheim
1948). The explanatory scheme incorporates logical deduction
whereby the explanandum can be deduced from the explanans
in a straightforward Aristotelean manner (Mitroff and Kilmann
1978, ch. 3).* In practice, the development of laws in the social
sciences invariably involves methods of statistical reasoning by
which a conformance between data and theoretical regularities
is established. Here the aim is to ensure that a relationship be-
tween the hypothesized occurrence and the real-world occur-
rence of variables of interest can legitimately be struck under all
conceivable conditions (Campbell and Stanley 1970). The ap-
proach is hypothetical insofar as no claim is made that any of
the laws and premises involved are infallible (Hempel and Op-
penheim 1948).

According to Echenique, ‘“‘A model is simply a representation of

relevant characteristics of a reality .. . ameans of expressing
certain characteristics of an object, or system, that exists, ex-
isted, or might exist’ (Echenique 1963, p. 1). Clearly, further
elaboration of this definition requires consideration of what is
meant by ‘“‘relevant characteristics of a reality’ and considera-
tion of alternative means of representation.

The first part of model making, the “selection of relevant charac-

teristics,” is seen to be a process framed by the intentions of the
model maker (Echenique 1963, p. 2f.). In other words, it is the
questions that the model is designed to answer that will deter-
mine the selection of relevant variables, antecedent conditions,
and so on. In this regard the concept of model does not really
differ from the idea of theory construction in the social sciences.
Both situations must begin with intentions, even if these are not
generally recognized—are, as Popper calls them, “private mat-

Architectural Positions and Their Realms of Inquiry 163

 ters’ (Popper 1959). How the actual selection of relevant char-
acteristics is appraised and verified is considered largely a
technical matter, involving good practice within the logical-
empirical system of theory construction.

Consideration of alternative means of representation is also

taken to be a technical matter, dictated by the situation at hand.
For example, the choice of means may be strongly determined
by how time, location, or other characteristics of the behavioral
setting under scrutiny either need to be represented or can be
incorporated into the model structure. It is also here that difficul-
ties can be encountered. For example, available means and
modeling “languages” may constrain the model builder’s inten-
tions. Such constraints may also result in the reformulation of a
problem in line with available techniques of representation, with
a concomitant abandonment of certain fundamental aspects of
the original problem under consideration. The problem focus
thus becomes shifted, and in ways that might later prove quite
damaging in the context of broader social areas of concern. For
example, the planning and design of a transportation system
within an urban area, based upon the analysis of data about past
travel behavior, may ignore aspects of the system that are inher-
ently socially regressive for certain groups (Such as gasoline
taxes for low-income groups). This ignorance may be due to an
underrepresentation of pertinent qualitative data, or it may be
the consequence of socially regressive effects that could not be
adequately described within the systematic framework originally
used for the planning and design exercise.
In summary, the process of model making generally conforms to
the following five steps, although not necessarily in the outlined
order. First, the existence of an object, setting, or system that is
of interest. Second, an intention, clearly expressed, enabling the
selection of appropriate characteristics of the object, setting, or
system. Third, a process of observation and abstraction enabling
the reality in question to be observed in relationship to selected
variables. Fourth, a process of translation, enabling the creation
of a suitable conceptual framework for organizing factual infor-
mation. And fifth, a process for testing and making conclusions
about the congruence between the model and reality. This last
operation is often referred to as ‘calibration,’ or the fitting of
some previously determined regularity to the specific circum-
stances of the locale under scrutiny.

Architectural Positions and Their Realms of Inquiry 164

 Models and Theory
To Echenique and others, the differences between models and
theories are principally matters of degree or status rather than
ones of kind. The intellectual orientation for both is clearly the
same (Echenique 1963, p. 3f.; C. Lee 1973, ch. 1; Collins 1976). In
fact, Echenique suggests that a model can stand as a theory to
the extent that it is a system of ideas that is held as an explana-
tion for a phenomenon or group of facts; he clearly means
theory in the ideal scientific sense.

The conformity of this concept to social science’s orthodox ap-

proach toward hypothetical-deductive theory construction is not
surprising. The latter is, more often than not, the acknowledged
ideal for the former. However, when it comes to the activities of
operational modeling, involving practical application for design
and policy planning, certain distinctions can be made between a
theory and a model. Alternatively, one can distinguish between
kinds of models, according to their predictive powers: theoret-
ical models are idealizations and weakly predictive, whereas
analog models based upon empirical data are not (Hesse 1963).

Operational Modeling
lra Lowry, an early luminary in the field of modeling and land-use
planning, makes the following distinction: “‘In formulating his
constructs the theorist’s overriding aims are logical coherence
and generality. ... He is ordinarily content to specify only the
conceptual significance of his variables and the general form of
their functional interrelationships. ... The model builder, on the
other hand, is concerned with application of theories to a con-
crete case, with the aim of generating empirically relevant output
from empirically based input” (Lowry 1965, p. 160).

In certain respects Lowry’s distinction can be seen as nothing

more than an operational one—as accommodation to the de-
mands of applying a model to specific situations. In effect, a
concern for the identity and direction of relationships is shifted
in operational model building to the joint specification of both
direction and magnitude. In other respects, however, particularly
in the practical realm, the distinction implies an extension of
theory. It is a situation where extant theory provides the building
blocks for model construction, but where the model is not
necessarily fully subsumed under theory and therefore becomes
something different. More precisely, the model is based on
theory but extends beyond theory in some respect.

The kind and magnitude of the difference or distinction can be

dictated by several circumstances (Lowry 1965, Steinitz and Ro-

Architectural Positions and Their Realms of Inquiry 165

 gers 1970). First, the resources available for model construction
and application may not fully permit a faithful theoretical ac-
count of the phenomenon to be made. Second, theory may gen-
erally cover the case in question but lack precision in particular
considerations of magnitude; for example, the direction of a re-
lationship may be known, but its functional form, in the formal or
mathematical sense, may not. Third, theory may not yet exist
covering some parts of the modeling exercise that are at any rate
deemed relatively insignificant, thus requiring a patching to-
gether of promising hypotheses that have not been tested. Fi-
nally, theory may be applied in a model, but under conditions of
application that extend beyond the reasonable limits of the
theory’s antecedent conditions.

The extension from theory to operational models in Lowry’s

sense can also be seen when one examines a fairly commonly
accepted hierarchical classification of models (Lowry 1965;
Steinitz and Rogers 1970, p. 8). Typically, four types of models
are identified, according to the general purposes of their appli-
cation. They are (1) descriptive models, (2) predictive models,
(3) explorative models, and (4) planning models.

The principal intention behind a descriptive model is explanation

of phenomena in the domain of interest. For instance, in a build-
ing or urban context a good descriptive model would reveal
much about the underlying structure of the building or urban
environment. It would reduce the apparent complexity of the ob-
served world to a coherent and rigorous framework. Further-
more, the descriptive model is logically essential for any of the
other three types. We cannot predict, explore, or plan, at least in
the view of this orthodoxy, in the absence of a fundamental de-
scription of the reality under study.

The merit of a descriptive model can be assessed according to

three broad criteria. First, the required ratio of input data to out-
put data must be small; that is, explanatory power must be high
(Lowry 1965, p. 160). Second, the conformance of model output
to direct observation must be high; that is, the model must be
accurate. Third, the model should be generally applicable—ap-
plicable beyond a specific spatio-temporal domain.

The purpose of a predictive model is to give a forecast of the

temporal disposition of the phenomenon under study. These
models can be of two types. Extrapolative predictive models are
those in which the spatio-temporal change process is repre-
sented by a continuation of past trends—for example, by using a
“line of best fit’’ applied to historical data. The obvious disad-

Architectural Positions and Their Realms of Inquiry 166

 vantage here is that the change process in the future, or even in
the present, may be discontinued or radically changed unbe-
knownst to the model maker. Recent models of settlement and
urban structure that make use of the “‘order by fluctuation” prin-
ciples propounded by Prigogine in chemistry have shown the
possibility of sudden “‘bifurcation’’ of phenomena due solely to
the timing of events (Allen 1981). This is a situation where the
system is capable of reaching an entirely different state, due not
to the types of processes at work per se but to their temporal
conjunction. For instance, an interaction of otherwise indepen-
dent and conventional land development decisions may radically
change the type and pattern of economic activities in a geo-
graphic location.

Conditional predictive models are those in which the mechanism

of cause and effect governing the change process is specified in
the general form “‘If X is present at time t,, then Y will transpire at
time to.’”’ Clearly, the advantage of this type of model is that the
process of change, or a significant portion of it, is endogenously
specified—specified within the model. It does not rely solely on
translation of past events into the future.

Explorative models are usually designed to allow the discovery,

by systematic speculation, of realities other than the one at hand
that may be logically possible. Such speculation usually pro-
ceeds by exploiting the systematic variation of basic parameters
used in a descriptive model. A case in point is Zwicky’s cele-
brated ‘‘morphological analysis” that leads to the invention of a
particular type of jet engine (Zwicky 1962). Similar examples may
be found in Simon's discussion of a ‘‘science of design” regard-
ing an instrument for telling the time (Simon 1969, p. 6f.).

Within the framework of this kind of model, a functional analysis

is usually performed on the type of system under consideration.
From this analysis various alternative means for achieving re-
quired functions are identified, such as means of combustion,
torque conversion, and so on. Compatible means, technically or
otherwise, are then combined covering the required functions.
Sometimes the results are novel solutions. An analogous ap-
proach can be employed in examining behavioral and cultural
settings of interest, particularly when there is a premium being
placed on new insights about underlying structure or behavior
patterns. The general approach has also been proposed and ap-
plied in architectural design situations, again principally for the
purpose of reexamining conventional practices.

Architectural Positions and Their Realms of Inquiry 167

 A planning model necessarily incorporates prediction, usually of
the conditional kind, but is extended to allow for the evaluation
of predicted outcomes in terms of goals. In other words, this
type of model is primarily developed for simulating the effects of
different decisions about an environmental and behavioral set-
ting, and evaluating those decisions or strategies against a
specified goal structure (an expression of ‘“‘good”’ performance).

Typically, the structure of planning models has the following

components. First, a specification is made of alternative pro-
grammatic or design interventions that might be applied through
some course, or courses, of action. Second, a predictive model
is developed, allowing the consequences of choosing any
specified alternative to be revealed. Third, there is a method of
scoring and ranking the consequences so revealed according to
some metric of goal achievement; that is, against some objective
function. Finally, there is.a mechanism for making the choice of
(presumably) the alternative that yields the highest score. For ex-
ample, a planning model developed expressly for the purpose of
resolving floor-plan layout problems requires specification of the
type of allocation procedures to be considered. It must also al-
low various room arrangements to be evaluated against specific
metrics of performance, such as adjacency and transportation
cost.

Classification of the models is hierarchical because one model

type forms the foundation for others, with planning models
located at the furthest extremity (see figure 53). The hierarchy
also illustrates the steps in the construction of a planning model,
including the use of an explorative model to test variables for
possible inclusion in the model of conditional prediction.

@ Problems of Interpretation
Of the types of models examined, planning models would most
frequently seem to be extensions beyond empirically based ex-
planatory theory. They incorporate goals presumably based on
normative positions, which are largely precluded from the idea
of explanatory theory. They seem to draw, as it were, on informa-
tion from elsewhere. In contrast, the other models appear to
conform more strictly to the boundaries of the idea of theory
advocated by a logical-empirical orthodoxy.

The question then becomes whether planning models can mean-

ingfully transcend their inherently explanatory function of ‘‘sci-
entific interpretation” to any significant degree. Can they
incorporate other values and qualities not directly accounted for
by this interpretation? The answer is probably no. For one thing,

Architectural Positions and Their Realms of Inquiry 168

 Coane)” ( Planning)

Past > Present > __ Future Reality

53
A hierarchy of model types.

a cursory survey of practical applications and experience with

planning models would reveal an assortment of metrics for scor-
ing achievements, each with its own requirements and limita-
tions for accommodating the specification of goals. This would
also apply to predictive models. For instance, cost-benefit analy-
sis in its various guises, a mainstay for discriminating among
various design alternatives, incorporates arcane formulae for
what is counted versus what is not (Mishan 1976). The very stric-
tures of these formulae, however, preclude any real advance be-
yond the realm of ‘‘facts’’ into the domain of ‘‘values,”’ except in
rather crude terms. For instance, affective aspects of choice be-
havior are largely precluded, and there is a tacit assumption of
economically rational behavior.

Yet it is with a second, related characteristic of planning models

that the in-principle restrictions imposed by the very ideals of
scientific interpretation come into play. In such models, out-
comes are evaluated in terms of goals. For this to be logically
accomplished, a reasonable correspondence must be struck be-
tween an expression of an outcome and an expression of a goal.
Given the inherent limitation of planning models as to movement
away from strictly factual expressions, the goals inevitably be-
come defined by, and therefore limited to, the currency of fac-
tual terminology. Furthermore, the converse also appears to be
true in the process of model construction itself. For example, in
econometrics and similar undertakings, there is the ubiquitous
problem of appropriately representing various policy options in
the ‘‘language”’ of a model (Domencich and McFadden 1975).

Architectural Positions and Their Realms of Inquiry 169

 The central issue here is not so much whether planning models
exhibit explanatory power or predictive accuracy as that a de-
pendent relationship can very easily develop between the pos-
sibilities for formulating goals or design interventions and the
technical exigencies of model development and application. Fur-
thermore, these are exigencies that seem to be largely due to the
pedigree, so to speak, of their ‘“‘value-free”’ theoretical underpin-
nings. But if design and policy formulation is to be freed from
the methodological restrictions that derive from the epistemol-
ogy of the logical-empirical orthodoxy, to what other guiding
theoretical orientation might it adhere? What are the alternatives
to the science of human behavior?

@ Alternative Positions
A number of authors have developed comprehensive and cogent
criticisms of the empirical orthodoxy’s idea of theory and the
possibility of a science of man. For example, Bernstein in his
classic work The Restructuring of Social and Political Theory
brings to bear arguments from analytical philosophy, phenome-
nology, normative political theory, and critical rationalism
(Bernstein 1976); Krimerman presents a well-organized anthol-
ogy representing both sides of the debate (Krimerman 1969). The
following brief account of critiques draws on these and other
works and is intended to give a taste of the kinds of arguments
that have been advanced, rather than anything approaching a
thoroughgoing account.

A Humanist Critique
One kind of criticism revolves around drawing a distinction be-
tween the “outward” manifestations of human activity and ‘“‘in-
ner’ systems of belief or thought that bear directly on this
activity. For example, Isaiah Berlin challenges the claim that all
real questions about man and his world can be answered
through empirical and formal methods of the kind we have been
discussing. To Berlin, ‘‘men’s beliefs in the sphere of conduct
are part of their conception of themselves and others... . This
conception, whether conscious or not, is intrinsic to their picture
of the world” and thus is central to social theory. Berlin does not
merely conclude that social theory must embrace beliefs and
convictions; he goes much further, asserting, ‘‘It is not simply a
case of the existence of subjective states of mind but rather self-
interpretation as constitutive of actions and practices”’ (Berlin
1963, pp. 13-15). As Bernstein observes, Berlin effectively pro-
vides a picture of social behavior in which we are dominated by
models of what the world is like and these models determine
both the content and the form of our beliefs, convictions, and

Architectural Positions and Their Realms of Inquiry 170

 behavior (Bernstein 1976, pp. 59-63). Any change in the guiding
model necessarily results in changes in both belief and behavior.

The primary issue here is that dichotomies between the ‘‘objec-

tive” and the ‘‘subjective,” between ‘‘fact” and ‘‘value,” entirely
misrepresent human action. Any proper description of objective
action is internally related to interpretations of what constitutes
action. This interpretation, in turn, is largely a matter of belief,
self-awareness, and value. For Berlin there is ‘‘human action,”
and to speak of its factual and value sides is ridiculous (Bern-
stein 1976, p. 61).

Another critique, in a similar vein, is propounded by the histo-

rian Collingwood in his discussion of the ‘‘a priori impossibility
of a science of man’”’ (Collingwood 1946, pp. 205-231, 315-320:
Krimerman 1969, pp. 7-25). Here he is referring to a science of
man based on analogy with the methods of natural science and
thus of the orthodox empirical variety. Collingwood makes a
case for the primacy of history in these matters through its fun-
damental concern with events that are consequential beyond the
realm of biological functions such as eating and sleeping. He
also makes a distinction between the “‘outside”’ and the “inside”
of an event, where the ‘outside’ corresponds to all the docu-
mented circumstances about the event and the “‘inside’’ refers to
the thought behind the event. In short, he advocates an inter-
pretative scheme for history in which the aim is to see through
the event to the thought that lies behind it. For Collingwood sci-
ence and its methods are about ‘‘processes of events” and their
interconnections, whereas history, or historical knowledge, is
about ‘“‘processes of thought.”’ Therefore, while he seems more
willing than Berlin to entertain a distinction between facts and
thoughts, he clearly concludes that a science of man, in the or-
thodox empirical sense, is an impossibility.

Structuralist Similarities and Divergences

In many respects the salient features of structuralism resemble
the idea of theory and general theory development advocated by
the empirical orthodoxy. At least this seems to hold among com-
mon themes and approaches to be found in the works of Lévi-
Strauss, Piaget, and Leach (Gardner 1981; Piaget 1970a, 1970b).
Structuralism is founded in an ‘‘ardent and powerfully held con-
viction that there is structure underlying all human behavior and
mental functioning.” This structure ‘“‘can be discovered by or-
derly analysis and . . . once discovered it has cohesiveness,
meaning and generality” (Gardner 1981, p. 10). What proponents
mean by structure takes on a variety of forms, usually expressed

Architectural Positions and Their Realms of Inquiry 171

 in terms of the subject matter in which they are dealing. For
Piaget in his developmental psychology, it is represented by the
immutable stages of cognitive development from conception to
adult life (Piaget 1965, 1970a, 1970b). For Chomsky it is the pres-
ence of a single deep structure that is the root of all language
(Chomsky 1965). There is a strong adherence to a ‘universal
base hypothesis,’ whereby all manner of human thought, activi-
ties, and experience can be traced to a common foundation,
regardless of culture, economic position, or other surface
manifestations of the human condition.

Apart from the similarity apparent in the foundational inclination

of theory, parallels with the empirical orthodoxy are also evident
within the framework of methods for theory construction. In sim-
plified form, this framework may be depicted as follows (Gardner
1981, p. 15f.). First, if regularities are apparent, or seem likely to
hold, then the behavioral phenomenon can be subject to struc-
tural analysis. Second, a diverse set of phenomena can be
related to one another once common factors have been
discovered. In the ‘developmental method” of structural analy-
sis espoused by proponents such as Piaget, this means distin-
guishing between two kinds of factors. They are ‘‘synchronic
elements,” or factors that do not exhibit temporal variation, and
“diachronic elements,” or those within the domain of inquiry
that change with time. These elements can be further distin-
guished into two subgroups, namely, “irreversible elements”
that move solely in one direction and “reversible elements” that
include shifts from one pole to another, actions and reactions,
and cycles of events. In ‘“‘nondevelopmental”’ or ‘“‘agenetic ap-
proaches’ it is assumed that all relevant factors are present in
some form throughout the period and across the domain of in-
terest. The third and final general step in the procedure is the
devising of a formal or informal model of underlying structures.
It is a model that must have the characteristics of accounting for
the data in an economic and comprehensive fashion and being
applicable to data yet to be collected. Thus, the model must be
divorced from concrete reality, to the extent that it is possible;
that is, it must not make specific reference to the particular case
being investigated. Models cannot be mere empirical generaliza-
tions, although they must be reflections of observations made by
empirical observation.

Clearly this framework coincides with the theoretical ideals

of the empirical orthodoxy. Beyond these general working
methods, however, structuralism in practice parts company with
the positivist empirical temper. Its intellectual heritage is pro-

Architectural Positions and Their Realms of Inquiry 172

 foundly influenced by a Rousseauan opposition to Cartesian and
other mainstream Enlightenment positions (Gardner 1981, pp.
15-47). Hence structuralism does concern itself with the affec-
tive and irrational nature of man. It has also placed considerable
emphasis on nonempiricist intellectual activities of theorizing,
such as ‘“‘arm-chair’ or “‘thought’’ experiments.

The Problem of Tacit Conventions

As proposed by Bernstein, yet another critique, or alternative
position, comes from the direction of analytical philosophy and
certain kinds of linguistic analysis closely associated with Witt-
genstein and others (Bernstein 1976).° Here theorists such as
Winch involve the notion of ‘‘forms of life’ as being fundamental
and given in any description of human behavior (Winch 1958,
Wittgenstein 1969, Bloor 1983). Furthermore, it is these forms of
life that constitute the essential questions to be pondered. The
necessary and sufficient ingredients that define and distinguish
forms of life are intersubjective conventions and agreements; for
example, speech is understood via complex intersubjective
agreements about language. These intersubjective conventions
manifest themselves, in turn, by ‘“‘rule following” behavior and
hence are logically indispensable from the idea of ‘making a
mistake” or the valuation of some things that count from those
that do not. In short, Winch concludes that social relations are
governed by the language, ideas, concepts, and beliefs, as well
as the reflective and unreflective actions, that constitute institu-
tional practices, and therefore cannot be separated from them in
any meaningful manner (Bernstein 1976, pp. 63-74). Like Berlin,
he seems to be arguing that a purely ‘‘objective”’ or “‘scientific”’
rendition of human behavior may run the risk of misunderstand-
ing the appropriate concept of behavior, by divorcing this con-
cept from a detailed explanation of outward manifestations.

The Phenomenological Alternative

The “phenomenological alternative,’ as Bernstein calls it, to the
empirical orthodoxy interposes concepts such as Husserl’s
Lebenswelt in the ‘scientific’? account of man and his world
(Bernstein 1976, pp. 117-169). The alternative might also include
Merleau-Ponty’s idea of situations and Heidegger’s being-in-the-
world (Mallin 1979, chs. 1-2; Heidegger 1962). Roughly speak-
ing, the Lebenswelt is the world we live in and the world of
everyday experience (Husserl 1975; Kockelmans 1967, p. 31f.). In
Husserl’s earlier accounts, at least, it covers the circumstances
surrounding commonplace events, including those that are af-
fective and intentioned. All human thinking and behavior in this
everyday world, argue the phenomenologists, has an indispens-

Architectural Positions and Their Realms of Inquiry 173

 able core of conceptual equipment that has no history and
embodies concepts, commonplaces, and views of the world
corresponding to the time when man first encountered himself.
This core does not change; as we might expect, the objective of
phenomenological inquiry is to reveal it and thus to address
questions about the ‘‘essence of things.”

According to Bernstein, where the phenomenologists part com-

pany with the empirical orthodoxy is in their conclusion that it
makes no sense to replace the core of conceptual equipment
with a ‘‘scientific’’ account (Bernstein 1976, pp. 126-135). Sci-
ence, like any other conceptual equipment, must presuppose
this core. For people like Merleau-Ponty, science is thus a sec-
ond-order expression of basic human experience and not
ontological (Mallin 1979, p. 20f.). This is not to say that all
phenomenologists necessarily assert ontological primacy for
their concept of everyday experience; Heidegger, for instance,
rejected Husserl’s schema as reductionist (Kockelmans 1967, pp.
228-230). In general, they argue for the juxtaposition and inter-
position of this kind of concept with that of objective science
and, therefore, a suspension of judgment about the reality of
either. The “‘bracketing’’ of both worlds that comes about
through such juxtaposition and interposition will result, so the
argument goes, in understanding through pure reflection, or the
enigmatic “transcendental turn.” It is only through this tran-
scendental turn that one reaches a complete phenomenological
understanding of the affairs of man, and positivist science fails
to make such a turn (Husserl 1970).

Logical Falsification and Ambiguity

Finally, in a more strictly logical vein, Duhem and later Fey-
erabend, both from the realm of the philosophy of science, have
attacked the commonly accepted notion that application of the
hypothetical-deductive method can result in falsification (Duhem
1914; Feyerabend 1978, p. 294; Mitroff and Kilmann 1978). In
fact, both argue, scientific experiments and procedures cannot
be decisively falsified.

Their argument begins with the assertion that scientists never

test a single hypothesis in isolation from other hypotheses, but
rather against a network of assumptions, auxiliary hypotheses,
concepts, and so on. Therefore, relationships with regard to a
hypothesis H; and an observation O can only be regarded as tak-
ing place within a network of n hypotheses. Thus, when observa-
tion of not O or O is in disagreement with the predictions of
theory, all one can legitimately conclude is that the hypothesis in

Architectural Positions and Their Realms of Inquiry 174

 question, Hj, is false. In fact, exactly which hypothesis is in error
can never be determined with complete assurance (Mitroff and
Kilmann 1978, p. 38f.).

Neither Duhem nor Feyerabend contend that falsification cannot

occur. They do, however, raise the issue that in order to falsify a
hypothesis, we have to assume the veracity of all the other hy-
potheses and auxiliaries entering into an experiment. Thus fal-
sification is ambiguous. Moreover, in a case where one is
considering the relationship between a hypothesis H and an ob-
servation O, if an experiment yields the result not O, the null
hypothesis can be concluded if and only if there are no assump-
tions that serve the original hypothesis and are also compatible
with not O. In other words, there would seem to be a certain
open-endedness to gradual elimination of false ideas through
experimentation (Feyerabend 1978, p. 35f.).

Arguments have been leveled at the ‘‘scientific interpretation” of

man and his world from other quarters. The attack from ‘“‘critical
theorists’ such as Habermas, for instance, has been omitted
from this discussion (Habermas 1979). So too have the doubts
expressed by otherwise staunch proponents of scientific inter-
pretation, such as Weber in his discussion of wertfrei (“‘value-
free’) investigations (Weber 1957). The foregoing discussion
should, however, have given at least some indication of the rea-
sonableness of raising fundamental questions about this form of
interpretation, in view of the plausible alternatives.

ESI
ee SaaS ae
Architecture from a Referential Interpretation

Over the last few decades there has been a noticeable shift of
emphasis in architectural discourse toward the world of archi-
tectural ‘objects’ and the use of its constituent elements as the
primary focus of design. ‘‘I really don’t think that architecture is
about social and political activity any more than | think politics
is about architecture,’ remarked Michael Graves at a recent
symposium. He went on to characterize architecture as “inven-
tion’ that makes up its own ‘‘text’”’ from the myths and rituals of
society—a text that in turn provides impetus to further invention
(Graves 1984). Such sentiments resonate with Wittgenstein’s
concept of ‘‘forms of life.’’ Arithmetic is just that, arithmetic, and
not a set of operations based upon mathematics and logic. To
engage in arithmetic one has only to know its constituent ele-
ments and not those of some other form of life, such as logic
(Diamond 1976, p. 271).

Architectural Positions and Their Realms of Inquiry 175

 This kind of preoccupation is by no means new. Each architec-
tural epoch has had its share of introspective interest in its own
constituent elements and their meanings. In the baroque era, for
example, precise architectural elements and compositional
principles from the Renaissance were transfigured, even
dematerialized, the baroque architects taking the principles of
the Renaissance as givens for their more willful explorations
(Millon 1983, pp. 9-10). The pluralism that is so characteristic of
current interest undoubtedly owes much, if only by reaction, to
the avant-garde stance of the modern movement, with its
concentration on abstract formalism and its steadfast denial of
stylistic reference or figural quality.

@ Strategies of Reinterpretation
At least three strategies can be seen at work in contemporary
efforts to reinterpret and develop the world of architectural ob-
jects. They are not styles in either the art-historical or the com-
monplace sense of the word. Rather, they may be regarded as
modes of interpretation with an essentially methodological incli-
nation: frames of reference with a judgmental as well as an ob-
jective dimension.

The three strategies will be referred to as work on the language,

bricolage, and the use of type. This is a somewhat arbitrary clas-
sification and one that is not at all novel (Colquhoun 1978a; Vid-
ler 1977, 1978; Rowe and Koetter 1978; Krier 1978; Dunster
1977). Work on the language encompasses both a consideration
of the formal component of architectural language, or the
“how,” and a consideration of the figural component, or the
“what” (Colquhoun 1978a, pp. 29-30). By contrast, bricolage
and the use of type will be discussed chiefly in relation to the
problem of choosing and reusing existing architectural objects,
or constituent elements. The classification is arbitrary because
the very applicability of the term Janguage may be questioned,
and because the use of type and bricolage may themselves be
regarded as work on the language insofar as they deal with mat-
ters of both figure and form. Yet verbal language is normally re-
garded as capable of describing something outside itself, but
with no particular agenda or purpose. People use the same
language to express many different ideas, whereas type and
bricolage are more ideological means of interpretation and
expression. In short, they are ways of using the language, or
more precisely, ways of appropriating and using constituent ele-
ments of architecture.

Architectural Positions and Their Realms of Inquiry 176

 Work on the Language: The Formal and Figural Traditions
Alan Colquhoun makes the following distinction between form
and figure: form applies to ‘‘a configuration with natural mean-
ing or none at all” (here ‘‘natural meaning” signifies meaning
without the overlay of intervening interpretative schema of a cul-
ture), whereas figure applies to ‘‘a configuration whose meaning
is given by culture.” Further, the figure organizes ideas and thus
is expressive and didactic (Colquhoun 1978a, p. 30f.). The two
components, he argues—and hence the formal and figural tradi-
tions in architecture—are not entirely disconnected, although
during certain eras the preoccupation may be more with one
component than with the other.

As is clear from the statement by Walter Gropius quoted earlier,

the modern movement had an undeniable preoccupation with an
abstract formalism. By denying any stylistic references of either
a formal or a figurative kind, the meaningful content of architec-
ture became indistinguishable from its form. This is not to say
that it had no meaning at all. On the contrary, it had a kind of
functional meaning that, as Colquhoun puts it, was ‘‘absorbed
directly into form” (1978a, p. 29).

In the hands of a consummate architect like Louis Kahn, a lyrical

architecture of the tectonic resulted from the use and expression
of frame, masonry wall, slabs, and vault, as evidenced in the
Kimball Art Museum in Fort Worth and the Salk Center in La
Jolla (figure 54). His was also an architecture in which the hierar-
chical disposition of “‘served and servant spaces,” very mucha
functional determination, played a considerable role in the over-
all organization. As Frampton and others have noted, when
Kahn’s architectural enterprise involved the need for an expres-
sion of monumentality or a question of saying ‘‘what,”’ he re-
sorted to the use of platonic arrangements such as squares and
circles, in much the same manner as Enlightenment architects
such as Ledoux and Boulée (Frampton 1985, Burton 1983). In
comparison to other figural devices that might be employed, this
seems like a last resort, and one that shows a certain discomfort
outside modernist doctrine.
The work of the ‘“‘New York Five” dramatized and extended the
geometrical abstractions of the modern movement in general
and of Le Corbusier in particular (Five Architects 1972). The
complex spatial interplay of solid and planar elements, beams,
and columns seemed more concerned with the very possibilities
presented by the world of such architectural objects than with
any other kind of meaning. As Peter Eisenman, one of the Five,

Architectural Positions and Their Realms of Inquiry 177

54
The Jonas B. Salk Center in La Jolla, by Louis I.
Kahn (1962-1966).

Architectural Positions and Their Realms of Inquiry 178

55
House Il, by Peter Eisenman (1969).

was to comment later, “‘It [the formal language] changed the re-
lationship between man and object away from an object whose
primary purpose was to speak about man to one that was con-
cerned with its own objecthood” (Eisenman 1978, pp. 21-22).

Eisenman’s own work seemed so completely to absorb func-

tional meaning into form that the architecture became a ‘‘pure’”’
work of art orchestrated according to its own internally consis-
tent formal rules (figure 55). Form seemed to be running the risk
of so totally subsuming function that it would transgress any rea-
sonable bounds. The ‘bedroom suite” in House VI, for example,
deconstructs the normal planar surfaces of wall and floor to the
point where questions can be raised about the suite’s actual
commonplace use (Gutman 1977). Eisenman’s work pushes Col-
quhoun’s definitions of form and figure to the limit. If the di-
dactic purpose is to explore architectural form for the sake of
art, then the work certainly can have a significant cultural mean-
ing beyond a “natural” one. As Eliot observed with respect to
the changing of meaning within a literary tradition, it is the effect
of new works to simultaneously rearrange the tradition and bring
about an alteration in the meaning of each text (Hirsch 1967, p.
215). Eisenman’s work would certainly seem to have accom-
plished this shift, to the extent that it calls for reexamination of
accepted traditions.

Architectural Positions and Their Realms of Inquiry 179

 The figural tradition has a long standing in architecture.
Panofsky eloquently describes the connection between Gothic
and Scholasticism and the synthetic power Gothic cathedrals
had for the purposes of transfixing, instructing, and transporting
the hearts and minds of the people (Panofsky 1967). The rich
iconography and allegorical elements of baroque architecture
formed a text whose meaning was also readily construed (Nor-
berg-Schulz 1980b). And classicism has spoken to us about or-
der and harmony as well as the place of man in the universe
(Kaufmann 1968).

After the hiatus represented by modernism a recovery of the fig-

ural tradition can be discerned. In fact, as Colquhoun argues,
such a tradition never fully disappeared. He cites oblique refer-
ences to the Beaux Arts on the part of modernists, as well as the
‘“neoclassicism’’ of many American works during the 1950s (Col-
quhoun 1978a, p. 29). Currently, however, the references are far
more literal and complete.

In the work of Venturi, Scott Brown, and others the recovery of

popular symbols is very much a hallmark. During a discourse on
the architecture of the American commercial strip—to continue
a discussion begun in the last chapter—various principles for
recovering a figural tradition were established (Venturi, Scott
Brown, and Izenour 1972). Buildings were classified as either
“ducks” or ‘decorated sheds.” A duck is a building whose con-
formation is a complete symbol or icon. A decorated shed is a
building to which symbols, often commonplace signs, have been
attached. All such signs and symbols were in common usage
and part of the ‘“‘vernacular’’ of the culture. It was argued that a
figural component to architecture could be developed that drew
upon ‘the ugly and the ordinary” as a legitimate rhetorical
resource.

The Venturi office has produced a number of decorated sheds.

It is both the distinctive and the ordinary features of the Guild
House in Philadelphia that demonstrate its figural qualities. Win-
dows look unabashedly like windows and, together with the
brickwork on the exterior, are quite suggestive in their ordi-
nariness. The ornamental horizontal white stripes near the top
of the building and the broad white panels are applied decora-
tion for the purpose of helping to organize the facade. The large
arched windows at the top center of the facade also help to or-
ganize the composition; in addition, they call attention to the
unique occurrence of a large common room (Venturi, Scott
Brown, and Izenour 1972, pp. 90-103). The recent proposals for

Architectural Positions and Their Realms of Inquiry 180

 the Laguna Gloria Art Museum, to be constructed in Austin,
Texas, are of a similar stamp (figure 56; Venturi, Rauch, and
Scott Brown 1985). The shed represents a very skillfully orga-
nized plan, and its exterior is decorated through the use of orna-
ment and appliqué, with such things as ‘‘stars of Texas” and
rather classical pilasters, a presumed commentary on Texas
culture.

The work of Charles Moore also employs figurative fragments—

particular roof forms, colonnades, alcoves, and niches—al-
though less as appliqué or shed decoration. For the most part
these elements are composed in a modern, albeit picturesque,
manner as an integral part of the overall work. Kresge College in
Santa Cruz exemplifies these characteristics (Moore 1967).

As Colquhoun points out, the strategy that seems to be most

involved in Venturi’s work and in Moore’s early work amounts to
“quotation” (Colquhoun 1978b, p. 35). Isolated figural elements
or motifs from outside the rest of the work itself are deliberately
introduced either to qualify or to offer additional commentary.
The interest of Venturi’s work derives from the intellectual foun-
dation of the choice of quotation and from the skill with which
the quotation is integrated into the architectural work. The prob-
lem of ‘‘misreading”’ is minimized by the currency and vernacu-
lar quality of the decoration or quotation. The Lone Star cannot
be completely misconstrued. Still, the inevitable parochialism of
today’s commonplaces would seem to limit the inherent power
of the strategy of quotation to unearth more universal meanings.
Can it be said of architecture, as of literature, that certain motifs,
figures, or tropes are handed down across the ages? If so, then
these would seem to represent the treasure trove for quotation.
The weakness of the vernacular is its more momentary meaning.

The recent scheme by Cesar Pelli for Robert R. Herring Hall,

housing the Jesse H. Jones Graduate School of Administration
(1982-1984), makes use of quotation, although the reach of the
references is far more modest and indeed is confined to the ar-
chitecture of the Rice University campus (figure 57). Here Pelli’s
renowned ability for systematic articulation and surface render-
ing of buildings (Pastier 1980) takes its cue from the intricate
patterns of masonry work and inlaid decoration that abound on
nearby buildings. As one reviewer remarked, however, the proj-
ect ‘‘reinterprets that tradition and reveals richer aspects of
[Pelli’s] design vocabulary” (Papademetriou 1985, p. 86). This is
not a case of mere contextual conformance but one in which the
architect's already distinctive approach is enriched by quotation

Architectural Positions and Their Realms of Inquiry 181

 Oy
EEOiLM OCK

56
The Laguna Gloria Art Museum in Austin, by
Venturi, Rauch, and Scott Brown (1985): sketch
and model.

Architectural Positions and Their Realms of Inquiry 182

 57
Robert R. Herring Hall at Rice University in Houston,
by Cesar Pelli and Associates (1982-1984).

Architectural Positioris and Their Realms of Inquiry 183

 and a strong sense of what it is to make a building. The result at
once provides a sense of continuity with the campus’s earlier
architectural tradition and offers a new statement about the
manner in which the tradition can be celebrated. As Pelli has
stated in a recent article with regard to the legitimacy of an exist-
ing building context, ‘‘The hope is that by repeating their forms
... somehow the value of the past will rub off... . This cannot be
a lasting condition. ... For me the solid ground... is given by
the basic relationship between the art of our buildings and their
system of construction” (Papademetriou 1985, p. 86).

The spatial organization of Herring Hall recognizes another

building tradition at Rice, namely, the creation of formalized
courts appropriated for use by surrounding departments. As did
Stirling and Wilford in their earlier addition to M. D. Anderson
Hall (1979-1981), Pelli ‘“wraps”’ or rather “‘layers” the building
around an outdoor garden to much the same effect as that of the
courtyard of the student center located opposite (figure 57). A
clear formal and figural relationship is thus struck, across an
open field, between the two building complexes.

In addition to quotation, more full-fledged deployment of a fig-

ural tradition in architecture is to be found in the historicist styl-
ing and conformation of a number of contemporary works. Many
might be dismissed as simply exercises in eclecticism that do
little to advance work on the language and the meaningful uses
to which it can be put. They might be seen as mere associative
uses of history—rifled, often indiscriminately, for the purposes
of styling. The recent work of Bofill and the Taller de Arquitec-
tura, however, is not to be taken lightly. Particularly in their new
town and other urban projects in France—in Saint-Quentin-
en-Yvelines, Paris, and Montpelier—a thoroughly Beaux-Arts
interest in classicism and a strongly rational sense of spatial
organization are pervasive. The interesting turn, however, is in
the manner of construction with which these figurative elements
are produced, and the types of accommodations to which they
are quite grandiosely applied.

The type of construction is close to the ultimate in sophistication

and efficiency. It employs industrial precast concrete produced
in specially prepared forms for the cladding, as well as highly
mechanized on-site “‘tunnel system” techniques for the shell of
the building (Dixon 1981). Rather than the design process being
subordinated to technical capacities for the production of de-
sired forms, the sophistication of the casting techniques is such

Architectural Positions and Their Realms of Inquiry 184

 that the designer is released, so to speak, to consider other
matters.

In most of the projects, such as Le Viaduc and Les Arcades du

Lac, the type of accommodation is predominantly mass housing
for middle-income wage earners. What one finds are several in-
versions of expected doctrine, especially of the modernist kind.
First, the mass housing has the conformation of a palace. It is
monumental, grand, and compositionally complete in a classical
sense (figure 58). Second, rather than a tectonic aesthetic or-
thodoxy that strives to absorb or express a fabrication technol-
ogy, the very sophistication of the technology has removed it
from matters of expression, as epitomized by the explosion of
unexpected figural forms. The resulting architecture is indeed
rhetorical in both a technical and a sociopolitical sense.

Certain problems emerge with this approach, particularly when

attempts are made to integrate domestic architecture with other
special and, one might say, more monumental uses. The ex-
pected hierarchy among institutional uses, which is normally ex-
pressed through an ascending scale of grandeur, material
quality, and monumentality, becomes difficult to achieve. This
dilemma is evident in the proposals for a largely nonresidential
civic complex to be placed on the other side of the lake from the
Arcades du Lac and Le Viaduc projects (Taller de Arquitectura
1984). The sheer monumentality of the existing housing
threatens to overwhelm the newer project proposal and makes
the design exercise that much tougher to resolve.

Bricolage
According to Lévi-Strauss, “a bricoleur is adept at performing a
large number of diverse tasks. . . . His universe of instruments is
closed and the rules of his game are always to make do with
whatever is at hand, that is to say, with a set of tools and materi-
als which is always finite and is also heterogeneous, because
what it contains bears no relation to the current project, or in-
deed to any particular project, but is the contingent result of all
the occasions there have been to renew or enrich the stock’’
(Lévi-Strauss 1966, p. 16). Colin Rowe and Fred Koetter in Col-
lage City call for the recovery of the bricoleur’s craft as a more
meaningful alternative to what they see as the emasculating in-
fluences of positivist planning and urban design doctrine (Rowe
and Koetter 1978). We are presented with the idea of making use
of past forms, fragments, and edifices in our cities, not as mere
sympathetic contextualism whereby one designs after the fash-
ion of what is around and about, but as a strategy that provides

Architectural Positions and Their Realms of Inquiry 185

58
Le Viaduc and Les Arcades du Lac in Saint-
Quentin-en-Yvelines, by Ricardo Bofill and the
Taller de Arquitectura (detail view of Le Viaduc and
general plan of both complexes, 1974-1978).

Architectural Positions and Their Realms of Inquiry 186

 NNAANNAAN
SSNs
ANASANAN NAANANAN
NANSNNAN
NNANNANNNN
SSAA
ANNNNANNAN
SSS
SSS
SS KR
NANANNANANN

NNANANANSN
NANNNNANSS
SSNS
SSENS
ASNANANANANAN
NANANANAN
SSS
ANANANS REE SAARNANANSSAAANAAANSAN
ARAN
NANANAANNS
AANA
SSAA
NSANS
ANNAN
NAAN
NSAANAN
AAAS
NAAN
XANAAANANAN
AAAS
NAAN
ANNAN
AN
AANA
NANNANN
ANS
NANANASANANANANANN
IOS
BIOO
OGQ
oqog
AONAN
SOG
HOO
OO
ow
YOOX AAAS
AANA
ASS
AAN
SD NAANANANANANANANANANAN
DADDDDDDDDDDDDDDDDDS
DDS
DDD
ODDDDADDADDDDADDD
SXANSSRAAAASAN
NARA
DD
DDD
DDD
DDS
tXXOOO?RRHAANNSNASNAAANS
ASAIN
ANAS
NAAN
NANNAN
AANA ASIANA
NANANSRANNANISANS
ANANSI
NASAANNASN
NAR
WANA
SX
NANA NANA
NASNS
ANS
SANS
AAS
ANAS
SAS
NNNARNANANANNNANN
DD
SDD
SADDD
DDD
DDD
DS NNNANNNASAANANANNANN
DDD SNARK
SCN
RN
SANSA

both continuity with the past and access to novel future works.
For after all, if we admit the possibility of recombining, reusing,
and reconstruing the old parts that are laid at architecture’s
doorstep, as it were, some measure of continuity is assured and
the new use to which the old parts are put advances the possibil-
ity of new meanings.

Colquhoun, in analyzing the work of Michael Graves, depicts a

shift away from an earlier preoccupation with abstract formalism
toward the expressive use and reinterpretation of traditional
architectural elements: ‘“‘Columns developed capitals, doors be-
came qualified with architraves and pediments, and ornamen-
tation adorned surfaces” (Colquhoun 1978b, p. 25). Certainly
in the Portland Public Services Building of 1979-1982 a brico-
lage can be seen (figure 59). The exterior, organized according
to a classical tripartite division, borrows heavily from the Beaux-
Arts facades in surrounding old Portland. The small roof pavil-
ions, obscuring mechanical equipment and the like, appear quite
literally as a collection of public buildings from a bygone era.
The overall appearance is very much in keeping with American
city halls built during the 1920s and 1930s. In part this is due to
the classical compositional principles involved, but it is also due
in large measure to the reinterpretation that has been placed
upon borrowed motifs and decorative elements (Jencks 1980,
p. 133f.).

Architectural Positions and Their Realms of Inquiry 187

The Portland Public Services Building in Portland,
Oregon, by Michael Graves (1979-1982): model
and sketches.

ffi!
ees Hi{ °
(adage cvotloa spate)

Architectural Positions and Their Realms of Inquiry 188

 Of course, the bricoleur’s craft, if it is to be of service to society
as a real alternative in matters of city planning and architecture,
must ‘address [itself] to a collection of oddments left over from
human endeavors’ (Lévi-Strauss 1966, p. 17) and transcend both
imitation and Rube Goldbergian invention. The product of the
bricoleur’s labors must have contemporary meaning. This re-
quirement raises issues of the choice of fragments or parts, their
integration, and their articulation, whether the purpose is polem-
ical or not.
In addressing these issues, Hirsch’s four criteria for good inter-
pretation seem most appropriate (Hirsch 1967, p. 236f.). For the
bricoleur must interpret the situation of the moment, even if only
through the “‘oddments” he has at hand. The four criteria are
legitimacy, correspondence, generic appropriateness, and co-
herence. By legitimacy is meant the permissibility of the use of
the oddments within prevailing public norms. Here some latitude
may be extended. What stands as accepted convention might be
quite far-ranging. And unlike those attempting to make new
formal or figural extensions to an architectural language, the
bricoleur faces the likelihood of public understanding based on
past experience. As one Portland resident said of Graves’s pro-
jected building, ‘‘l look forward to a public building that offers
more than the blank facades and mirror walls of modern archi-
tecture. ... The ‘temple’ design shows sensitivity to the design
of neighboring buildings” (Jencks 1980, p. 138).

The criterion of correspondence requires that the bricoleur ac-

count for all the conditions inherent in the situation of applica-
tion and in the oddments themselves. The choice, integration,
and articulation of architectural elements must resolve all the
problems at hand and do so in a legitimate manner. The criterion
of generic appropriateness is most simply illustrated by a lin-
guistic example: You don’t use the standards of, say, scientific
discourse to measure or interpret casual conversation. In archi-
tecture, the bricoleur must be careful not to use forms that
are inappropriate to their situations. The exterior of worker
housing that looks like a palace might be such a transgression.
Alternatively, the interpretative framework that allows such
transgressions to be made must be apparent and consistently
applied. In the work of Bofill cited earlier, sociopolitical com-
mentary might well be the appropriate basis for polemical articu-
lation of grand neoclassical facades on mass housing.

The fourth criterion is perhaps the hardest to meet. Coherence

requires that the meaning be obvious and make sense: more

Architectural Positions and Their Realms of Inquiry 189

 sense in fact than an alternative arrangement of architectural
elements. As we have seen in earlier chapters dealing with such
matters as ‘‘satisficing decisions’ and the unbounded nature of
design problems, knowledge that a given proposal is the best is
well-nigh impossible to attain. Nevertheless, such a quality,
should it prevail, would ensure conformance with the other three
criteria.

The danger of an architecture of bricolage is that legitimacy,

correspondence, and generic appropriateness cannot be main-
tained in the light of contemporary situational conditions that
have little or no congruence with past experience. The modern
urban world is notable for the rise of new institutions and means
of transaction. To equate the conditions of modern transporta-
tion, for example, with the streets, roads, and byways of old,
even if in a consciously polemical vein, is to deny the relevance
and ignore the opportunities of these conditions.

The undeniable strength of an architecture of bricolage is the

sense of cultural continuity, which may serve as a stabilizing in-
fluence in the face of sharp disjunctions with tradition or ca-
price. This strength does not entirely lie in the preservation of
the status quo or ‘“‘remembrance of things past,’ however.
Bricolage offers the opportunity for new kinds of meaning by
direct reference to and use of old, familiar, and rediscovered ar-
chitectural elements, fragments, and edifices. The strategy de-
pends, in large measure, on architectural elements’ retaining
their meanings. Nothing would confound the strategy more than
if these meanings were to shift and change on a whim, or during
the passage of time. The elements would surely lose their iden-
tity as tools and materials in the bricoleur’s craft shop.

Type
The “type idea’”’ has exercised the minds of many in the Western
philosophical tradition. From Plato’s ‘“‘ideas’”’ to modern lin-
guistic conceptions of a type as a shared mental object, the rela-
tionship of the ‘‘one and the many” inherent in the type idea has
been a subject for profound contemplation and impassioned dis-
pute for centuries (Hirsch 1967, p. 265f.). A central issue con-
cerns the ability of a type, which is one thing, to stand for or
represent more than one thing. Our trees and their trees are all
“tree.”

In architecture today, type often stands at the center of debate. It

is widely recognized as fundamental to the formulation of signifi-
cant and legible work, both for its ‘““one and the many” proper-
ties and, by extension, for the sense of continuity of tradition

Architectural Positions and Their Realms of Inquiry 190

 that is offered. According to the definition promulgated by the
eighteenth-century theorist Quatremére de Quincy, ‘‘the art of
regular building is born of a preexisting source,” and that preex-
isting source is the idea of type. ‘‘The word type presents less
the image of a thing to copy or imitate completely than the idea
of an element which ought itself to serve as a rule for the model.
Thus one should not say .. . that a statue or the composition of a
finished and rendered picture has served as a type for the copy
that one made. But when a fragment, a sketch, a thought of a
master, a more or less vague description has given birth to a
work of art in the imagination of an artist, one will say that the
type has been furnished for him by such-and-such an idea, motif
or intention” (Quatremére de Quincy 1977, p. 148).

In this discussion Quatremére de Quincy is careful to distinguish

between type and model, the latter being possessed of precisely
identifiable properties. He also, however, makes a distinction be-
tween “‘idea, motif, or intention” and type, in which type has the
quality of a definable rule that is provided by the idea, motif, or
intention. Consequently, while type definitely belongs to the
realm of ideas, it is not to be seen as something altogether
vague and fuzzy.

This kind of lofty abstraction is little different from definitions to

be found in disciplines other than architecture,° and does little
to influence architecture of its own accord. For that we might
turn to the group Known as the neorationalists and their pro-
gram of inquiry centered on the city, or what Vidler has referred
to as the ‘Third Typology” (Vidler 1978).

In seeking to overcome the difficulties of positivist interpreta-

tions of cities according to their functional elements (such as
land uses) and technical characteristics (such as infrastructure
systems), the neorationalists see the city as being most meaning-
fully composed of ‘‘fragments,”’ or architectural elements. These
elements, in turn, are not reinventions of past architectural types
and not just repetitions of past typological forms. According to
Vidler, such elements can only be defined, assembled, or reas-
sembled if they meet the following three criteria of meaning:
first, meaning ‘inherited from ascribed meanings of the past ex-
istence of the forms’; second, meaning ‘‘derived from the spe-
cific fragment and its boundaries and often crossing between
previous types’’; and third, meaning “‘proposed by a recomposi-
tion of those fragments in a new context” (Vidler 1978, p. 31).

With the first criterion Vidler echoes Rossi's idea of the ‘‘city
type” as ‘‘the collective memory of its people made up of objects

Architectural Positions and Their Realms of Inquiry 191

 and spaces, .. .a memory which in turn shapes the future, .. .
for when a group is introduced to a space they transform it into
their own image, but at the same time they yield and adapt to
material things [objects and spaces of the city] which resist this
transformation” (Rossi 1982, p. 130). It is a criterion that allows
quite heavy reliance to be placed on collective past meaning(s)
to ground and validate the work. As will be discussed later in
this chapter, the multiplicity of meanings with which a fragment
is imbued can sometimes confuse significance with meaning
and deny the very grounding that is being sought.

The work of the brothers Krier and others provides insight into
the second criterion by suggesting how the fragments should be
defined (Krier 1979, AAM 1978). For example, there is the ‘‘con-
ception of urban space’’—streets and squares—‘“‘as the primary
organizing element of an urban morphology,” from which it fol-
lows that buildings alone do not form a describable (meaningful)
space (figure 60). There is the dictum that housing zones must
be ‘“‘transformed into complex parts of the city, into cities within
the city, into ‘quartiers’ which integrate all the functions of urban
life.” According to this vision of the city, ‘“‘the street and square
are precise spatial types, the block is a result” (Krier 1978). Many
if not all these blocks are zones of mixed use in which the per-
ceived modern tendency toward separation of function is delib-
erately denied. The resulting idea of city form is reminiscent of
certain eighteenth-century accomplishments, although for differ-
ent reasons; as critics have argued, it does not follow that the
spatial conformities of one way of life are applicable to those of
another (Frampton 1980, p. 292). Neorationalist proposals are
viewed by some as having a faintly reactionary orientation, re-
moved from the sociopolitical locus of contemporary urban
conditions.

The third criterion for assembly, that meaning be imparted by

the “‘recomposition of . . . fragments in a new context,” recalls
the standard definition for another sort of human activity. The
genre of spoken language called the joke requires just such a
comprehension of new meaning through a shift in context. In the
hands of someone like Rossi, assembly in this manner takes on a
critical sociopolitical stance. According to Vidler’s interpreta-
tion, Rossi’s Regional Hall project for Trieste of 1974 (figure 61),
with its references to eighteenth-century prisons assigned to the
seat of local government, raises questions regarding the ambig-
uous condition of civic government (Vidler 1978, p. 32). The jux-
taposition of an open arcade suggests a lack of confinement and
is thus a contradiction of a prison. Other such metaphysical op-

Architectural Positions and Their Realms of Inquiry 192

 60
Three models (types) of urban space: 1. blocks re-
sulting from a street-and-square pattern, 2. streets
and squares as a result of positioning blocks, and
3. streets and squares as precise spatial types.

Architectural Positions and Their Realms of Inquiry 193

 S
.
. . :
y .
> . ‘ . )
are S
S
LIED XXX
N ‘ SN
Yj Ge GY “Wy Y
Y SRY. SG
yyy y SN BESS
: =
S
:
LM MYj bb Y Mares Lae
tip ==
- e
on =

: L
| ;
Preis!
i
lila
[ ia
Boe
=
= a
eee =2
gaa _
posseeee re
goa
oe
csnoseeeeae
oe

ms ssos
The
ee
8EEB
| | SUGUGGSBGGRR

ae eee
61
Project for the Regional Hall in Trieste, by Aldo
Rossi (1974).

positions can be seen deployed in Rossi’s work (‘‘prisons and

schools,” for instance), whereby past fragments are used to
strive toward new meanings (Braghieri 1982).

@ Problems of Meaningful Interpretation

in this second realm of inquiry, where architecture is seen
primarily in relationship to itself and its constituent elements, a
number of general issues of interpretation arise. Several specific
examples have already been noted in passing, and the criteria
for ‘“‘good”’ interpretation raised in connection with the bricoleur
could well be applied to other strategies. Three issues in particu-
lar stand out, however—issues that in one way or another attend
architectural form making of the introspective and referential
kind under discussion. They are the question of privacy versus
accessibility in the employment of an architectural language; the
confusion between “significance” and “‘meaning’’; and the prob-
lem of adequately integrating new formal constructs, as well as
expressing new social institutions.

The Privacy of Language

With regard to written text material, the evolving field of herme-
neutics—roughly speaking, the science of interpretation—has
dwelt upon a paradoxical situation that confronts a reader in
obtaining knowledge from the material (Gadamer 1982, p. 113f.;

Architectural Positions and Their Realms of Inquiry 194

 Bernstein 1983, pp. 30-48). This paradox is sometimes referred
to as the “hermeneutic circle’ and can be described in the fol-
lowing manner (Hirsch 1967, p. 259; Rickman 1976, p. 246f.).
“The meaning of a text is a complex of submeanings or parts
which hang together.” Further, ‘‘it is not merely a mechanical
collocation but a relational unity in which the relations of the
parts to one another and to the whole constitute an essential
aspect of their character as parts . . . and [of] the character of
the whole”’ (Hirsch 1967, pp. 258-259). It follows ‘‘that we cannot
perceive the meaning of a part until we have grasped the mean-
ing of a whole” (Hirsch 1967, p. 259), and vice versa. A genuine
dialectic can be seen to occur between our idea of the whole
and our perception of the parts that constitute it. Once the
dialectic, or “circle,” has been entered into, neither the ‘‘parts-
side” nor the ‘‘whole-side’”’ is totally determined by the other,
and of course for meaningful interpretation to ensue, the circle
must not be a vicious one.

A great deal of scholarship has been developed around this

paradox and the way in which the circle might be broken so as
to arrive at a plane of real understanding—if it can be broken at
all. Some, like Gadamer, hold that the process of grappling with
the paradox is inherently open-ended (Gadamer 1975, p. 251f.).
For our immediate purposes, however, this description of the
paradox serves to establish an important dilemma posed by a
work, such as a piece of architecture, with regard to meaning.

In the dialectical relationship between parts and whole in a work

of architecture, the parts must be apprehended and understood
if one is to make sense of the whole, and the whole must be
apprehended in some manner that facilitates understanding of
the parts. Understanding, then, requires intelligibility on the part
of the overall organization and composition of the work and on
the part of its surfaces, framing devices, material, and decora-
tion. In the absence of a truly expressive capacity, the architec-
tural object will remain mute and thus have no meaning.
As described earlier, one of the problems of an artificially con-
trived abstract formalism is that it can stretch the public expres-
sive capacity of the architectural work toward a limit at which
meaning is difficult to sustain. By the same token, rampant ad
hoc use of figural devices, like adjectives and nouns composed
on a page in some private way, can also fail to make public
sense. This is not to say that either abstract formalism or figural
devices have no place in architectural expression. On the con-
trary, to deny their place is to deny architecture. For example,

Architectural Positions and Their Realms of Inquiry 195

 geometries of various kinds are certainly abstract and funda-
mentally concerned with the ordering of spatial objects, includ-
ing architecture. They should and must be explored for their
formal expressive possibilities. But like the idea of types men-
tioned earlier, geometries do have rules, and because of this
they can be held up to a more public scrutiny and accountability.

The real dilemma is where the boundary line might be drawn

between the public realm and a more private world—assuming,
as this discussion does, that architecture is a broadly based cul-
tural enterprise and hence that the accessibility of a work is im-
portant. Several criteria, on the order of ‘“‘balancing tests,’’ were
proposed in the last chapter in connection with the issue of the
“engagement” of a work. One was that architecture should not
be considered arbitrary if it provides the means for obtaining
understanding. Nevertheless, the difference between a mere pro-
vision of means and real understanding can be considerable; the
work must gain meaningful substantiation in the ‘‘real world”’
outside itself if it is not to be “relegated to virtual worlds created
in the image of private languages and meanings” (Colquhoun
1978b, p. 37).

Significance and Meaning

A popular contention today is that architecture is autonomous,
effectively leading a “‘life of its own.’ More precisely, what is
proposed is that architecture, like other cultural works such as
literature, can have a meaning independent of authorial inten-
tion and that such meaning can change during the course of
time. This raises the question whether meaning can be sustained
in the absence of permanent norms on which to base judgment.
In other words, if we can have as many different meanings as
spectators, then are we not likely to have no lasting meaning at
all?

The obvious counter to this is that at any given moment there is

some unanimity about meaning on the part of spectators (Hirsch
1967, p. 212f.). They are all part of the same culture, read the
same books, breathe the same air, and so on. As Hirsch points
out, however, even if we accept this as so, it still leaves unan-
swered whether the meaning of the work has changed or
whether the meaning of collective interpretations has changed
(Hirsch 1967, p. 213). In fact it seems quite possible that any
Original meaning that the work may have had must be under-
stood in order for it to have any further meanings. As Hirsch
argues, it is clear that “‘significance’’—that is, present rele-
vance—cannot readily be fused with meaning. Furthermore, an

Architectural Positions and Their Realms of Inquiry 196

 autonomous, ‘living’ work does not really free the specator
from the “shackles of historicism,” it only destroys the basis for
agreement and objective study (Hirsch 1967, p. 214). On the con-
trary, to paraphrase Husserl, it would seem that the meaning of a
work of architecture must be an “intentional object,’ corre-
sponding to the designer’s intentions, which remains unchang-
ing through the course of time (Husserl 1913, p. 96f.).” In short,
the meaning of an architectural work is that aspect of the de-
signer’s intention which, under architectural conventions, may
be shared by others.
Of course this kind of declarative statement does not settle the
matter; what is actually shared by others may be unintentional.
But it does mean that architecture is not a purely public object,
whose character is solely determined by public norms. This
would imply the operation of some system of probabilities in the
adjudication of meaning; or, as Cassirer put it, meaning arises
from the “reciprocal determination” of ‘public [conventional]
possibilities and subjective specification of those possibilities”
(Cassirer 1975, p. 178; Hirsch 1967, p. 225). Returning to the in-
terpretative paradox, at least some semblance of architectural
convention must be present in order to constrain or otherwise
direct the knowing gaze of the spectator. As we have seen, many
form-making strategies, such as the use of formal and figural
devices, bricolage, quotation, and type, axiomatically require a
substantial measure of fixity of meaning; otherwise, the rhetoric
and cultural continuity to which designers aspire would be to-
tally ineffective. The presence of this fixity of meaning also sug-
gests that mistakes, misquotations, and misappropriations can
be made, and that the history and the original circumstances of
the architectural device, fragment, or oddment in question are
consequential to its contemporary use.

New Constructs and Institutions

As societies move through time, discoveries are made, new
functions and sociopolitical arrangements are instituted along-
side the old, and the complexion of how natural resources are
harnessed undergoes change. The results are frequently laid at
the doorstep of man’s cultural enterprise. New institutions
emerge, requiring expression through architecture; some se-
verely tax or transcend traditional practice. For example, the
modern automobile city presents institutional arrangements and
movement rituals that are materially if not radically different
from those of cities of earlier times. The problem of interest to us
here is how to provide appropriate expression to these entities,

Architectural Positions and Their Realms of Inquiry 197

 arrangements, and rituals. This, as modern functionalism was to
discover, involves more than merely housing the new functions
and providing basic systems of movement.

The enterprise of cultural interpretation and integration is an

essential part of a living tradition. We have seen many failures,
either in the guise of the emasculated, mechanistic, and frag-
mented forms of functionalism or in anachronistic and whimsi-
cal uses of architectural elements and devices of other epochs.
The problem seems to call for concerted development and ex-
tension into new dimensions and new forms of architectural ex-
pression. So might we regard the heroic efforts of the modern
movement. On the other hand, the integrating aspect of tradition
seems to require the establishment of continuity with the past;
and in any case, since cities are not built in a day, newcomers
have a responsibility to make a nod toward old-timers. It is not at
all clear that the existing constituent elements of architecture
can be regarded, after the pattern of language, as a more or less
complete set of words and grammatical rules with which texts
may be written and novel thoughts expressed. In any event, natu-
ral languages change through time and become themselves the
objects of changing needs and institutional rearrangements.

Architecture is also susceptible to changes in the very material

and dimensional aspects that fundamentally determine its con-
stitution. Advances or discoveries in material and fabrication
technologies can have import. Similarly, as suggested earlier,
geometry itself by now offers many different systems for organiz-
ing space. In certain brands, the system immediately moves
beyond three-dimensional Euclidean space into the realm of
“hyperspaces.”’ The representational schemes that go with these
geometries also affect the view of space and therefore its
conformation.

The directions in which such discoveries and inventions might

influence architecture are difficult to predict and even some-
times difficult to discern after the fact. In some cases, such as
that of new geometries, they seem to represent opportunities
and promising lines of investigation for work on the language.
Nevertheless, exploration of this kind for its own sake, although
it may eventually suggest some important contributions, at the
moment is not necessarily architecture. Furthermore, it seems
unlikely that the new contributions can be appreciated without
reference to tried and true procedures, again raising the issue
of the necessarily integrative aspect of tradition and its
advancement.

Architectural Positions and Their Realms of Inquiry 198

ee BRON)
A Convergence of Issues

To summarize, we have two realms of inquiry. One is concerned

with the making of architecture in the light of an external vision
of man and his world. The other is concerned with architecture
in itself.

At the risk of caricature, interpretation in the first realm can be

described as tending to adhere to the hypothetical-deductive
system of theory construction and empirical observation favored
by the social sciences. As pointed out earlier in the chapter, the
emerging idea of theory is one that makes categorical distinc-
tions between facts and values. Indeed, some purists would ar-
gue that value aspects of propositions about man and his world
can be neither true nor false and therefore cannot be included in
the realm of theory at all. Others might treat the same proposi-
tions about values as disguised factual claims and thus misplace
them within factual rather than normative discourse. Below the
loftier heights of such theoretical speculation, prevalent logical-
empirical orthodoxies have tended to bequeath a reductionist
and naive functionalism to the design and physical planning pro-
fessions. As Bernstein persuasively argues, it is also a theoret-
ical perspective that in more extreme cases fosters an attitude of
disinterested observer, with a concomitant faith in ‘‘value-free”’
objectivity as the inevitable avenue for social progress (Bern-
stein 1976). What has resulted is a disguised form of ideology in
which a dependent relationship is often unwittingly struck be-
tween methodological exigencies and limitations and the formu-
lation of courses of action: ends become defined by means
instead of by some other standard. The horizon of policy formu-
lation, planning, and design is determined by a ‘‘scientific view”’
of man and his world and not by some other vision. When such a
view is inappropriate or ignores many dimensions of the situa-
tion at hand, the consequences can be dire indeed.

That a certain disenchantment would occur within the areas of

planning, urban design, and architecture was probably inevita-
ble. There has been a growing suspicion of the kind of “‘liberal
faith” implicit in the role of urban theorists and practitioners as
aloof observers and dispensers of information and technical
know-how. ‘‘Value-free’”’ research and application have become
open to criticism as masked ideology that lends false legitimacy
to technical control (Sch6n 1983, ch. 1). On a more practical
level, a sense of professional crisis has developed. At first this
appeared to be just a problem of technique, quality of data, ana-

Architectural Positions and Their Realms of Inquiry 199

 lytical methods, and so on. Surely with a little more persever-
ance and added resources such problems could be overcome.
Repeated failures of planning and design applications to live up
to expectations, however, pronounced a requiem, first for large-
scale models in planning (Lee 1973, Harris 1967), then for quan-
titative techniques to be found more squarely within design
disciplines, and finally for the naive functional form making of
architects.

Not surprisingly, perceptions of these kinds of difficulties have

had mixed professional and institutional responses. Technical
shifts and more radical reorientations toward the possibility of
other interpretative frameworks have occurred. For example, the
inherent pluralism of irreducible concepts about tradition, cul-
ture, and ‘‘forms of life’’ has been expounded in so-called ethno-
methodologies and especially in hermeneutic critiques of the
modern era (Winters and Peckham 1984). These themes, how-
ever, are beyond our present scope.
Within the second realm of inquiry there is a tendency to adhere
to the rhetorical domain of architectural objects and organizing
compositional principles. Thus the distinction between architec-
ture and other disciplines is sharply and narrowly drawn. In more
extreme cases the kind of theory that is advanced fosters an in-
trospective attitude that is disengaged from technology and ar-
chitecture’s social purpose. Here ‘‘art for art's sake’ can become
an operative maxim that threatens to relegate architecture to the
rarefied fringes of consumer culture.

Means-ends and fact-value dichotomies, so evident in theorizing

of the logical-empirical variety, are also to be found among
theoretical positions of a recognizably normative kind. Many
normative speculations about architecture fail to find sufficient
grounding outside the immediate realm of architectural catego-
ries and the private languages of design. The result is that ques-
tions of ends, surely the heart of normative discourse, remain
implicit and vague. Consequently, these normative positions are
actually cast only in the form of means. Such discourse, if it re-
mains centered solely on prescriptive devices and architectural
categories—on questions of “‘right’’ rather than on questions of
‘‘good’’—seems certain to fail to generate sustaining normative
doctrines. It will also suffer many of the same problems as do
theoretical perspectives that confuse ends and means under the
influence of more ‘‘scientific’”’ kinds of interpretation. Even if the
confusion is not so blatant, the theoretical discourse may suffer
from a certain self-referential quality that might well confound

Architectural Positions and Their Realms of Inquiry 200

 efforts to provide adequate substantiation for the positions in-
volved. If one position seems to be more or less as good as an-
other and choosing between them is merely a matter of taste,
why bother much with fundamental considerations of appropri-
ateness, comprehensiveness, or validity?

Disenchantment with the expressive possibility of abstract mod-

ern formalisms has led to attempts to recover a figural tradition
in architecture. These attempts involve a number of strategies,
including quotation, bricolage, and certain uses of type, all of
which turn on the resurrection and reintroduction of architec-
tural elements from past eras. Meanwhile, other formal exercises
continue, in a seemingly endless quest for different ordering sys-
tems and means of manipulating spatial geometries.

Grounding the meaning of an architectural work has become

very much an issue, as a consequence of the dismantling of the
confidence placed in the doctrine of ‘‘form follows function.”
The idea of architecture as a language, capable of a broad reper-
toire of referential and rhetorical statements, has taken on a new
vitality, and acknowledgment of the importance of preserving
cultural continuity with the past has further reinforced an al-
ready introspective and backward-looking theoretical stance.
But the spectre of cities cluttered with self-referential ironic
statements, or simply architecture for architecture’s sake, looms
large in this orientation. Neither of these would seem to advance
our Cultural enterprise.

Again not surprisingly, we are looking at what might more rea-

sonably be regarded as two sides of the same coin. For architec-
ture, with its unique pedigree of art and science, plays between
the two realms. Furthermore, if we set aside the procedural stric-
tures of both realms in favor of their enabling prejudices, a com-
mon problem may be seen. It is the problem of recovering the
social purpose of architecture beyond the often insightful but
emasculated and reductive constructs from our logical-empirical
interpretation of man and his world. It is also the problem of
making the ennobling aspect and substance of architecture
more accessible and a part of society. The two facets of the
problem are clearly intertwined. Without social purpose it is dif-
ficult to imagine how a broad understanding of and meaning for
architecture can be established. On the other hand, without en-
gagement it is difficult to see how architecture can be conveyed
in the cultural mainstream, even with a strong sense of purpose.

chitectural Positions and Their Realms of Inquiry 201

Notes
Rae ee are simply conditions whereby we experi-
Chapter 1 ence something—whereby what we en-
counter says something to us’ (Gadamer
1
1976, p. 9). As Hirsch observed, Vorurteil is
Procedure for the protocols took the follow- Gadamer’s version of the hermeneutic prin-
ing format. First, subjects were invited to ciple that was first clearly perceived by
participate with the promise, unless waived, Schleiermacher and then elaborated by
of having their identity and association with Dilthey and Husserl (Hirsch 1967, p. 258).
the experiments remain confidential. Sec- Vorurteil should be translated as “‘prejudg-
ond, interview sessions were conducted ment” in order to avoid the pejorative sense
with each subject from the beginning to the that “prejudice” takes on in English; | use
end of a particular design project. Except at the term enabling prejudice to convey this
the start and the end, these sessions were positive construction.
largely self-scheduled by the subjects to
coincide with points at which they felt com-
fortable discussing their work. The general SE ae care ae
aim of the protocol analyses was in the di-
Chapter 2
rection of ‘validity of measurement” rather
than “reliability.” Third, each subject was 1
asked to keep a “sketchbook,” or set of Bazjanac and Rittel cite eleven properties of
working notes and drawings, and to record wicked problems. In addition to the four de-
the sequence of design ideas of any other scribed in the text, there are the following:
jottings about the project. Fourth, each in- no exhaustive list of admissible operators
terview was tape-recorded in two parts. The for solving such problems; for every wicked
first part allowed the subject to describe problem there is more than one possible ex-
and explain the work to date. The second planation; every wicked problem is a symp-
part allowed questions to be asked for the tom of another higher-level problem; no
purpose of clarification. At no time was an wicked problem and no solution to one has
overtly critical attitude taken toward the a definitive test; each wicked problem is a
subjects’ work. Finally, the tape recordings “one-shot” operation; every wicked prob-
and drawings were analyzed for the purpose lem is unique; and the wicked-problem
of reconstructing the design protocols that solver has no right to be wrong (Rittel and
are presented here in summary form. Webber 1972: Bazjanac 1974, pp. 9-10).
The reasons why real-time protocol analysis Z
was not used include the limited availability David Hartley (1705-1757) is generally rec-
of resources and the interference that real- ognized as the founder of the school of psy-
time recording procedures would seem to chology known as associationism. In 1749
impose on the process itself. For example, it he published his Observations on Man.
is unusual for designers to describe aloud Other leading figures of the movement that
what they are doing as they do it. Also, the held sway for some two hundred years were
technical paraphernalia that would be re- John Gay, John Stuart Mill, and Alexander
quired for real-time recorded observa- Bain.
tion would be obtrusive, unless carefully
masked. The aims of the protocol analyses 3
were more modest and more in the direction Embodiment of this view is hardly an acci-
of capturing the broader contours of design dent, since associationism grew out of the
rather than specific relationships such as philosophy of John Locke (1632-1704), es-
time of response to the problem encoun- pecially as expressed in his Essay Concern-
tered, levels of anxiety, and the like. ing Human Understanding (1690).

2 4
Prejudices in Gadamer’s sense of the word Wundt (1832-1920), often referred to as the
(Gadamer 1975). ‘Prejudices [Vorurteile] are founder of modern experimental psychol-
biases of our openness to the world. They ogy, was the pioneer of introspective tech-

Notes to pages 2-42 203

niques whereby subjects’ sense data could 12
be indirectly observed. Let us assume a decision-making process
with the following general form:
5
The mental state was referred to as Bewusst- Ri— st; [S — si] > R{T4 etc.,
seinslage, or “‘imageless thought” (Hum-
Ri = the ith rule chosen at time t;
phrey 1963, ch. 3).
si = observed characteristics of the pro-
6
posed solution from application of the
This corresponds to Wallas’s four steps
ith rule at time t;
(Wallas 1926), which were extended by re-
searchers such as Rossman (Rossman 1931) S = characteristics, or properties, of the
to (1) observation of a need, (2) analysis of required solution as known to the de-
the need, (3) survey of available information, signer; and
(4) formulation of objective solutions, (5)
critical analysis of solutions, (6) birth of a [] = the difference between the proposed
new idea, and (7) experimentation and test- and required solution properties at
ing. time t (episode f), as it appears to the
designer.
7
Perkins quarrels with this idea as an expla- “In the case of a well-defined problem (that
nation of creativity on the basis that biso- is, S are unambiguous) a particular rule
ciation is not a distinctive psychological Riis applied and a proposed solution s} is
process. Hence it does not clarify creativity obtained. Any resulting discrepancies
but rather begs the question (Perkins 1981, (through the comparison [S — s/]) are used
pp. 91-99). in order to select another rule R/74 during
the next episode, or to subsequently modify
8 the rule under consideration.
The term “chunk” was introduced by Miller
For the case of an extremely ill-defined
(1956, p. 93) to denote a grouped set of in-
problem, however, where S are largely un-
put data in a familiar unit—that is, a familiar
known, such an evaluation between S and s}
unit of information. Simon (1979, pp. 50-61)
cannot be sustained with any confidence.
presents an extended discussion of chunks,
Here Ri— si— S by definition. Hence [S —
concluding among other things that the size
si] = 0. In other words, no discrepancies are
of chunks may vary depending upon context
likely to be found and s} may or may not be
(word chunks, chunks of poetry, etc.).
a solution to the problem at hand. Further-
9 more, problem solving can only continue by
The term derives from the Greek heuriskein, simply starting again.
“to discover.”
If we admit a certain “selective inattention”’
10 in the constancy of appreciation with which
In other words, it is a view that runs against rule Ri is applied to yield proposed solution
the Cartesian concept of objective founda- si, then the switching model might take the
tional knowledge that stands, as it were, by following form:
itself.
Ris stand—>S
11 “/Si=3s4)20
Utzon’s own commentary indicates an addi- but Ri R/z1,..., etc.
tional purpose for the superstructure of sail-
In short, even in ill-defined problem-solving
like buildings: “In the Sydney Opera House,
situations, rule Ri, upon reflection by the de-
the idea has been to let the platform cut
signer, is likely to be considered part of a
through like a knife and separate primary
constellation of rules that are linked or re-
and secondary functions completely” (Ut-
lated in some manner (=A}), and problem
zon 1963, p. 117).
solving could continue with the use of one
of these related rules.

Notes to pages 42-104 204

During the subsequent application of rules what makes something a correct step in
in this manner, even under conditions of composing a variation . . . is that it is con-
trial-and-error, the required solution proper- nected pertinently with what precedes it;
ties (S) will become clearer. Thus something what makes a connection pertinent is
approximating the first kind of rule applica- settled by the reaction of practitioners”
tion can take place. In fact, for the solution (Mounce 1981, pp. 116-118).
of wicked problems, where there is no de-
finitive stopping rule (that is, S is poorly de- 6
“Vernacular” in the sense of a derivative of
fined), only this type of sequence seems
plausible. a more formal language. To a certain extent
the term “popular” might seem synonymous
with vernacular. For instance, ‘‘popular ar-
GRR See ann Se Si chitecture”’ is a term often used to refer to
Chapter 3 indigenous local forms of building, to local
materials and production systems. Yet ver-
1
nacular expressions are variations of “high
“The office building is a house of work of
style,” while popular architecture is con-
organization, of clarity, of economy... the
cerned with the development of forms out-
maximum effect with the minimum expendi-
side high style.
ture of means” (Mies van der Rohe 1923, p.
89). 7
For example, Stern’s assertion that contem-
2
porary architecture “represents a complex
First enunciated succinctly by Louis Sulli- interaction between the often conflicting is-
van: “Form is everything and anything. ...
sues that characterize the modern world.”
According to their nature ... some forms “It is difficult,’ he continues, “to sustain a
are definite and some are nebulous... the belief in Modernism’s struggle to reestablish
form exists because of the function” (Sulli- a monolithic style, because such a style
van 1934, pp. 40-41).
would possess an intolerably inadequate
3 palette with which to address the complex-
These terms are borrowed from Paul Frankl ities of modern life’ (Stern 1980, p. 39).
(Raumzellen as “space cells’ and Korper- 8
formen as ‘‘mass forms’’) and represent a In making this kind of point Eisenman noted
conceptual scheme for the analysis of ar- that the development of the formal language
chitectural compositions (Frankl 1914). of architecture ‘fundamentally changed the
4 relationship between man and object away
The Existenzminimum of lodging was a from an object whose primary purpose was
preoccupation of “Der Ring” architects, the to speak about man to one which was con-
Bauhaus, and later the CIAM. For instance, cerned with its own objecthood”’ (Eisenman
the CIAM II in Frankfurt (1929) was devoted 1978, p. 22).
to ‘“‘minimum-living-standard housing.” Le 9
Corbusier developed the notion of the “‘bio- [We] can... represent a corpus of knowl-
logical unit” of housing, corresponding to edge... by a deductively closed set of sen-
a standard of 14 square meters per occu- tences ... containing a set [to be called the
pani. And at the same CIAM congress of ‘urcorpus’]. The [urcorpus] consists of all
1929 he was to repudiate the idea of the logical truths, etc. ... [in the corpus of
“minimum house’”’ as being possible with knowledge] .. . as well as any other items
traditional techniques (Le Corbusier 1967, p. ... which might qualify as incorrigible in the
29). sense that they are immune from removal
5 from the standard of serious possibility”
After Wittgenstein, /nvestigations, in which, (Levi 1980, pp. 7, 12-13). The concept also
for instance, the matter of a musical theme
has much in common with Landau’s notion
and its variations is analyzed. “‘In short, of the ‘hard core”’ (after Lakatos) in the

Notes to pages 104-148 205

context of “architectural programmes,” ex- Third, there is the Law of Implication (Mo-
cept here the hard core seems to take ona dus Ponens): if p then q; p is true; therefore
more explicit presence in programme devel- q is true (Mitroff and Kilmann 1978).
opment (Landau 1982, pp. 303-309).
5
Bernstein groups together philosophers like
EA ES ARS Wittgenstein, Ryle, and Austin as providing
Chapter 4 the basis for this position; to them we might
add Russell and Moore.
1
The distinction of the two realms is by no 6
means novel. In certain respects it coincides For example, ‘‘| consider type to be a mental
with Simon’s definition of an artifact as the object or, if one prefers, an idea. The essen-
meeting point (‘interface’) between the “‘in- tial feature of a type idea is its ability to
ner’ environment, described according to subsume more than one experience and
“the substance and organization of the ar- therefore to represent more than one expe-
tifact itself,” and an ‘‘outer’ environment, rience” (Hirsch 1967, p. 265).
“the surroundings in which it operates”’ (Si-
7
mon 1969, p. 7). In quite a different vein, it
In very contracted form Husserl’s analysis
coincides with Vidler’s distinction between
proceeds as follows: (1) different intentional
his ‘second typology’’—‘‘architecture as
acts (for instance, acts by which | perceive
simply a matter of technique’’—and his
objects) can exist; (2) different intentional
“third typology,” in which architecture
objects (for instance, the object | am per-
“stresses its continuity of form and history”’
(Vidler 1978, pp. 23-32).
ceiving) can exist but each is unchanging;
(3) intentional objects may be reproduced
2 by different intentional acts and remain self-
“Generally coincides” in the sense that few identical (the same) through these repro-
proponents of the view are positivists in the ductions; (4) meaning is the sharable
strict sense of Ernst Mach and the Vienna content of an author’s intentional object;
Circle. They have nonetheless been pro- and (5) since this meaning is unchanging
foundly influenced by the positivist temper. and interpersonal, it may be reproduced by
For instance, a central role is assigned to the mental acts (appreciation) of other per-
empirically based explanatory theory very sons (Husserl 1913; Hirsch 1967, p. 217f.).
similar to Comte’s idea of a ‘‘positive sci-
ence’ based on precise observation, hy-
pothesis making, experimentation, and
regularities of natural cause and effect
(Bernstein 1976).

3
Or more precisely, Z2weckmassigkeit und die
neue Sachlichkeit—‘‘purposiveness/expedi-
ency/functionalism and the new objectivity/
practicality.”

4
Such a system of logic primarily involves
three principles. First, there is the Law of
Contradiction, under which no proposition
can be true and false at the same time (i.e.,
if p then q). Second, there is the Law of the
Excluded Middle, under which proposition
p and q is either true or false but not both.

Notes to pages 148-197 206

Bibliography
AAM, ed. (1978). Rational Architecture: The Allen, P. M. (et al.) (1981). Models of Urban
Reconstruction of the European City. Brux- Settlement and Structure as Dynamic Self-
elles: Editions des Archives d’Architecture Organizing Systems. Washington, D.C.: U.S.
Moderne. Department of Transportation, Systems
Akin, Omer (1978). ‘‘How Do Architects De- Analysis Division, February.
sign?” In G. Latombe, ed., Artificial Intelli- Anderson, Stanford (1984) “Architectural
gence and Recognition in Computer-Aided Design as a System of Research Pro-
Design, pp. 27-34. New York: North Holland. grammes.” Design Research.
Akin, Omer (1982 [1979]). Models of Archi- Archer, Bruce (1963-1964). “Systematic
tectural Knowledge: An Information Pro- Methods for Designers.” Design, 172,174,
cessing View of Architectural Design. Ann 176,179,181,185,188, April-August.
Arbor, Michigan: University Microfilms
Arieti, Silvano (1976). Creativity: A Magic
International.
Synthesis. New York: Basic Books.
Alberti, L. B. (1955 [1726]). Ten Books on
Asimow, M. (1962). Introduction to Design,
Architecture (De Re Aedificatoria). London:
Englewood Cliffs, New Jersey: Prentice-Hall.
A. Tiranti.
Attoe, Wayne (1978). Architecture and Criti-
Alexander, Christopher (1964). Notes on the
cal Imagination. New York: John Wiley and
Synthesis of Form. Cambridge, Massachu-
Sons.
setts: Harvard University Press.
Alexander, Christopher (1965). ‘‘A City Is Bachelard, Gaston (1979). The Poetics of
Not a Tree.” Architectural Forum, April. Space. Boston: Beacon Press.

Alexander, Christopher (1981). The Linz Barker, Roger G. (1968). Ecological Psychol-
Cafe. New York: Oxford University Press. ogy: Concepts and Methods for Studying
the Environment of Human Behavior. Stan-
Alexander, C., S. Ishikawa, and M. Silver- ford, California: Stanford University Press.
stein (1968). A Pattern Language Which
Generates Multi-Service Centers. Berkeley, Bartlett, Sir F. (1961). Remembering. Lon-
California: Center for Environmental don: Cambridge University Press.
Structure. Batty, M. (1971). “A Rational Approach to
Alexander, Christopher, et al. (1975). The Design.” Architectural Design, 41, pp. 436—
Oregon Experiment. New York: Oxford Uni- 439, 498-501.
versity Press. Bazjanac, Viadimir (1974). “Architectural
Alexander, Christopher, et al. (1977). A Pat- Design Theory: Models of the Design Pro-
tern Language. New York: Oxford University cess.” In William R. Spillers, ed., Basic
Press. Questions of Design Theory, pp. 8-16. New
York: North-Holland.
Alexander, Christopher (1979). A Timeless
Way of Building. New York: Oxford Univer- Berlin, Isaiah (1962). “Does Political Theory
sity Press. Still Exist?” In Peter Laslett and W. G. Run-
ciman, eds., Philosophy, Politics and Soci-
Alexander, C., and M. Manheim (1962). ety (second series). Oxford: Blackwell.
“HIDECS 2: A Computer Program for the Hi-
erarchical Decomposition of a Set within an Bernstein, Richard L. (1976). The Restruc-
Associated Graph.”’ Civil Engineering Sys- turing of Social and Political Theory. New
tems Laboratory Publication No. 160, MIT, York: Harcourt Brace Jovanovich.
Cambridge, Massachusetts. Bernstein, Richard L. (1983). Beyond Objec-
Allen, Gerald, and Richard Oliver, eds. tivism and Relativism: Science, Hermeneu-
(1981). Architectual Drawing: The Art and tics and Praxis. Philadelphia: University of
the Process. New York: Whitney Library of Pennsylvania Press.
Design. Blondel, Jacques-Francois (1771-1777).
Cours d’architecture, Ou traité de la décora-
tion, distribution et construction des bati- Cassirer, Ernest (1975 [1955]). The Philoso-
ments. Paris (9 vols.). phy of Symbolic Forms, Vol. 1: Language.
New Haven, Connecticut: Yale University
Bloomer, Kent C., and Charles W. Moore
Press.
(1977) Body, Memory and Architecture. New
Haven, Connecticut: Yale University Press. Catanese, Anthony J. (1972). Scientific
Methods of Urban Analysis. Urbana, Illinois:
Bloor, David (1983). Wittgenstein: A Social
University of Illinois Press.
Theory of Knowledge. New York: Columbia
University Press. Chadwick, George (1972). A Systems View
of Planning. New York: Pergamon Press.
Bonta, J. P. (1979). Architecture and Its In-
terpretation: A Study of Expressive Systems Chomsky, Noam (1965). Aspects of the
in Architecture. New York: Rizzoli. Theory of Syntax. Cambridge, Massachu-
setts: MIT Press.
Boring, E. (1950). A History of Experimental
Psychology. New York: Appleton-Crofts. Churchman, C. West (1967). “Wicked Prob-
lems.’’ Management Science, 4, no. 14, pp.
Braghieri, Gianni (1982). Aldo Rossi. Bar-
B-141, and B-142.
celona: Estudio.
Clawson, Marion, and Peter Hall (1973).
Broadbent, Geoffrey H. (1966). “Creativity.”
Planning and Urban Growth: An Anglo-
In S. A. Gregory, ed., The Design Method.
American Comparison. London: Johns Hop-
London: Butterworths.
kins University Press.
Broadbent, Geoffrey (1973). Design in Ar-
Collingwood, R. G. (1946). The /dea of His-
chitecture: Architecture and the Human Sci-
tory. New York: Oxford University Press.
ences. New York: John Wiley and Sons.
Collins, Lyndhurst, ed. (1976). The Use of
Broadbent, G., and A. Ward, eds. (1969). De-
Models in the Social Sciences. Boulder, Col-
sign Methods in Architecture. New York:
orado: West-View Press.
Wittenborn.
Colquhoun, Alan (1967). “Typology and De-
Bruner, Jerome S. (1961). ‘The Act of Dis-
sign Method.” Perspecta, 12, pp. 71-74.
covery.” Harvard Educational Review,
Winter. Colquhoun, Alan (1978). ‘Form and Fig-
ure.’ Oppositions, 12, Spring, pp. 29-37.
Bruner, Jerome S. (1967). “The Conditions
of Creative Thinking.” In H. Gruber, G. Colquhoun, Alan (1978b). “From Bricolage
Tewell, and M. Wertheimer, eds., Contempo- to Myth: Or How to Put Humpty-Dumpty To-
rary Approaches to Creative Thinking. New gether Again.” Oppositions, 12, Spring, pp.
York: Atherton Press. 1-18.
Bruner, Jerome S., Jacqueline J. Goodnow, Conrads, Ulrich (1964). Programs and Mani-
and George A. Austin (1967). A Study of festos on 20th-Century Architecture (trans-
Thinking. New York: John Wiley and Sons. lated by Michael Bullock). Cambridge,
Massachusetts: MIT Press.
Burton, Joseph (1983). ‘Notes from Volume
Zero.” Yale Architectural Journal, 20, pp. Consortium of East Coast Schools of Ar-
70-91. chitecture (1981). Architectural Education
Study, Vols. 1 and 2. New York: Andrew W.
Campbell, Donald T., and Julian C. Stanley Mellon Foundation.
(1970 [1963]). Experimental and Quasi-Ex-
Craik, Kenneth H., and Ervin H. Zube, eds.
perimental Designs for Research. Chicago:
(1976). Perceiving Environmental Quaiity.
Rand McNally.
New York: Plenum Press.
Carlian, Jean Paul (1979). ‘“‘The Ecole des
Crane, David A. (1964). “The Public Art of
Beaux-Arts: Modes and Manners.” Journal
City Building.” Annals of the American
of Architectural Education, 23, no. 2,
Academy of Political and Social Sciences,
November, pp. 7-17.
March.

Bibliography 210
Davis, Albert J., and Robert P. Schubert Eisenman, Peter (1979). ‘“Aspects of
(1976 [1974]). Alternative Natural Energy Modernism: Maison Domino and the Self-
Sources in Building Design. New York: Van Referential Sign.” Oppositions, 15/16, pp.
Nostrand Reinhold. 118-129.
De Leon, Augusto (1972). A Computer-
Fann, K. T. (1970). Peirce’s Theory of Ab-
Aided Markovian Model of the Design Pro-
duction. The Hague: Martinusnijhoff.
cess. Houston, Texas: Rice Architecture
Computer Laboratory. Ferguson, Francis (1975). Architecture,
Cities and the Systems Approach. New
Diamond, C., ed. (1976). Wittgenstein’s Lec-
York: George Braziller.
tures on the Foundations of Mathematics,
Cambridge 1939. Brighton: Harvester Press. Feyerabend, Paul (1978 [1975]). Against
Method: Outline of an Anarchistic Theory of
Dickey, John W., and Thomas M. Watts
Knowledge. London: Verso.
(1978). Analytical Techniques in Urban and
Regional Planning. New York: McGraw-Hill. Five Architects (1972). Five Architects:
Eisenman, Graves, Gwathmey, Hejduk and
Dietz, Albert G. H., and Lawrence S. Cutler
Meier. New York: Wittenborn Art Books,
(1971). Industrial Building Systems for
Inc.
Housing. Cambridge, Massachusetts: MIT
Press. Frampton, Kenneth (1974). “Apropos Ulm:
Curriculum and Critical Theory’. Opposi-
Dixon, John Morris (1981). ‘““Wage-Earner’s
tions, 3, pp. 17-36.
Versailles.’ Progressive Architecture, Octo-
ber, pp. 94-97. Frampton, Kenneth (1980). Modern Archi-
tecture: A Critical History. New York: Ox-
Domencich, Thomas A., and Daniel McFad-
ford University Press.
den (1975). Urban Travel Demand: A Behav-
joral Analysis. New York: North-Holland. Frampton, Kenneth (1982). Modern Archi-
tecture and the Critical Present. London:
Duhem, Pierre (1914). La théorie physique:
Architectural Design Press.
Son objet, sa structure. Paris.
Frampton, Kenneth (1983). “Towards a Criti-
Dunster, David (1977). ‘““A Comeback for Ar-
cal Regionalism: Six Points for an Architec-
chitectural Theory.” Progressive Architec-
ture of Resistance.” In Hal Foster, ed., The
ture, May, pp. 80-83.
Anti-Aesthetic: Essays on Postmodern Cul-
Durand, J. N. L. (1802). Précis de lecons ture, pp. 16-30. Port Townsend, Washing-
d’architecture. Paris. ton: Bay Press.
Frampton, Kenneth (1985). “Architecture
Eastman, Charles (1969). ‘“‘Towards a Theory
and the Tectonic Culture.” Lecture Series at
of Automated Design.” Proceedings, Inter-
the School of Architecture, Rice University,
disciplinary Conference on Decision-Aids,
Spring.
Ohio State University, May 16, p. 30.
Frankena, W. K. (1963). Ethics. Englewood
Eastman, C. M. (1975). Spatial Synthesis in
Cliffs, New Jersey: Prentice-Hall.
Computer-Aided Building Design. New
York: John Wiley and Sons. Frankl, P. (1914). Die Entwicklungsphasen
der neuen Baukunst. Leipzig.
Echenique, M. (1963). Models: A Discus-
sion. Working Paper No. 6, Land Use and Gadamer, Hans-Georg (1975). Truth and
Built Form Studies, University of Cam- Method. New York: Crossroad.
bridge, March.
Gadamer, Hans-Georg (1976). Philosophical
Egber, Donald Drew (1980). The Beaux-Arts Hermeneutics (translated by David E. Linge).
Tradition in French Architecture. Princeton, Berkeley, California: University of California
New Jersey: Princeton University Press.
Press.
Eisenman, Peter (1978). ‘“Postscript: Graves Gadamer, Hans-Georg (1982). Reason in the
of Modernism.” Oppositions, 12, Spring, p. Age of Science (translated by Frederick G.
22f.

Bibliography 211
Lawrence). Cambridge, Massachusetts: MIT Harrison, Andrew (1978). Making and Think-
Press. ing: A Study of Intelligent Activities. Sus-
sex: Harvester Press.
Gardner, Howard (1981). The Quest for
Mind: Piaget, Lévi-Strauss and the Struc- Hayes, J. R. (1978). Cognitive Psychology:
turalist Movement. Chicago: University of Thinking and Creating. Homewood, Illinois:
Chicago Press. Dorsey.
Ghiselin, Brewster, ed. (1952). The Creative Hayes, John R. (1981). The Complete Prob-
Process. Berkeley, California: University of lem Solver. Philadelphia: Franklin Institute
California Press. Press.
Gombrich, E. H. (1965). Art and I/lusion: A Heidegger, Martin (1962). Being and Time
Study in the Psychology of Pictorial Repre- (translated by John Macquarie and Edward
sentation. New York: Pantheon. Robinson). London: SCM Press.
Gordon, W. J. (1961). Synectics. New York: Hempel, Carl G., and Paul Oppenheim
Collier Books. (1948). ‘The Covering Law of Scientific Ex-
planation.’”’ Philosophy of Science, 15, no. 2,
Graves, Michael (1977). “The Necessity of
April, pp. 135-146, 172-174.
Drawing: Tangible Speculation.” Architec-
tural Design, 47, no. 6, pp. 384-394. Herdeg, Klaus (1983). The Decorated Dia-
gram. Cambridge, Massachusetts: MIT
Graves, Michael (1984). “‘Monumentality and
Press.
the City: Forum Transcript, December 12,
1981.” Harvard Architecture Review, 4, Hesse, Mary (1963). Models and Analogues
Spring, p. 40. in Science. Brighton: Harvester Press.
Gregory, S. A., ed. (1966). The Design Hillier, Bill, John Musgrove, and Pat O’Sulli-
Method. London: Butterworths. van (1972). “Knowledge and Design.” In
W. J. Mitchell, ed., Environmental Design:
Gutman, Robert (1977). “‘House VI.” Pro-
gressive Architecture, June, pp. 57-68.
Research and Practice, pp. 29-33. Los
Angeles: EDRA.
Habermas, Jurgen (1979). Communication Hirsch, E. D., Jr. (1967). Validity in Interpre-
and the Evolution of Society (translated by tation. New Haven, Connecticut: Yale Uni-
Thomas McCarthy). Boston: Beacon Press. versity Press.
Habraken, N. J. (1972). Supports: An Al- Holmes, Ann (1985). ‘““New Austin Museum
ternative to Mass Housing. New York: Shown as Work in Progress.”’ Houston
Praeger. Chronicle, April 23.
Halfpenny, William (facs. ed. 1968 [1724)). Hubbard, William (1981). Complicity and
Practical Architecture. London: Bowles. Conviction: Steps toward an Architecture of
Hall, Edward T. (1969). The Hidden Dimen- Convention. Cambridge, Massachusetts:
sion. Garden City, New York: Anchor MIT Press.
Books. Huet, Bernard (1978). ‘“Small Manifesto.” In
Handa, Rumiko (1983). “Abduction Theory AAM, ed., Rational Architecture: The Re-
and Its Application to Bruno Taut’s Interpre- construction of the European City, p. 54.
tation of Japanese Architecture.’ Unpub- Bruxelles: Editions des Archives d’Architec-
lished paper, Ph.D. Program in Architecture, ture Moderne.
University of Pennsylvania, December. Humphrey, G. (1963). Thinking: An In-
Handler, A. B. (1970). Systems Approach to troduction to Its Experimental Psychology.
Architecture. New York: Elsevier. New York: John Wiley and Sons.

Harris, Britton (1967). “The Limits of Sci- Hunt, Morton (1982). The Universe Within: A
ence and Humanism in Planning.” American New Science Explores the Human Mind.
Institute of Planners Journal, September, New York: Simon and Schuster.
pp. 324-335.

Bibliography 212
Husserl, Edmund (1913). Logische Unter- Kostof, Spiro, ed. (1977). The Architect:
suchungen. Halle: Max Niemeyer. Chapters in the History of the Profession.
New York: Oxford University Press.
Husserl, Edmund (1970). The Crisis of Euro-
pean Sciences and Transcendental Phe- Krier, Leon (1978). “The Reconstruction of
nomenology (translated by David Carr). the City.” In AAM, ed., Rational Architec-
Evanston, Illinois: Northwestern University ture: The Reconstruction of the European
Press. City, pp. 38-42. Bruxelles: Editions des Ar-
chives d’Architecture Moderne.
Husserl, Edmund (1975 [1962]). /deas: Gen-
eral Introduction to Pure Phenomenology Krier, Rob (1979). Urban Space (Stadtraum).
(translated by W. R. Boyce Gibson). New New York: Rizzoli.
York: Collier.
Krimerman, Leonard |. (1969). The Nature
and Scope of Social Science: A Critical An-
Ingram, Gregory K. (1979). “Simulation and
thology. New York: Appleton-Century-
Econometric Approaches to Modeling Ur-
Crofts.
ban Areas.” In Peter Mieskowski and
Mahlon Straszheim, eds., Current Issues in Krueckeberg, Donald A., ed. (1983). /n-
Urban Economics, pp. 130-164. Baltimore: troduction to Planning History in the United
John Hopkins University Press. States. New Brunswick, New Jersey: The
Center for Urban Policy Research.
Ivins, William M. (1973). On the Rationaliza-
tion of Sight. New York: Da Capo Press. Landau, Royston (1968). New Directions in
British Architecture. New York: George
Jencks, Charles, and George Baird, eds.
Braziller.
(1969). Meaning in Architecture. New York:
George Braziller. Landau, Royston (1982). ‘Methodology of
Research Programmes.” In B. Evans, T. A.
Jencks, Charles, ed. (1980). Post-Modern
Powell, and R. J. Talbot, eds., Changing De-
Classicism. London: Garden House Press.
sign, pp. 303-309. New York: John Wiley
Jones, J. C. (1970). Design Methods. New and Sons.
York: John Wiley and Sons.
Lane, Barbara Miller (1968). Architecture
Julian de la Fuente, Guillermo (1968). “The and Politics in Germany, 1918-1945. Cam-
Venice Hospital Project of Le Corbusier.” bridge, Massachusetts: Harvard University
Architecture at Rice, 23, April. Press.

Kahn, Louis I. (1961). ‘Order Is... .”’ Zodiac, Lang, Jon, Charles Burnette, Walter
8, June, p. 20. Moleski, and David Vachon (1974). Design-
ing for Human Behavior: Architecture and
Kaliski, John (1983). “Learning from the the Behavioral Sciences. Stroudsburg,
Park Regency.” Cite, Spring, pp. 8-13. Pennsyivania: Dowden, Hutchinson, and
Kaufmann, Emil (1968 [1955]). Architecture Ross.
in the Age of Humanism. New York: Dover. La Patra, J. W. (1973). Applying the Systems
Kockelmans, Joseph K., ed. (1967). Phenom- Approach to Urban Development. Strouds-
enology: The Philosophy of Edmund Husserl burg, Pennsylvania: Dowden, Hutchinson,
and Its Interpretation. Garden City, New and Ross.
York: Anchor Books. Le Corbusier (1951). The Modulor. London:
Koestler, A. (1964). The Act of Creation. Faber and Faber.
London: Hutchinson. Le Corbusier (1958). The Chapel at Ron-
Koffka, K. (1935). Principles of Gestalt Psy- champ. London: Architectural Press.
chology. New York: Harcourt, Brace. Le Corbusier (1959) Towards a New Ar-
Kohler, Wolfgang (1929). Gestalt Psychoi- chitecture. London: Architectual Press.
ogy. New York: Liveright. Le Corbusier (1965). Oeuvre complete:
1910-1929. Zurich: H. Girsberger.

Bibliography 213
Le Corbusier (1967). The Radiant City. New Millon, Henry A. (1983). Baroque and
York: Orion Press. Rococo Architecture. New York: George
Braziller.
Le Corbusier and Pierre Jeanneret (1926).
“Five Points towards a New Architecture.”’ Milne, Murray (1970). “CLUSTER: A Struc-
Almanach de I’Architecture Moderne, Paris. ture-Finding Algorithm.” In Gary T. Moore,
ed., Emerging Methods in Environmental
Lee, Colin (1973). Models in Planning. New
Design and Planning. Cambridge, Massa-
York: Pergamon Press.
chusetts: MIT Press.
Lee, Douglas B. (1973). “Requiem for Large-
Mishan, E. J. (1976). Cost-Benefit Analysis.
Scale Models.” American Institute of Plan-
New York: Praeger Special Studies.
ners Journal, May, pp. 163-178.
Mitchell, William J. (1970). “Switching On
Levi, Isaac (1980). The Enterprise of Knowl-
the Seven Lamps.” In Henry Sanoff and Sid-
edge: An Essay on Knowledge, Credal
ney Cohn, eds., EDRA-7. Chapel Hill, North
Probability and Chance. Cambridge, Massa-
Carolina: University of North Carolina.
chusetts: MIT Press.
Mitchell, William J. (1977). Computer-Aided
Lévi-Strauss, Claude (1966). The Savage
Architectural Design. New York: Petrocelli-
Mind. Chicago: University of Chicago Press.
Charter.
Lowry, Ira S. (1965). ‘“A Short Course in
Mitroff, lan |., and Ralph H. Kilmann (1978).
Model Design.” American Institute of Plan-
Methodological Approaches to Social Sci-
ners Journal, May, pp. 158-166.
ences. Washington, D.C.: Jossey-Bass.
Lynch, Kevin (1960). The Image of the City.
Moore, Charles (1967). ‘Portfolio: Architec-
Cambridge, Massachusetts: MIT Press.
ture.” Perspecta, 17, pp. 178-218.
Lynch, Kevin (1981). A Theory of Good City
Moore, Gary T., and Reginald G. Golledge
Form. Cambridge, Massachusetts: MIT
(1976). Environmental Knowing. Strouds-
Press.
burg, Pennsylvania: Dowden, Hutchinson
Maldonado, T. (1972). Design, Nature and and Ross.
Revolution: Toward a Critical Ecology. New Morris, Robert (facs. ed. 1971 [1734]). Lec-
York: Harper and Row. tures on Architecture Consisting of Rules
Maldonado, T., and G. Bonsiepe (1964). Founded upon Harmonic and Arithmetical
“Science and Design.” Ulm, 10/11, May. Properties in Buildings, etc. London:
Bowles.
Mallin, Samuel B. (1979). Merleau-Ponty’s
Philosophy. New Haven, Connecticut: Yale Mounce, H. O. (1981). Wittgenstein’s Trac-
University Press. tatus: An Introduction. Chicago: University
of Chicago Press.
Merleau-Ponty, Maurice (1962). Phenome-
nology of Perception (translated by Colin Negroponte, N. (1970). The Architecture Ma-
Smith). London: Routledge and Kegan Paul. chine. Cambridge, Massachusetts: MIT
Mesarovic, M. D. (1964). Views on General Press.
Systems Theory. New York: John Wiley and Negroponte, N. (1972). The Soft Architec-
Sons. ture Machine. Cambridge, Massachusetts:
Meyer, Hannes (1928). “‘Building.’’ Bauhaus, MIT Press.
Zeitschrift fur Gestaltung (Dessau), 2, no. 4, Negroponte, N., ed. (1974). Reflections on
p. 153f. Computer Aids to Design and Architecture.
Mies van der Rohe, Ludwig (1923). ‘‘Work- New York: Mason and Lipscomb.
ing Theses.” De Stijl, 6, p. 89. Newell, A., J. C. Shaw, and H. A. Simon
Miller, George A. (1956). “The Magical Num- (1957). “Elements of a Theory of Problem
ber Seven, Plus or Minus Two: Some Limits Solving.” Rand Corporation Report P-971,
on Our Capacity for Processing Informa- March.
tion.” Psychology Review, 63, pp. 81-97.

Bibliography 214
Newell, Allen, J. C. Shaw, and Herbert A. Perkins, P. N. (1981). The Mind’s Best Work.
Simon (1967). ‘The Process of Creative Cambridge, Massachusetis: Harvard Univer-
Thinking.” In H. Gruber, G. Terrell, and M. sity Press.
Wertheimer, eds., Contemporary Ap-
Perrault, Claude (1683). Ordonnance des
proaches to Creative Thinking, pp. 63-119.
cing espaces de colonnes selon la méthode
New York: Atherton Press.
des anciens. Paris: Jean Baptiste Coignard.
Newell, Alan, and Herbert A. Simon (1972).
Pevsner, Sir Nikolaus (1961). An Outline of
Human Problem Solving. Englewood Cliffs,
European Architecture. London: Pelican
New Jersey: Prentice-Hall.
Books.
Norberg-Schulz, Christian (1965). Intentions
Piaget, J. (1965). The Child’s Conception of
in Architecture. Cambridge, Massachusetts:
the World. Totowa, New Jersey: Littlefield
MIT Press.
Adams.
Norberg-Schulz, Christian (1968). ‘Less or Piaget, J. (1970a). Structuralism. New York:
More?” Architectural Review, April, pp.
Basic Books.
257-258.
Piaget, Jean (1970b). Genetic Epistemology.
Norberg-Schuiz, Christian (1980a). Genius New York: W. W. Norton.
Loci. New York: Rizzoli.
Pohlman, Richard W. (1982). ‘““A System for
Norberg-Schulz, C. (1980b). Architettura tar- Recording Behavior and Occupying De-
dabarocca. Milan: Electa. sign.” In Omer Akin and Eleanor F. Weinel,
eds., Representation in Architecture, pp.
Owen, Charles L. (1970). ““DCMPOS: An AI-
121-138. Silver Springs, Maryland: Informa-
gorithm for the Decomposition of Non-
tion Dynamics.
Directed Graphs.” In Gary T. Moore, ed.,
Emerging Methods in Environmental Design Polya, Gyorgy (1954). Mathematics and
and Planning. Cambridge, Massachusetts: Plausible Reasoning, 2 vols. Princeton, New
MIT Press. Jersey: Princeton University Press.
Polya, G. (1957). How to Solve It. London:
Panofsky, Erwin (1967 [1951]). Gothic Ar-
Doubleday.
chitecture and Scholasticism. Cleveland,
Ohio: Meridian Books. Popper, Karl R. (1959 [1934]). Logic of Sci-
entific Discovery. London: Hutchinson.
Papademetriou, Peter C. (1980). ‘Putting Up
a Good Front.” Progressive Architecture, Portoghesi, Paolo (1983). Postmodern: The
July, pp. 58-59. Architecture of tiie Post-Industrial Society.
New York: Rizzoli.
Papademetriou, Peter C. (1985). “Pattern
and Principle.” Progressive Architecture, Preiser, Wolfgang, ed. (1978). Facility Pro-
April, p. 86. gramming. Stroudsburg, Pennsylvania:
Dowden, Hutchinson and Ross.
Pastier, John (1980). Cesar Peli. New York:
Whitney Library of Design. Price, Cedric (1965). ““Non-Plan.” Architec-
tural Design, May.
Peirce, C. S. (1965). Collected Papers of
Charles Peirce (edited by C. Hartshorne and Price, Cedric (1966a). ‘‘Ptb Life Condition-
P. Wiess). Cambridge, Massachusetts: Har- ing.” Architectural Design, October.
vard University Press. Price, Cedric (1966b). ‘Potteries Thinkbelt.”’
Pena, William M. (1977). Problem Seeking: New Society, 2, June.
An Architectural Programming Primer. Bos- Proshansky, Harold M., William H. Ittelson,
ton: Cahner Books. and Leanne G. Riulin, eds. (1970). Environ-
Perez-Gomez, Alberto (1983). Architecture mental Psychology: Man and His Physical
and the Crisis of Modern Science. Cam- Setting. New York: Holt, Rinehart and
bridge, Massachusetts: MIT Press. Winston.

Bibliography 215
Quatremére de Quincy (1977 [1825]). “Type”’ Sanoff, Henry (1977). Methods of Architec-
(translated from an essay in the Encyclo- tural Programming. Stroudsburg, Pennsyl-
pédie méthodique, Paris). Oppositions, 9, p. vania: Dowden, Hutchinson and Ross.
148f. Schnaidt, Claude (1965). Hannes Meyer:
Building Projects and Writings. New York:
Rapaport, Amos (1969). House, Form and
Architectural Book Publishing.
Culture. Englewood Cliffs, New Jersey:
Prentice-Hall. Schon, Donald A. (1983). The Reflective
Practitioner: How Professionals Think in Ac-
Rapaport, Amos (1977). Human Aspects of
tion. New York: Basic Books.
Urban Form: Towards a Man-Environment
Approach to Urban Form and Design. New Schon, Donald A. (1984). “Problems,
York: Pergamon Press. Frames and Perspectives on Design.” De-
sign Research.
Rawls, John (1971). A Theory of Justice.
Cambridge, Massachusetts: Harvard Univer- Scott, Mel (1969). American City Planning
sity Press. since 1890. Berkeley, California: University
of California Press.
Rickman, H., ed. (1976). Dilthey: Selected
Writings. New York: Cambridge University Scott Brown, Denise (1977). ‘““Suburban
Press. Space, Scale and Symbols.” Via, 3, pp. 41-
47.
Rittel, Horst W. J. (1972). “On the Planning
Crisis: Systems Analysis of the First and Scully, V., Jr. (1979). Modern Architecture.
Second Generations.” Bedrifsakonomen, 8, New York: George Braziller.
pp. 390-396. Silverstein, Murray, and Max Jacobson
Rittel, Horst W. J., and Melvin M. Webber (1978). “Restructuring the Hidden Program:
(1972). Dilemmas in a General Theory of Toward an Architecture of Social Change.”
Planning. Working Paper No. 194, Institute In Wolfgang F. E. Preiser, ed., Facility Pro-
of Urban and Regional Development, Uni- gramming. Stroudsburg, Pennsylvania:
versity of California, Berkelely. Dowden, Hutchinson and Ross.
Rossi, Aldo (1982). The Architecture of the Simon, H. A. (1957). Models of Man. New
City (translated by Dianne Ghirardo and York: John Wiley and Sons.
Joan Ockman). Cambridge, Massachusetts: Simon, Herbert A. (1969). The Sciences of
MIT Press. the Artificial. Cambridge, Massachusetts:
Rossman, J. (1931). The Psychology of the MIT Press.
Inventor. Washington, D.C.: Inventors Simon, Herbert A. (1970). ‘Style in Design.”
Publishing. In John Archae and Charles Eastman, eds.,
Rowe, Colin, and Fred Koetter (1978). Col- EDRA-Two: Proceedings of the Second An-
lage City. Cambridge, Massachusetts: MIT nual Environmental Design Association
Press. Conference, Pittsburgh, Pennsylvania, pp.
1-10.
Rowe, Peter G. (1972). “A Question of Ar-
chitecture, a Matter of Style.” Architectural Simon, Herbert A. (1973a). ‘Does Scientific
Design, August, p. 466f. Discovery Have a Logic?” Philosophy of
Science, December, pp. 471—480.
Rowe, Peter G. (1982). ‘“A Priori Knowledge
and Heuristic Reasoning in Architectural Simon, Herbert A. (1973b). ‘Structure of III-
Design.” Journal of Architectural Educa- Structured Problems.” Artificial Intelli-.
tion, 36, no. 1, pp. 18-23. gence, 4, pp. 181-201.
Ryle, Gilbert (1949). “On Knowing How and Simon, H. A. (1979). Models of Thought.
Knowing That.” In The Concept of Mind. New Haven, Connecticut: Yale University
London: Hutcheson. Press.
Skinner, B. F. (1953). Science and Human
Behavior. New York: Free Press.

Bibliography 216
Smithson, Alison, ed. (1968). Team 10 Venturi, Robert (1966). Complexity and Con-
Primer. London: Studio Vista. tradiction in Architecture. New York:
Sommer, Robert (1969). Personal Space: Museum of Modern Art.
The Behavioral Basis for Design. Englewood Venturi, Rauch and Scott Brown (1985).
Cliffs, New Jersey: Prentice-Hall. “The Laguna Gloria Art Museum: An Exhi-
Steinitz, Carl, and Peter Rogers (1970). A bition of a Work in Progress.” Farish Gal-
Systems Analysis Model of Urbanization lery, Rice University, April 8-24.
and Change. Cambridge, Massachusetts: Venturi, Robert, and Denise Scott Brown
MIT Press. (1968). ““A Significance for A&P Parking
Stern, Robert (1980). ‘Classicism in Con- Lots, or Learning from Las Vegas.” Ar-
text.” In Charles Jencks, ed., Post-Modern chitectural Forum, March, pp. 37—43f.
Classicism. London: Garden House Press, Venturi, Robert, and Denise Scott Brown
pp. 89-91. (1971). “Ugly and Ordinary Architecture, or
Studer, R. (1965). ‘Christopher Alexander’s
the Decorated Shed (Parts 1 and 2).” Ar-
chitectural Forum, November, pp. 64-67,
Notes on the Synthesis of Form.”’ Architec-
ture Association Journal, March. and December, pp. 48-53.
Venturi, Robert, Denise Scott Brown, and
Suckle, Abbey (1980). By Their Own Design:
Steven Izenour (1972). Learning from Las
Ten Architects Discuss Their Process of De-
Vegas. Cambridge, Massachusetts: MIT
sign and Construction. New York: Whitney
Press.
Library of Design.
Vickers, Sir G. (1983). Human Systems Are
Sullivan, Louis H. (1934). Kindergarten
Different. New York: Harper and Row.
Chats. Chicago: Scarab Fraternity Press.
Vidler, Anthony (1977). “The Idea of Type:
Summerson, John (1979). The Classical Lan-
The Transformation of the Academic Ideal,
guage of Architecture. Cambridge, Massa-
1750-1830.” Oppositions, 8, pp. 95-115.
chusetts: MIT Press.
Vidler, Anthony (1978). “The Third Typol-
Tafuri, Manfredo (1976). Architecture and ogy.” In AAM, ed., Rational Architecture:
Utopia: Design and Capitalist Development. The Reconstruction of the European City,
Cambridge, Massachusetts: MIT Press. pp. 28-32. Bruxelles: Editions des Archives
Taller de Arquitectura (1984). Lecture pre- d’Architecture Moderne.
sentation of work, School of Architecture, Wade, John W. (1977). Architecture, Prob-
Rice University, November. lems and Purposes: Architectural Design as
Thorndike, E. L. (1931). Human Learning. a Basis Problem-Solving Process. New
Cambridge, Massachusetts: MIT Press. York: John Wiley and Sons.

Thornley, D. G. (1963). ‘Design Method in Waldman, Peter D. (1982). “The Design Nar-
Architectural Education.” In J. C. Jones and rative.” Journal of Architectural Education,
D. G. Thornley, eds., Conference on Design 35, no. 4.
Methods. Oxford: Pergamon. Wallas, G. (1926). The Art of Thought. New
Tolman, E. C. (1938). “The Determiners of York: Harcourt, Brace.
Behavior at a Choice Point.’”’ Psychology Watson, J. B. (1930 [1924]). Behaviorism.
Review, 435, pp. 1-41. New York: W. W. Norton.

Utitz, Emil (1923). ““Zweckmassigkeit und Watts, Ronald D. (1966). ‘““The Elements of
Sachlichkeit.’’ Dekorative Kunst, 26, p. 144. Design.” In S. A. Gregory, The Design
Method, p. 85. London: Butterworths.
Utzon, Jorn (1963). “Platforms and
Plateaus.” Zodiac, 10, p. 117. Weber, Max (1957 [1922]). The Theory of So-
cial and Economic Organization (translated
Van der Velde, Henry (1907). “Credo.” Vom by A. H. Henderson and Talcott Parsons).
Neuen Stil, p. 150. Glencoe, !Ilinois: Free Press.

Bibliography 217
Wertheimer, Max (1945). Productive Think-
ing. New York: Harper and Row.
Wiebenson, Dora, ed. (1982). Architectural
Theory and Practice from Alberti to Ledoux.
Chicago: Architectural Publications.
Winch, Peter (1958). The Idea of a Social
Science and Its Relation to Philosophy. Lon-
don: Routledge and Kegan Paul.
Wingler, Hans M., ed. (1969). The Bauhaus.
Cambridge, Massachusetts: MIT Press.
Winters, Ed, and Andrew Peckham, eds.
(1984). “Architecture and Hermeneutics.”’
Issue, 4, March.
Wittgenstein, L. (1969 [1933—1935]). Blue
and Brown Books. Oxford: Blackwell.
Wittkower, Rudolf (1971 [1962]). Architec-
tural Principles in the Age of Humanism.
New York: W. W. Norton.
Woods, Shadrach (1964). ‘Project for the
Free University of Berlin.’ Architectural De-
sign, August.
Woods, Shadrach (1965). ‘Free University,
Berlin.” In World Architecture, vol. 2. New
York: Viking Press.
Wright, F. L. (1943 [1932]). Autobiography.
New York: Hawthorn.

Ziegler, Oswald L. (1973). Sydney Builds an

Opera House. Sydney: Oswald Ziegler
Publications.
Zwicky, Fritz (1962). “Morphology of Propul-
sive Power.’’ Monograph on Morphological
Research No. 7, Society for Morphological
Research, Pasadena, California.

Bibliography 218
\\
Illustration Credits
12 36
Julian de la Fuente, Guillermo (1968). ‘The Five Architects: Eisenman, Graves, Gwath-
Venice Hospital Project of Le Corbusier.”’ mey, Hejduk, Meier (1972), pp. 32, 34. New
Architecture at Rice, 23, April, composite York: Wittenborn Art Books, Inc.
from pp. 8, 12, 16, 27.
37
13 Paraline and perspective drawings after Al-
After Watts 1966, p. 85. len and Oliver 1981, pp. 19, 20.

14 41
After Archer 1963, p. 6. Schnaidt, Claude (1965). Hannes Meyer:
Buildings, Projects and Writings, p. 18. New
17-21, 26
York: Architectural Book Publishing Co.
Basic drawing of the facade of San Sebas-
tiano after Wittkower 1971 (1962), p. 52. 42
Le Corbusier (1965). Oeuvre complete, p.
22
191. Zurich: H. Girsberger.
Le Corbusier (1951). The Modulor, p. 117.
London: Faber and Faber. 43
Plan redrawn from Process Architecture, 8,
27
no: 9; 1979; p. 62:
Alexander, Christopher (1964). Notes on the
Synthesis of Form, composite from pp. 137, 48
142, 151, 153. Cambridge, Massachusetts: Progressive Architecture, 1, January 1979,
Harvard University Press. p. 85.

29 49
After Ziegler 1973, p. 28. Landau, Royston (1968). New Directions in
British Architecture, p. 72. New York:
30 George Braziller.
Jencks, Charles, ed. (1980). Post-Modern
Classicism, p. 64. London: Architectural De- 50
sign and Academy Editions. Lane, Barbara Miller (1968). Architecture
and Politics in Germany 1918-1945, p. 110.
32 Cambridge, Massachusetts: Harvard Univer-
Davis, Allen J., and Robert P. Schubert sity Press.
(1976). Alternative Natural Energy Sources
in Building Design, p. 78. New York: Van 51
Nostrand Reinhold. Photograph by L. G. Rowe.

33 55
AAM (1978). Rational Architecture: The Re- Five Architects: Eisenman, Graves, Gwath-
construction of the European City, p. 70. mey, Hejduk, Meier (1972), p. 37. New York:
Bruxelles: Editions des Archives d’Architec- Wittenborn Art Books, Inc.
ture Moderne.
56
34 Model photograph and sketch reproduced
Alexander, Christopher, Sara Ishikawa, and with permission of Venturi, Rauch and Scott
Murray Silverstein (1968). A Pattern Lan- Brown.
guage Which Generates Multi-Service Cen-
58
ters, p. 18. Berkeley, California: Center for
Environmental Structure. Plan abstracted and redrawn from Jencks
1980, p. 54, and photograph a detail from
Progressive Architecture, October 1981, p.
95. (Original photograph by Deidi von
Schaewen.)
59
Model photograph from Architectural Rec-
ord, 8, August 1980, p. 100. Sketches from
Wheeler, Karen Vogel, ed. (1982), Michael
Graves: Buildings and Projects 1966-1987.
New York: Rizzoli International.

60
AAM (1978). Rational Architecture: The Re-
construction of the European City, p. 58.
Bruxelles: Editions des Archives d’Architec-
ture Moderne.

61
Braghieri, Gianni (1982). Aldo Rossi, pp.
120, 121. Barcelona: Editorial Gustavo Gili,
S.A.

|llustration Credits 222

Index
Act theorists. See Brentano Bricolage, 176, 185-190, 201
Alberti, L. B., 57, 69, 120 Bruner, Jerome, 46, 110-111
Alexander, Christopher, 71-73, 88, 122, Buhler, K., 42
127, 142 Burnham, Daniel, 20, 22, 79
Alexandria, 82, 85
Alley Theater, 125-126 Cassirer, Ernst, 197
Analogy. See a/so Heuristics Centre Pompidou. See Plateau Beaubourg
anthropometric, 80 Chomsky, Noam, 172
canonic, 82 Chunking, 55
iconic, 82 CIAM, 155, 157
literal, 80 Collingwood, R. G., 171
takeover, 96 Colquhoun, Alan, 177, 179, 180, 187
Anderson, Stanford, 91 Commercial strip, 127-128
Appreciative contexts, 76. See also Frames Constancy of appreciation, 94, 104, 105
of reference Conventionist position, 124, 129-130,
Archer, Bruce, 49 142,144
Archigram, 138 Craik, Kenneth H., 162
Architectural categories, 121-122. See also
Der Ring, 155
Normative architectural positions
Descartes, R., 99, 173
analytical framework for, 123-124
Design behavior. See also Problem-solving
assignment of priority to, 145-147
behavior
autonomous, 130
characteristics of, 91-110
devices for, 124-134
Design episodes
examples of, 121-122
examples of, 3-31, 136
substantiation of, 134-148
rule making in, 76
Asimow, Morris, 47—48
structural manifestations of, 34-36
Associationism, 42
structure of, 3
ATBAT Project, 65-66
Design methodology, 37, 110-111
Attoe, Wayne, 119
Dessau Siedlungen, 155
Backtracking, 20, 95 Dinocrates, 82
definition of, 35-36 Duhem, Pierre, 174-175
discussion of, 107-109 Durand, J. N. L., 64
Barker, Roger G., 162
Eastman, Charles, 56
Barragan, L., 134
Echenique, Marcel, 163, 165
Bartlett, Sir F., 44
Ecole des Beaux Arts, 46, 180, 184, 187
Bauhaus, 116, 155, 157
Ecole Polytecnnique, 46
Behaviorism, 44—46
Eisenman, Peter, 95-97, 131, 133, 153,
phase and rigid-state models of, 45
177,179
stimulus-response models of, 44—45
Eliot, T. S., 179
Behrens, P., 155
Enabling prejudice, 37
Berlin, Isaiah, 170, 171, 173
Enlightenment, 173, 177
Berlin Free University, 88, 92-93
Environmental relations, 85-86, 101, 162
Bernstein, Richard, 141, 154, 170, 173,
Ethnomethodology, 200
174, 199
Euclid, 98-99, 198
Bisociation, 46
Blondel, J. F., 47, 64 Fact-value dichotomy, 169-170, 171, 175,
Bloomer, Kent, 80 199, 200
Bofill, Ricardo, 184-185, 189 Feyerabend, Paul, 174-175
Bonatz, G., 155 Formalist position, 124, 130-134, 144
Boulée, 177 Form follows function, 124, 137-138, 155—
Bounded rationality, 39 157, 201
Brasilia, 158 Forms of life, 141, 173, 175, 200
Brentano, Franz, 42, 43
Frames of reference, 76, 103-104. See also House in Mathew Street, 130
Appreciative contexts; Forms of life House Il, 97, 179
broad classes of, 153-154 Hubbard, William, 144
foundational, 141 Huet, Bernard, 131
interpretative, 153 Husserl, Edmund, 173-174
relativist, 141 Hypothetical-deductive system, 163
Frampton, Kenneth, 49, 98, 134, 177
Frankena, W. K., 143
Iconic model, 47
Franzen, Ulrich, 125-126 Information processing theory of problem
Functionalist position, 124, 125, 142, 144, solving, 50-74
155-158 background, 50-51
naive, 155 commonplace examples, 52—54
postulates and directions in, 51-56
Gadamer, Hans Georg, 195 Interpretation. See also Scientific interpreta-
Generate-and-test procedures tion; Hermeneutics
definition of, 59 criteria for, 189-190
rules for, 74 meaningful, 194-198
strategy for, 74 open-ended, 195
Gestalt movement, 43-44
concept of Gestalten in, 43 Jencks, Charles, 130
concept of mindset in, 43 Johanson, John, 31-32
concept of schemata in, 44 Jonas B. Salk Center, 177, 178
Goddard Library, 32 Julian de la Fuente, Guillermo, 28-31
Graves, Michael, 97-98, 131, 175, 187-188,
Kahn, Louis I., 118, 177-178
189
Kimball Art Museum, 177
Gropius, Walter, 157, 177
Know-how, 112, 199
Gugelot, Hans, 48
Koestler, Arthur, 46
Guild House, 180
Koetter, Fred, 157, 185
Habermas, Jurgen, 175 Koffka, K., 43
Habraken, J. N., 65 Kohler, Wolfgang, 43
Halfpenny, William, 64 Kresge College, 181
Hall, Edward T., 162 Krier, Léon, 131, 192
Heidegger, Martin, 173-174 Krier, Rob, 1, 192
Hermeneutics, 195 Krimerman, Leonard |., 170
Herron, Ron J., 139 Kulpe, O., 42
Heuristic reasoning. See a/so Heuristics
Laguna Gloria Art Museum, 181-182
definition of, 75-77
Lakatos, Imre, 91
end-justification of, 96, 102
Lane, Barbara Miller, 155
logical structure of, 77-78
Lang, Jon, 162
Heuristics. See also Heuristic reasoning
Language
appreciation of, 94
autonomous, 130
classes of, 80—93
classical, 64, 88
constraints, 78-79
figural component of, 176-185
definition of, 75
formal, 88
examples of, 88
formal component of, 176-185
order of application of, 107-108
historicist, 130, 184
referential bases of, 95-96
modeling, 164
switching of, 103-107
pattern, 88, 127
technology of, 76
privacy of, 194
Hill climbing, 61
work on the, 153, 176
Hirsch, E. D., Jr., 189, 196
Leach, Edmund, 171
Hochschule fur Gestaltung, 48-49
Hollein, Hans, 82, 84

Index 226
Le Corbusier, 28-31, 65, 82, 91, 117, 118, Moore, Gary T., 162
WS), Wl, We WAS EE ley, al7e7/ Morris, Robert, 64
Ledoux, 177 Municipal Control Building for Quail Valley,
Les Arcades du Lac,.185-—187 130, 131
Le Viaduc, 185, 187
Lévi-Strauss, Claude, 171, 185 Negroponte, N., 56
Logic, 154 Neorationalists, 153, 191, 192
abductive, 101, 102-104 Newell, Allen, 51, 75
Aristotelean, 163 New York Five, 177
coherence of, 134-140 Niemeyer, Oscar, 158
deductive, 101, 103 Norberg-Schulz, Christian, 120-121, 135,
falsification of, 174 146
inductive, 101, 103 Normative architectural positions. See a/so
interior, 35 Architectural categories
logical-empirical, 163, 168 analytical framework for, 123-124
modeling, 169-170 categorical systems for, 119-121, 135
Lowry, Ira, 165 doctrinaire versions of, 119-120
Lynch, Kevin, 1, 162 as doctrines about ends, 140-143
as doctrines about means, 143-145
Mach, Ernst, 98-99 examples of, 115
Magdeburg Siedlung, 155 frames of reference for, 153-154
Maison Domino, 91 ideological basis of, 121
Maldonado, Tomas, 48 kinds of, 124-134
Mallin, Samuel B., 76 logical coherence of, 134-140
Man-environment relations, 162 logical structure of, 116
M. D. Anderson Hall, 184 substantiation of, 134-148
Means-ends analysis surface features of, 116
examples of, 62-64 urcorpus of, 148
rules for, 74
strategy for, 62 Oklahoma Theater Center, 32
use of, 104-105 Operations research, 48
Merleau-Ponty, Maurice, 76—77, 79, 110, 174
Panofsky, Erwin, 180
Meyer, Hannes, 116-117, 119, 121, 135, 139,
Park Regency Apartments, 127-129
155
Peirce, C. S., 102
Mies van der Rohe, Ludwig, 118
Pelli, Cesar, 181-7184
Mitchell, William J., 56, 111
Perkins, P. N., 75, 76
Models
Perrault, Claude, 64
computer, i161
Peters School, 117
concept of, 162, 163-164
Pevsner, Sir Nikolaus, 115
descriptive, 166
Phenomenology, 173-174
econometric, 99-100
concept of bracketing in, 174
explorative, 166, 168
concept of lebenswelt in, 173
hierarchical classification of, 166, 168, 169
Piaget, Jean, 171, 172
iconic, 47
Piazza d'Italia, 130, 132
languages for, 164
Planning
operational, 162, 165-168
empirical data for, 161
planning, 166, 168
as instrumental action, 161
predictive, 166
instruments of, 158-159
problems of interpretation of, 168-170
models, 166, 168
rigid-state, 45, 110
orthodoxy, 158
staged-process, 46-50
Plateau Beaubourg, 33-34
theory and, 165
Plato, 190
Modern Movement, 155-158
Popper, Sir Karl, 163
Moore, Charles, 80, 130, 132, 181

Index 227
Populist position, 124, 125-129, 142, 144 Robert R. Herring Hall, 181-184
Portland Public Services Building, 187-188 Rogers, Richard, 33-34
Preiser, Wolfgang, 162 Ronchamp Chapel, 82
Price, Cedric, 138 Rossi, Aldo, 1, 131, 159, 191, 192, 194
Priene, 82, 85 Rowe, Colin, 157, 185
Problem. See also Problem space; Problem Royal College of Art, 49
solving Royal Institute of British Architects, 49
definition of, 39, 52
ill-defined, 40-41, 104-105 Sanoff, Henry, 162
well-defined, 40, 59, 104 San Sebastiano, 57, 59, 62, 64, 69-70
wicked, 41, 78 Satisficing, 39, 190
Problem solving. See a/so Problem; Prob- Scientific interpretation, 49, 122, 140, 154—
lem space 175, 199, 200
behavior, 51-74 alternatives to, 170-175
creative, 39, 51, 110-111 Schon, Donald, 94
serial aspects of, 106 Schultze-Naumberg, Paul, 156
theories of, 51-74, 108 Scott Brown, Denise, 180
Problem space, 51—54, 65, 68. See a/so Scully, Vincent, 140
Problem; Problem solving Shaw, J. C., 51, 75
definition of, 51-52, 65-68 ~Short-term memory, 55
examples of, 52—54 Significance and meaning, 194, 196-197
new information in, 94-95 Simon, Herbert A., 39, 51, 75, 109, 167
superficial enrichment of, 97 Siza, Alvaro, 134 —
transformation of, 100 Situations, 76-77. See also Merleau-Ponty
Problem-space planning Skinner, B. F., 45
alternative methods, 73-74 Smith, Thomas Gordon, 130
categories of, 67 Solution generation strategies, 56—76
evaluation of, 95, 104 Sommer, Robert, 162
hierarchical decomposition methods, Staged-process model, 46-50
67-71 Archer's, 49
hierarchical decomposition-recomposi- Asimow’s, 47
tion methods, 67, 71-73 background, 46—47
purpose of, 65 of the Hochschule fiir Gestaltung, 48-49
Protocol Stirling, James, 110, 184
applications, 55 Structuralism, 171-173
definition, 1-2, 51, 53 Style
examples of, 3-27 concept of, 109-110, 146
Puzzle forms, 46 determination of, 37
historicist, 184
Quatremeére de Quincy, 191 Suckle, A., 28
Quotation, 181, 201 Sydney Opera House, 33, 82-83
Systems approaches, 162. See a/so Models;
Rapaport, Amos, 162 Planning
Rawls, John, 145-146, 147
Representation Taft Architects, 130, 131
geometric aspects of, 98-99, 144 Taller de Arquitectura, 184-185
mathematical expressions of, 99-100 Taut, Bruno, 157
modes of, 97-98, 164 Team Ten, 157
referential aspects of, 98 Teleological doctrines, 143
Rigid-state models, 45-46 Theory and practice, 149-150
application of, 110 Thorndike, E. L., 39
fundamental steps in, 45 Thornley, Denis, 49
Ringturm Travel Agency, 82, 84 Trevi Fountain, 130
Rittel, Horst, 40

Index 228
Trial-and-error procedures
distinguishing features, 56—57
random, 56
use of, 105
Trieste Regional Hall, 192, 194
Type
building, 85, 130
city, 191-192
elemental, 85, 87
formal, 122
idea, 190-191
as model, 85-87
organizational, 85, 87
as third typology, 191
as typologies, 85-87, 104
use of, 176

Unitarian Church (F. L. Wright), 82

Urban morphology, 192
Utitz, Emil, 156
Utzon, Jorn, 33, 82, 83, 134

Van der Velde, 118

Venice Hospital, 28-31
Venturi, Robert, 27, 137, 138, 180, 181
Venturi and Rauch, 127, 129, 180
Vernacular, 131, 134
Vidler, Anthony, 87, 153, 191, 192
Villa Stein—De Monzie, 118
Vitruvius, 120, 121

Waldman, Peter, 82
Walking City, 139
Watson, J. B., 45
Weber, Max i75
Weissenhof Siedlung, 155
Wertheimer, Max, 43
Winch, Peter, 173
Wittgenstein, Ludwig, 173, 175
Wittkower, R., 69
Woods, Shadrach, 88, 92-93
Wright, Frank Lloyd, 28, 82
Wundt, Wilhelm, 42
Wurzburg School, 42—43, 55
concept of Aufgabe, 43, 44, 49

Zube, Ervin H., 162

Zwicky, Fritz, 167

Index 229
 as | 7

a ee ee
; Om = )
, 7] eS
a ~ <€,

reas & .

7 2 ow é
» *

Un a

OO

a i

7 ae Se
a oe me Oe
<= ~ i—- » ©

= —
oa a

— ane)

—_ =a
- 49006
: —— SS “9 = ela S27) Pres

fe
en
ee
i
A
a
ann ep eo
_
ee at
ues
shee
arcs
= ovr
aon. i 7

owe ma (OY

- 6 —

: |
<2) =

_ =
= .
2 J

2 a

ew =

at 7
SS a
—-* :
=o ee - £5
eon —

= => 8
7 7 7
- 4

a yy Pre
= - U0
a a 7 ve

a oan

=e
 7
:

i /
i
i » +.
// |
The MIT Press
Massachusetts Institute of Technology
Cambridge, Massachusetts

NATIONAL GALLERY oF
ART
32 D800786 18485p.,
P

|| ISBN 0-262-68067-X
90000>
ae ele Sb
16.95
i}

9 "780262"680677 | |