THE DEVELOPMENT OF DISCOURSE:
A LANGUAGE FOR COMPUTER ASSISTED CITY DESIGN
by
William Lyman Porter
B.A. Yale University
(1955)
B. Arch. Yale School of Art and Architecture
(1957)
Submitted in partial fulfillment
of the requirements for the
degree of
Doctor of Philosophy
in City and Regional Planning
at the
Massachusetts Institute of Technology
August, 1969
Signature of Author.....
................
Department of Urban Studies an
Planni
.
August 18, 1969
Certified by........
hesis Supervisor
Accepted by..............
....
..
..
.....
.........
.II
..
••.•
Chairman, Departmental Committee
on Graduate Students
Archives
1969
25IST.
'
NOVLkIA,
R1E
-
ABSTRACT
The
A
of
Development
for
Language
Computer
As
by William Lyman
ISCOURSE:
isted City Design
Porter
Submitted to the Department of Urban Studi es and
in partial fulfillment of the
Planning on August 18, 1969
Philosophy in City
requirement for the degree of Doctor of
and Regional Plannin
DISCOURSE is a language and computer system to assist
large
the city designer in the process of the design of
scale physical environments. The system can accommodate
large amounts of data describing the environiment which can
and new categories of the description
be readily changed,
can be added at any time.
which
data
as well as new
DISCOURSE permits the designer to devel1op his own ways
of working by having a data structure which i s not committed
to any particular set of environmental descr iptors and yet
which allows the des igner to speci fy hi s own without
extensive bookkeeping; and by having a larg e vocabulary of
building-block manipula tions out of wh ich th e designer can
fashion his own special requests. The d es i gne r can alter the
alter it subje ct to criteria which
description directly or
he specifies.
In Chapterll is the record of an experiment in the
to the
replication of a design which was fundamental
analyse
to
attempts
III
Chapter
system.
development of the
the process of city design in functional terms which will
point toward the use of a computer system. It is argued that
what designers do when they design can be thought of as
describing) the environment,
"describing" (and further
it. The
"testing"
"transforming" the description, and
language of DISCOURSE accommodates these three activities;
its structure and manipulations are presented in Chapter IV.
Chapter V contains an example of the use of DISCOURSE to
simulate the use of recreational facilities in an urban
environment.
The emphasis in system development to-date has been to
build the capability of the computer system to handle
internally the data and manipulations used by designers and
to permit the designer to create and change both. Plans for
the immediate future include improving the ease of entering
improving the quality of the
large quantities of data,
display, and facilitating communication with other sources
of data and systems for manipulation.
Thesis Supervisors:
Aaron Fleisher
Professor of City and Regional Planning
Kevin A. Lynch
Professor of City and Regional Planning
iii
ACKNOWLEDGEMENTS
DISCOURSE is the product of the efforts and skills of
many. My thanks to those who have been most involved is not
meant to suggest that, by the support which they gave me,
they were not making progress in their own right, progress
which deserves reporting from their points of view, just as
this dissertation reports from mine.
I would like to express my special thanks to Professor
Aaron Fleisher whose insights and criticisms are an integral
part of the thinking which underlies this work. I would like
to thank Professor Kevin Lynch for his quidance particularly
in the early phases of the work when the direction was hard
to find. My thanks go also to Dr. John Boorn with whom much
of the early thinking was done.
Many of my colleagues in the Departments of Urban
Studies and Planning and of Architecture have contributed to
the development of the ideas and have given me their support
and patience as this document was being prepared. My thanks
are to all of them.
The design of the first version of the DISCOURSE system
was the work of Katherine J. Lloyd. She deserves much of
the credit for bringing clarity and order to the idea of
"DISCOURSE" at its inception.
Wren McMains is reponsible for the development of the
of DISCOURSE, and, beside
second and present version
of the system, he has
design
the
to
bringing his ingenuity
ideas about
particular
many
been able to help generalize
"mapping"
on
work
able
done
design. Stanley Hoderowski has
him and
to
go
and other features of the system; my thanks
computer
the
of
to others who have part of the development
system.
TABLE OF CONTENTS
PAGE
CHAPTER
1.
INTRODUCTION .
.
.
.
.
. .
.
.
. .
..............
City Design
City Designing .
.
. .
.
..
. .
.
. ........
.
.
The DISCOURSE Language
. . .
Outline of the Remaining Chapters
.
.
.
.
.
3
.
11
..
.
15
..
AN EXPERIMENT IN THE REPLICATION OF A DESIGN . .
II.
The Experiment . . .
. .
.
.
. .
....
6
.
.
.
.
.
1
.
.
.
Summary of Descriptors and Transformation Rules.
Results.
.
.
.
.
.
.
.
.
.
.
Strategy for Replicators . .
III.
.
.
. .
.
0
0
25
a
0
*
45
Functions of Testing .
THE DISCOURSE LANGUAGE .
.
.
. .
60
.
Functions of Transformation
.
64
. ..
68
69
Basic Structure of the Language
Characteristics of ATTRIBUTES: C
and CHARCONS
.
System Components
V.
0
21
FUNCTIONAL ANALYSIS OF CITY DESIGN
Functions of Further Description
IV.
*
19
. . .
.
. .
. . .
. .
.
.
.
.
RVA
93
109
SIMULATION OF ENVIRONMENTAL USE
112
Description of the Experiment
112
Uses of the Experiment .
The Experiment.
. .
.
.. .. ......
Files for the Experiment .
.
.
.
114
117
123
V
CHAPTER
VI.
PAGE
FUTURE DEVELOPMENT OF DISCOURSE
Problem Oriented Features
.
.
Use Prototypes .......
.
.
.
.....
.
..
.....
Graphics and Modes of Representation ..
APPENDIX
.
.
.
.
. .
BIOGRAPHICAL NOTE
FOOTNOTES
.
.
.
BIBLIOGRAPHY .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
135
.
.
.
.
136
.
.
.
.
143
.
.
.
.
.
.
.
.
149
.
.
.
.
.
.
.
.
185
.
. .
. .
.
.
.
.
.
129
132
...
Description of the Environment .
.
.
.
. . ..
.
.
.
186
196
LIST OF FIGURES
F I GURE
PAGE
(1.01) DISCIOURSE placed with respect to
. . . . . . .
other higher level languages
.
.
.
.
.
.
18
(2.01) - (2.07) Figures illustrating the experiment
36-42
in the replication of th e design of Ciudad Guayana.
(4.01)
Data structure of DISCOURSE
(4.02) - (4.10)
Illustrations of the
. . ..
0 *
operators in DISCOURSE
. .0.
. . . . . 75
logical
.
*
a
.
.
.
- (4.21)
Illustrations of the DISCOURSE
language including CHARVARS and CHARCONS
. . . .
.
76-83
(4.11)
. 98-108
(6.01)
Places identified by names, not coordinates. . 147
(6.02)
File to refer to places with names
.....
148
CHAPTER I
INTRODUCTION
rest of the thesis.
the
background to
introduction and
is an
This chapter
It describes: 1. what the writer means by
"city design"; 2. describes the characteristics of city design
the
summarizes
the following
an outline of
4. concludes with
language; and
DISCOURSE
the
of
design
for the
criteria
3.
behavior;
solving
problem
designers'
and
problems
chapters.
has
DISCOURSE
and
generating
studying alter native
These configurat ions
environment.
of a
implications
of pol icies
set
of
the
the design,
the
heart of
the
configurations
be
may
is
which
in
designer
the
assisting
do with
to
design, or a representation of the erivironment which is useful
in determining what the design proble .m is.
DISCOURSE is not a tool fo r data gathering though through
its
the
use
information he
the
needs.
communication
designers although
of
a
get
may
designer
ideas
to
fut ure this
in the
of
sense
people
might be
what
help in
been developed
Nor has
design
better
other
than
one of
its
uses.
DISCOURSE
should help the designer
system which
mastery
over some
generating
of
the
conceived from
has been
his material
design ideas.
It is
begi nning as
sense of
to achieve a
the
during
devoted
to the
a
time he
is
designer's
communication with himself.
The major
difficulty in
the development
of a
computer
CHAPTER
PAGE 2
INTRODUCTION
I:
language to assist the city desi gner is that what the designer
not been well describe d.
does has
vocabulary
and
computer, one
grammar
a
syst em
computer
which
appropriately
designer's ways of
working one must have a
of what
A second dif ficulty
line
they are.
the designer
between what
well as
best, as
to
the
fairly clear idea
the dividing
is that
does
the computer
and what
c)f the
the nature
in order to design
and,
res ponds
a
the vocabulary
idea of what
mus t have a clear
on
implemented
be
whi ch can
and grammar of the city designer are;
a
design a
For in order to
criteria by
which that
judgment is made, are not clear.
In the development of DISCOURSE
we have tried to program
This
what we could describe of what the designer does. (1.01)
both difficulties
still leaves
of what
how much
not know
described;
unresolved:
the designer
one does
first,
to be
does remains
and second, one does not know whether an activity,
better be performed
describable, may not
even though
designer than the computer. And
there will be
by the
to further complicate matters
a moving frontier between
what is programmable
and what is not.
user, it
the subtle
will force him
issues which, if he were to
he might not have
surely
work
presently
to satisfy the
system is rich enough
If the
in
being
to face.
an
used
through design
design in the traditional manner,
He may work more
unfamiliar
by
to think
demands of
way.
designers
Many
will
slowly, he will
techniques
suddenly
not
become
CHAPTER I:
INTRODUCTION
available.
The change
PAGE 3
in design aids will
and opportunities for working on
from those
problems which are different
past and may
of the
offer constraints
change design
designer may become increasingly resourceful
own
much of
abilities as
in the use of his
customarily done
he has
But it
aid of the computer. (1.02)
accomplished with the
be that
what
behavior. The
and designers as
know it now
design as we
which
the goal
continues to
augment the skills
DISCOURSE, to
govern
of
the development
the designer as
may
they are
presently trained are not adaptabl e to the new techniques.
so
is
If
of
we know
him, will have been ill chosen.
I.
The city designer
cities over time
spatial form of
is concerned with the
what society does
as cause and as effect of
is concerned with it as means
and feels;
he
and
concerned
he is
CITY DESIGN
with it
to the wide
particularly sensitive
environment affords
as
to himself
an end
in
range of
and to
to other goals;
itself. He
is
experiences the
others.
His
special
skill is in the manipulation of the elements which affect that
experience.
environments
enhancing:
His
be
contribution is
own normative
of
high
responsive
to
quality
present
contributory to the development of
which make up the society.
These norms
and
and
that
usually that
they
future
be
needs,
life
and
the individuals and groups
(1.03)
have been further
safety, health, and comfort of
specified to
include the
people, adequacy of facilities
CHAPTER I:
INTRODUCTION
PAGE 4
and of settings for the conduct of everyday life and business,
a wide
range of diversity of
many others
ranging from
facilities and of
settings and
creature comfort
to transcendental
may design not only
the public spaces
pleasure. (1.04)
The city designer
of towns
and cities,
but also
the suburbs,
over spaces, industrial areas, parking
strip commerce,
systems of
arterials, left
areas, junk yards, and
movement, communications,
public
services and the other systems which affect the ecology of the
total environment as well as key aspects of those systems; and
the
city's
graphics,
designer may work
furniture, lighting
at the location of
and
color.
The
activities and physical
facilities, at the design of pieces
of the environment or its
systems,
result
at
the
processes
development or at
which
in
environmental
the policies which would set
the stage for
better design.
The designer might
corridor for
a new
address a problem of the
highway to
be built
design of a
through an
already
developed area.
The goals he will have to contend with may be
many:
federal
of
the
satisfactorily to
and
other roads
state
agencies
to
connect
and to
minimize the
costs of
land acquisition and site preparation; of the city to minimize
the
loss
relocation
politically
to
its
of
tax base
residents
damaging; of
clarity in orientation,
and
and
minimize
businesses
the drivers
the
which
on the
good lighting and well
problems
of
might
be
road who
need
marked access
PAGE 5
CHAPTER I:
INTRODUCTION
points; of
the residents who
generated by the use of the
highway and minimum breaks in the
movement to adjacent areas;
residential pattern to allow free
of the would-be
politician who wants clear
to build his reputation and
in the
who are involved
and other
highways, utility systems
design of
issues upon which
consolidate his constituency; and
other professions and groups
of the
from nuisances
require freedom
components of
the urban environment who have their own design criteria,
The designer engages in many activities during the course
of his work:
finding out what
the area are, who
local environment,
are;
in
the physical characteristics of
lives in there, how they use
and value the
who changes the area, and what the ir
plans
trying alternative locations and designs of the corridor
where
detail
needed I
pedestrian crossings,
road connections;
studying
the
possibilit ies
developments,
ai r rights
and
for
regional
the environment wil 1 change
predic ting how
with and without the kind s of interventions which woul d result
from his influence,
and value
imag :ining how
the changed
criteria for
a
the institu
the design ideas , spelli
of
new
institutional
purposes clearly and
people will percei ve, use,
cha nging
ona 1
arrangements associ ated
out the environmental
arrangements;
and
to develop criteria by
alternati ves may be evaluated.
developing
environment,
with
impl ications
trying
to
state
which the design
CHAPTER I:
INTRODUCTION
PAGE 6
We face the typical situation in large-scale
objectives numerous, poorly related, and
design:
possibly shifting, constraints
imperfectly known and
changing, the range of alternatives limitless and not
occurring along
a single continuum;
even the
potential variables not certain. (1.05)
To these observations must be added the necessity for the
spatial,
city designer to deal with
components: his problem area
so large
that it connot
short time;
simultaneously from
he is considering issues
changes to that environment
period of time;
very long
is usually large geographically,
be comprehended
any point nor even in a
which may imply
temporal and experiential
and
extending over a
concerned with
he is at once
the easily and the not so easily measured. (1.06)
CITY DESIGNING
II.
and
City design
problems fall
ill-defined
problems.
criteria, for
into
designer
The
what constitute
the category
all at once.
the criteria.
implicitly,
for those
A problem
environment and satisfactory values
"Problem",
perceived
for
exists between known
is a condition in which apparently no fit
of
reaching
satisfactory values
criteria, and for design alternatives
alternative states of the
is
of complex
therefore,
solutions,
and
it
includes at
is
in
part
least
the
inappropriateness of the solutions perceived which creates the
problem.
As he
designer
nears the achievement
is usually
able to
of design
demonstrate
alternatives the
that his
proposal
overcomes major disadvantages of the existing environment plus
INTRODUCTION
CHAPTER 1:
existing
development
PAGE 7
trends,
which his
disadvantages
evident to him
even
though
design overcame
the
not have
may
until very late in the design
precise
process.
been
It is
as if the designer were engaged in a process of finding almost
simultaneously problem and solution. (1.07)
Ways of Working
Some designers may prefer to work from the formulation of
goals and tests
measured,
which
by which the achievement of the
to the
inventing
would ensure
prefer to
work in
possibilities
of
policies and
their achievement;
other ways,
for action
specifications
other designers
for example,
to goals
goals may be
which
may
from particular
can reasonably
be
achieved or from attempting to explore the environmental needs
of some particular group of people
with those people in order
to develop both goals and appropriate actions.(1.08)
The designer may limit his area
a single
goal, a single
of Inquiry by working on
small geographic area,
one activity
system such as the recreation or transportation system, trying
to
improve
the
environment
population, or looking
for
a
single
at the impact of a
environmental change or the constraints
group
in
the
specific system of
on it such as zoning.
(1.09)
As the city designer works he attends to different scales
of the
environment in
aspects of
the problem.
response to
He may
thinking about
be interested in
character of certain points in the environment.
different
the design
At the larger
CHAPTER 1:
scale he
PAGE 8
INTRODUCTION
the qu al ities
may specify
points, at the smaller scal e he
further specify the qualiti !
the
to
s
those
present at
may work out the character of
Or, a t
area in detail.
the
to be
smaller
might
scale, he
be satisfied by the designer
!
of form, structure and mate I als.
i
design the designer will probably
At the early stages of
related systems, and
be tested. As
be
very
about the
finding out
more time
spend
environment and
the standards by which the
and
problem
the design
design is to
the description may
he proceeds the changes to
large
begin
will
Later, he will probably
formulated more clearly.
its
to
be
spend more
time testing the various design alternati ves he has developed;
the
changes to the
description may
beginning to the
end his major attention
from information
gathering and the
goals to
assemble them
trying to
smaller.
be
From
the
will probably shift
inve ntion of
in some
designs and
acceptable design
alternative. (1.10)
Although sequences of activities
solving have been
as
a
identified they may be no
things
checklist of
designer:
Juring creative problem
the order
to
remembier for
more useful than
any
in which any desij gner works
particular
may be very
usefully his own. (1.11)
Graphic Media
Maps
or
spatial
designer's way of
displays
working.
are
fundamental
The visual representation
design is the traditional focus
to the
of the
of the designer's act ivit ies;
CHAPTER I:
PAGE 9
INTRODUCTION
aware of many aspects of
considerations of patterns makes him
(1.12)
the design.
means
by
simulations
and
models
dimensional
view, sections, perspectives, three
plan or map
environment:
the
of representing
ways
different
designer uses
The
of
movies,
photographs and videotape, to mention a few.
But none of
these which display motion
to
change over time permit des ign changes
be
through space or
made readily.
It
is difficult a lso to record large amou nts of data in map form.
be
varl•ables can
Several
clTarity, but
particularly
si ngle d rawing
if the variables take
there cannot be many variables
And a
a
mapped on
with
many values,
before the drawing is unclear.
prolif eration of symbols,
keying location
to specific
information, is also difficult to read.
The desijgner
tight fit
his design
may deal with
wi th the real world.
At one extreme
the description
ambiguously d efined.
which are
And, at thie
or a
his sketches
the land and may refer to
may be only a pproximately placed on
categories of
in a loose
either scarcely
or
other extreme, the sketches
or models may be intended as a Pýreci se scale representation of
some piece of
about
the
the real world or of
Norld
like
road
combinations of loose drawings
some specific information
there
are
with detailed information,
for
ali gnments.
And
example, traf fic flow diagrams in which the width of the lines
represents the exact
volume of traffic, but
the land is approximate.
the placement on
Large
environmental des i gn,
scale
translati on,
through
help the
because
information
the
designer works
growth
the
of
it.
of
of the material
Certain kinds
with
population
population
the
means
may
better
diagrammatic
graphic or
of
or
groups
population
to translate
distribut ion
the
by
Either
is symbolized in the new
had
tabular,
in
example,
income,
which
scale
concepts.
as increas ing his understanding
of having
displayed
smaller
designer
providing new associatio ns with what
mode as well
like
vi sual and non visual
design includes both
may,
PAGE 10
I NTRODUCTION
CHAPTER I:
or
density
income
circulation
form:
for
pattern
of
of
be
for
against
several
an
air
from
i ts
terminal.
But
the
where
idea
inception, the violence
is visual
spatial
or
don e it by translation
less than by translation into
symbol may be
into a visual
a verbal symbol.
Ideas may need to be nurture d in their relatively untranslated
states during part of their development.
The designer
ways
than
geometries,
the
operates on the
verbal fiel d:
symmetrical
goodness
of
In different
form,
transformations, doodling
play all yield new justaposi tions
which enrich the vocabulary of
different
and
form
and other new relationships
possibilities for designing as
well as for testing the desi gn.
pattern yields
visual field
information not
And the scanning of a visual
always expected
and problems
which one may not be verball y aware of.
Just
what accounts
for the
insights
which a
designer
CHAPTER I:
PAGE 11
INTRODUCTION
gains from graphic
displays and from visual
not at
Nor is
i t clear how translation
from one
representation to anothe.r is accomplished.
What is
all clear.
mode of
clear is
manipulations
that the designer needs
a wide variety
of displays
which can be readily manipula ted as he works.
In
of
amounts
desig ning involves
summary, city
noin-visual,
and
visual
non-geographic material, much of which
during the course of the
the
des i gn.
see
the
pre-experience the
in an
is appended or changed
design
designed envi ronment.
unpredictable order on all
large or small
and
The designer requires vivid
of
consequences
large
geographic
and manipulable displays easi ly t:ranslatable
to
use of
one to the other
and
changes
The
designer works
or part of the
scale, and with aI "tight" or
to
problem, at
"loose" fit with
the real world.
THE DISCOURSE LANGUAGE
III.
The DISCOURSE language has been
designed as a high level
problem oriented language in which the data structure would be
transparent to
the
the user and in
manipulations existing
language.
It affords
the
which there would be
in any
general purpose
opportunity
language, responds rapidly to answer
to
program
most of
computer
in
the
user requests, and has a
limited range of displays. Soon it should have a wide range of
vivid
displays, have
sources and should
ways
of
interfacing with
have associated with it
other
data
teaching programs
which make It easier to learn and to use than at present.
CHAPTER I:
assisting any u ser
user
vernacular are
Those
other end of
by many
own
user's
the
to
oriented" languages
sometime s called "problem
at t he
are close to
which are usable
close
are
w hich
for
languages
computer
grou p range from those which
own vernacul ar to those
groups.
and those
of
design
the
to
Approaches
the user's
PAGE 12
INTRODUCTION
the range,
"general purpose"
The terminol ogy is somewhat misleading because, as
languages.
indicated in
of languages
whole spectrum
the
f igure (1.01),
which is being discussed is already "problem oriented" in that
within it are designed to
the languages contained
be used in
scientific and commercial applications. (1.13)
Some
(usually
either be
somewhat lower)
in
higher level
languages.
They
is an
may
to be
program or remain
The
were written.
language in which they
latter type of language is
DISCOURSE
be written
translated into an object
interpreted by the
other
can
languages
higher level
sometimes called an "interpreter".
in
"interpreter" written
level
the higher
language of AED. (1.14)
Languages
at the
top end
of
the spectrum
tend to
be
closely tied to a particular problem or set of problems and it
is in this sense that the label "problem oriented" is helpful.
For
example,
STRUDL
is
a
structural engineers written
ICES:
the
problem
oriented
in ICES using two
Command Definition Language (CDL)
manipulations language,
very much
Some typical lines typed by the
language
for
subsystems of
and ICETRAN which is
like FORTRAN.
(1.15)
engineer user in STRUDL might
PAGE 13
INTRODUCTION
CHAPTER I:
be:
TYPE SPACE FRAME
UNITS INCHES KIPS
JOINT COORDINATES
X 0.0 Z 0.0 Y 0.0 SUPPORT
where SUPPORT, for
of wh ich
example,
is one of two
have been anticipated by the
them after the
only use
at the
engineer wishes to perform
The computations which the
joint.
of restrai nt
indicating the degree
is FREE
options, the other
ICES system programmers; he need
been entered,
appropriate data have
and the computations will be performed.
Where many of the tasks to be performed are well defined,
Where the
appropriate.
offers the
his own
and
task is
new subsystem using
ICETRAN.
Even
though
ICES
well defined,
not as
the engineer-programmer
opportunity to
be very
STRUDL level vay
oriented language at the
a problem
to define
Definition Language
the Command
of
modification
subsystems
is
relatively easy the ICES system seems particularly appropriate
for
assisting in
tasks
which can
be
well
defined at
the
outset.
The major justification of
oriented languages
is to
building higher level problem
relieve the user
having to deal with the complexities
data
internal to
the computer
and
different components of the computer
to
the user
facilities
in
is
of the
burden of
of the management of the
the communication
system.
among
What one brings
terms of
input,
manipulations, and
another
matter
and
depends
upon
output
the
PAGE 14
INTRODUCTION
CHAPTER I:
of problem
user's problems and manner
characteristics of the
solving. (1.16)
Unlike some other fields, city design does not have a set
of well
Therefore,
built.
user
defined sub-problems
to
program
manipulations in
for which
the possib ility
DIS COURSE offers
the
in
order
language
that he be
facilities could
with
w ide
a
able to modi fy it
be
for the
range
of
to s uit
himself and in order that designer-programmers can improve the
system without having
to rely entirely on
the expert syst ems
programmers who constru cted the system in the fi rst place.
The rapid response
time of the system
has been achie ved
by virtue of the time s haring systems at M.I.T. which give the
user the appearance of having a large computer fully dedica ted
The reason for the rapid respon se is to al low
to his own use.
the
user the
opportun ity to
learn
question he has just posed without
he would if
he had to submit
result of
from the
the
getting off the subject as
and wait until
a deck of cards
the next da y for the results.
Until
recently work on the language has been concentrated
on the deve lopment of the storage, retrieval and manipulations
facilities.
Now the
creating a wide
associated
stress in development will
range of displays,
with getting
reading in complicated
data
from
and on
the input problems
other sources
and large amounts of
And the res ults of the present
be placed on
and
with
geographic data.
phase will include an up-dated
learner's primer plus examples of the use of the system.
PAGE 15
INTRODUCTION
CHAPTER I:
IV.
OUTLINE OF THE REMAINING CHAPTERS
nor the
to make
process of design is well enough understood
system obvious. The reflection can
the design of the computer
be seen
the language
the fact that neither
This thesis reflects
spent analysing
shares of time
in the roughly equal
design and describing the language to accommodate design.
is
anything
which
approximated by
computer could
could
even
a
be
that a
so clear
sequence of
instruct ions
computer is
an algori thm.
A
out.
out on some
be carried
which can
there
of instructions
a sequence
carry them
whether
which
does
designer
the
of
the issue
to
addressed
Design",
of
Replication
the
Experiment in
"An
II,
Chapter
The same issue restated asks if there is anything the desi gner
does which can be replicated by means of an algorithm.
In
Chapter
designers
design
which
end
of the
master
near the
fast growing city
Guayana, a
was
II, a
by
favored
Ciudad
planning for
in southeastern
the
Venezuela, was
selected as a subject for the replication. (1.17) An algorithm
the designers' thinking about
was developed to substitute for
the
design;
environment was
generated
and,
at the
a design
algorithm seemed
from
a
representation
beginning of
not unlike
to be
the planning
that of
capable of
of
how
phase, it
the designers.
refinement to
the
The
produce an
even better approximation of the design, or of modification to
produce some of the other designs
for the same city which had
PAGE 16
:INTRODUCTION
CHAPTER 1
been cons idered by the designers at other times.
luster:
the algorithm,
components of
The
if
they could,
designer?
to describe
a function.
which computes
are
algorithm, one needs
algorithm is
a procedure
functions, then,
Algorithms and
for which
are functions
But there
closely allied.
in a computer
be useful
To write an
because an
a function
new
more
rearranged, generate
by being
than one plausible design, might they
language for the
on a
then, took
no
algorithm exists. (1.18)
Chapter Three,
Design", is
"Functional Analysis of City
an effort to analyse city design in functional terms which are
more independent
Ciudad
Guayana
situations:
design
these
of
applicable to a wide range
are
of
"functions
description", the derivation and inference
further
of new d ata of the
description, "functions of transformation",
the data of
"functions
distinct
experiment. Several
design" emerged which seemed to be
of
than the
problem situation
of a particular
the alt eratio n of
the description, and "functions
of tes ting", the
comparison of data of the description with norms.
The
major
achievement
of
the
DISCOURSE
language,
explicated in Chapter IV, "The DISCOURSE Language", is
can accommodate
reference, the user
"places"
and
and
data structure
these functions. Although the
and manipulations
of DISCOURSE
are independent
can most easily think of
"descriptors of places":
ATTRIBUTES, names
which specify
that it
of empirical
the 1anguage as
a grid
or qualify
of LOCATIONS;
what
is
at
CHAPTER I:
INTRODUCTION
LOCATIONS.
Examples in
PAGE 17
Chapter IV
serve
to illustrate
the
basic features of the language for this interpretation.
Chapter
V,
"Simulation
Environmental
of
Use",
moves
somewhat closer to how a designer might use the language.
experiment ,which
simulates
facilities, demonstrates
language
functions
developed in
Description", and
are discussed
the
the use of
of
recreational
"Functions of
"Functions of Measurement". In
a
highly explicit
in the
categories
the
of
Chapter III:
some of the issues
context of
use
functions written
two
to
correspond
which
people's
which can be
The
of
Further
the chapter
considered in
formulation
of a
design
problem.
Chapter VI,
"The Future
ahead - pointing out tactics
the short run to city
issues surrounding
environment richer than
for increasing its usefulness in
designers, discusses the elusive issues
of "graphics and modes of
the
Development of DISCOURSE", looks
representation", and raises some of
the
need for
a
description of
that easily afforded by
the
DISCOURSE as
it stands.
In the
Appendix is the
developed in
used
for some
mig•ht want.
the same
implementation of
Chapter II, demonstrating
of the
how DISCOURSE
bookkeeping which
Also included is
the design
an additional
decision rules but with
the algorithm
analyst
experiment with
a minor change in
which caused major changes in the results.
can be
the data
CHAPTER
I:
I NTRODUCT ION
general purpose system building languages
special
purpose design languages
specific problem oriented languages
FOR SCIENTIFIC,
PAGE 18
*---eg.
LISP,PL-1,
AED, ICES,
FORTRAN
+-------DISCOURSE
*---eg. TRANSET
HIGHER LEVEL LANGUAGES
COMMERCIAL AND RESEARCH APPLICATIONS
LOWER LEVEL LANGUAGES
assembler languages
machine languages
on the
DISCOURSE occupies a range
Figure (1.01)
continuum within higher level languages offering to the user a
language in which he can store, retrieve and manipulate data
about his environment without having to do any of his own
the line
the top part of
programming (represented by
with
facilities
the
offering
and
associated with DISCOURSE),
the
adapt
and
to
modify
programming
own
which he can do his
(1.19)
system to his own special requirements.
PAGE 19
CHAPTER II
AN EXPERIMENT IN THE REPLICATION OF A DESIGN
II
Chapter
is
of
replication
beginning
of
language
and
record
the
for
a design,
the
the
the
was
the
DISCOURSE
be
will
short
a
perspective within
to clarify the
historical note
underlies
however,
First,
system.
in
experiment
that experiment
in
which
thinking
an
of
which the
experiment was conceived.
semina r on the city design
This exper iment grew out of a
process. (2.01) The
seminar started as an
the skill of th e city
designer.
The parti c ipan ts harbored two
first that designing
suspicions:
eff ort to describe
could be rat ionalized,
that
is that the ways in which desi gner s worked were , by and large;
rational
city
-
desig n
designer's
project,
among
even though the y mi ght
largely
was
ways of
kinds
favored
ndividually
is,
simi lar
in
srmla
1 in
number
- t he
most
act
of
topography,
i
t hat
were
arising
the
rivent ional,
and second that
worki ng
a nd, possibly,
desi gns
co
not know it,
visua
y
from
vities
(map)
loc al
to be
that
project
the
after
differences
differences
accommodated,
oatterns.
Behind
in
and
the
apparent variety there might be a kind of behavior which could
be described, and,
therefore,
replicated.
Recording how the designer works, trying to determine the
types of
information he uses,
his sketches, and
his analytical
procedures and
asking him to keep as complete
a record as
PAGE 20
II: REPLICATION
CHAPTER
possible of what he was doing and thinking might have been one
of research.
method
But
would
that method
have taken
an
enormous amount of time, not only on the part of the analysts,
but on
There
the part of designers
would have
been
whose work was
of communication
problems
as well
of confidentiality,
designer and
in order to be sure
study many designers
that the hypotheses
And this
which could
in hypotheses
have resulted
of the
necessary to
being advanced were of more than parochial interest.
approach would
the
problem of
as the
it would have been
research activity. Moreover,
with
was changing because
much the design
recognizing how
being analysed.
not
easily have been tested.
behavior.
This
completely
described
substitute
designer's
activity,
the behavior
of
been,
callied
have reproduced the
sketches, and
verbal comments which
when
faced
success, it then
about the
with the
same
initia
the processes i
of
a
the
designing"
The test
programming
the designer
might have been possibl
similarity of
"ci ty
same record
algorithms, to
done
aspe ct
algorithms.
means of
by
have required
e very
for
have been translated into
would have
success
woIuld
approach
designer's
would have to
have simulated the
might have been to
A second approach
the
of drawings,
would have
With
conditions.
to have speculated
the human
mind to
those which were programmed.(2.02)
The
behavior
which
Newell and
simulate was in the context of
Simon
were
trying
to
a very well defined problem in
PAGE 21
CHAPTER II: REPLICATION
was
to
added
ever.
then
distant
some
seemed
more
and
sketches
Moreover, the
drawings,
designer over
the entire
of a
verbal comments
behavior
of design
simulation
the
designers,
of
behavior
eccentric
apparently
the
design problems
of city
the ill-defined nature
when
logic;
course of
a
design project were not readily available.
reflec ted
procedures
the procedures
interested in
conventions
of
design;
required.
not
was
proce ss
design
tl he
we
because they
were
looking
the
well
design er.
accurate protocol of desii gn
by-product of an
the
of
behavior
the
for how
regard
without
the designer
of
reproduce the
we could which would
create procedures however
product
selected, was to
and the one which we
A third approach,
The
behavi or during
we
Instead
were
might embody
for
the
the
logical
equivalent of the designer. (2.03)
I.
experime nt was
The
aspect
of
the favored
growing industrial
this
writer was
design was the
level of detail.
were:
industry,
THE EXPERIMENT
an
attempt
design
for
to r e pl i cate
only
Ciudacd Guayana,
a
city in southeastern Vein ezuela
familiar. (2.04)
pl an for the use
The cho s en
of the lain d at
The categories of "uses"
major
city center,
and
one
fast
with which
aspect of
the
a very gross
(or "activities")
urbanization
-
a
residual- category for developed land consisting principally of
residence.
The "design" which was to be replicated was assumed to be
CHAPTER II: REPLICATION
"the
order
of
activities."
PAGE 22
occurrence
The
formation
of
the
various
commenced
desi
the environment
from
at
lopment
2.05)
developing the region.
describe an "initial
with the
in 1961,
it was
de
locations
of
these
first departure
the
corporation
the
time of
charged
with
It was necessary, therefore, to
s tate" of the enviro nmen t
including the
three types of activiti es selected.
The next step was to pick the variabl es from the existing
environment which
It
three activities.
some of
with
the locationis of the
woul d most influence
seemed reasonable
the vari ables
design discussions:
slope
preferred by industry;
which
had b een
of the land: 0-5
6-15 percent
for any
Orinoco River
partly
pro cess ed
percent was most
the riv er:
channel to the ocean
highly valued by in dust ries exporting
volume
and
the
nore than 15 pe rcent
purpose; (2.06 )
had a de:ep sea
centra 1 to
stil 1 buildable for both
industrial and resident ial purposes; and
unacceptable land
at least,
to start,
whic
raw materials and
finished
products;
and
the
was
arge
the
existing activities in the three categorie:s selected.
In order for t he representation of the design an d
states of
the envi ronment to
location, a grid
be unambiguous with
w as laid over the land which
of squares
within wh ich
differently
stated,
the limits
of one
the vari ables
points at whi ch
square could
be placed).
respec
provided a
could be
the data
ini
placed
falling wi
The grid
size
selected was the maximum which would still retain the identity
PAGE 23
CHAPTER II: REPLICATION
of the
The sacrifice was
initial and design states.
accuracy of
the description, but
of the
a description
the gain was
with fewer parts very much simpler to manipulate.
Figure (2.01), the
the replication was drawn
state" for
"initial
existing environment in 1 961
square one mile on a side;
the squares by
and the
wi th each g rid
and the variables were assigned to
the arbitrary rule that if more
than half
the
square was occupied by the variable in question it became t rue
of the entire square.
figure (2.02) , could then
The designer's "design state",
be drawn using
the des4:ription and
the selected variables of
grid size plus the order and
placement of the activities over
time which were part of the design.
The
of
problem
described:
write the
was,
replication
set
of
rules which
completely
then,
transform
would
figure (2. 01) into figure (2.02).
The t ransformation
the quanti ty of
activities, the order in which
they would be
of the
environmental
relative importance
assigned,
and the
variables
considered.
For
located on flat land next to
river if possible.
reduce
the
cost
and
be
other urbanization and next to a
Flat land was
of
was to
example, urbanization
thought by the designers to
urbanization,
urbanization could take advantage of
did exist;
considering
developed by
rules were
land next to a river
adjacency
to
existing
the urban services which
seemed desirable.
Next
in priority would be sloped land next to existing urbanization
PAGE 24
CHAPTER II: REPLICATION
arid,
if
category
of
If land
river.
would
urbanization
then
available
was
a
to
possible, next
neither
not
be
assigned.
The
apparent
only
retrospect,
description was created, one which,
the
other variables
presence in a
of
the
operated on
a location,
a useful
variable of the
description, depended
for
its
found within that
function, center of gravity, which
values found elsewhere.
can be seen that there were
designers'
in counter distinction to
location not on values to be
location but on a specific
in
to the center
proved to be
Thus a new
"compactness".
rules;
of the
form of the city. Proximity
of the existing activities
way of expressing
site became
different
part
been
"compactness" had
discussions of the
of gravity
of
trial
the
during
"compact"
development
ends of the
out the
letting it run
rather than
the
keeping
for
necessity
Also, in the
rules, it
three other functions needed:
next to a location,
and nearest of one
at
class of
locations to some other location.
Finally, the quantities of new activities were taken from
the design drawing, and the
order freely interpreted from the
design drawing and indirectly from the economists' projections
for the city.
PAGE 25
CHAPTER II: REPLICATION
II.
1.
SUMMARY OF DESCRIPTORS AND TRANSFORMATION
RULES
Description
1.1 Grid size:
one mile squares
1.2 Variables and their values:
1.21 slope 0-5 percent
1.22 slope 5-15 percent
1.23 unacceptable land (more than 15
percent or flooded)
1.24 river
1.25 deep sea channel
1.26 industry
1.27 urbanization
1.28 major city center
1.29 center of gravity of locations with
industry, urbanization, or major city center
1.3 Functions needed for the description:
1.31 variables occur at the location
1.32 center of gravity
2.
Rules
2.1 Quantitative measure of change required: 4 square
24 square miles of urbanization;
miles of major city center;
24 square miles of industrial land.
2.2
Assignment
urbanization, 3
orderings:
square miles
3
of industry,
center, 3 square miles of urbanization,
square
miles
1 square
of
mile of
and 3 square miles of
PAGE 26
CHAPTER II: REPLICATION
industry four times in succession.
2.3 Transformation rules
of flat
or sloped
proximity to the
center of
other urbanization, location
adjacency to
to a river, and
land, adjacencey
of
function
a
is
urbanization
of
location
The
gravity of activities.
to
deep
adjacen cy
channel,
sea
of adjacency
indust ry is a function
The location of
industry,
other
to
and
proximity to the center of gravi ty of all activities.
The location
location on
of the
is a
function of
to other major
center, and
major center
flat land, adjacenc y
proximity to the center of gravi ty of all activities.
elaborations
the
figure
(2.07)
a re
of 2.3,
above,
w hich,
In
"decision
when coupled
trees",
the
with
assignment orderings, generated figure (2.03).
Additional functio )ns needed for the rules:
2.4
2.41 variable does not occur at a location
2.42 adjacency
2.43
nearest
(of
2 variables
or
more)
to
a
location
III.
THE RESULTS
Figure (2.03) represents the result of the application of
the
rules
to
the "initial
state".
The
differences
chief
between figures (2.02) and (2.03) are in the use of the sloped
land.
The
differences
can be
further
figures (2.04), (2.05), and (2.06)
noted
by
examining
which compare the patterns
PAGE 27
CHAPTER II: REPLICATION
with the pattern generated by
generated by the decision rules
the designers for each of the three activities:
urbanization,
industry, and c enter.
The resu Its raise some issues to be investigated further,
issues which it would be practical
rules were
cornput er. Better "fit"
i mp lemented on a
migh t
design
to investigate only if
have been achieved
the
with the
the assignment
by varying
The generated des ign seems to be relatively
orderings. (2 .0 7)
insensitive t0 assignment orderings except in the later stages
when the abil it:y
to predict events is not
great.
Clearly if
all
act ivities wanted the same variabl
at the location, that
is,
the transformation rul es for each
ere the same, then the
desi gn configuration
of
the
land
desire,
the distribu tion
wou d depend only upon
the
and
ordering.
assignment
This
phenomenon can I e obser ved in assignments 15 to 25 of ind ustry
This extreme might
and urbanizatioi
city
where the
boundries
was
proximity to th
The other
supply
limited
of total
or in
a
land
be reached in a gr owing
within the
metropolitan
muni cl pal
region
wihere
region al core was highly valued.
extreme , when
in design because
same land,
may be reac hed
try to fit
eac
partic ularity of the site
activity.
In
this
case,
compete for
no activities
the
the
designers often
to the approp riate
transformation
rules
for
different activ ities would have not similar branches and would
determine the
configurations; the assignment
have no influence.
orderings would
PAGE 28
CHAPTER II: REPLICATION
Where assignment
configurations,
orderings do have
there
may
other
be
an influence
rules
on the
which
can
be
formulated for the competition for land among activities which
political resources available for
represent their monetary or
the
purpose
land acquisition
of
facilities
appropriate
instead
and
of
of
construction
the
point
in
the
time
of
assignment.
The assignment orderings used in the experiment were very
the
detailed in
overly
later stages,
the
and they
would happen;
of what
designers' sense
respect to
of development with
the initial stages
gross for
were
could
fault which
a
easily be remedied by different assignment orderings.
example, more
have
been
reflect
changed during
the
changing
the
growth
locational
activities as the city grew in size.
real changes
in the activities
more to
industry and
transformation rules
The
residence.
land for
sloped
flat land for
weight to
process in
might
order
the
of
criteria
for
to give,
rules
the transformation
by varying
achieved
have been
the designers' design might
Better "fit" with
to
various
And, in order to reflect
themselves as the
city grew,
new activity types might be introduced with new transformation
rules.
One could
land,
have varied
for example,
changing
the qualities
attributed to
"unacceptable"
to "sloped"
the
at
location 7,13, a site favored by many of the designers for the
major city center,
and one where the original
decision as to
CHAPTER II: REPLICATION
which
category
PAGE 29
to use
in
figure
(2.01) for
the
"initial
state", 6-15 percent or "unacceptable", was difficult to make.
One could
introduce new
information, for
land owned by Orinoco Mining
Company,
example, that
the
locations 5,11 6,9 6,10
6,11 7,9 7,10 7,11,, would not be ava ilable for development for
at
least half
"unacceptable"
growth period
thle
and
would
For this design
the one mile grid size was
too large in some areas, but needles sly small
If the differences
design could
the
to
ascribed
experimenters,
as
Ian id during that time . And, the grid size might
have been varied.
designers'
be classed
lack
in others.
between the generated design
not be redu ced, the
of
ingenui ty
or to the designer:
on
and the
failure could be
the
part
of
the
either his artistry or his
error. More generally, if the differ ences between many designs
by
generated
designs done
reduced
by
rules
in d ifferent
by designers in
rules
situations
tho
seemingly arbitrary, then somethin g
or
creative
contributio n
"Creat i ve" , when
used as
violations
system
of a
might
the delay beyond
of the
another.
of one
The element of surprise
only be
case
and
the
indentif ied.
thought of
as Iocal
exageration of
to some important space;
wall of enclosure.
rule system
each
been
the previous spa cing (point
next opening in a
violations
have
of rules :
in
many
of each designer's un ique
above, can be
height of one facade in relation
the
situations could
were different
which
and
may well
the
or
of expectat ion)
Of
course, the
be consistent
with
mixed with joyous curiosity
PAGE 30
CHAPTER II: REPLICATION
can yield may be consistent
through a sequence
movement
regular
of
violations
designed
carefully
which
with a system
the designer is at least
of behavior in environments of which
intuitively aware.
thought of as those which
But creative ideas can also be
violate
conventional
regularity strongly implied.
this sense, the
The more
with some
way
a global
in
ideas
new
"creative" a n idea , In
and the
it
more of observation is changed by
more wide ranging are its repercussions throughout the bod' y of
concepts making up the present theory of design. (2.08)
ideas
Conventional
are
the
ideas
conventional
readjustment:s in the
a fter
factors external
conventional ideas cannot be
the
taken place.
body of knowledge have
hindsight, a nd without reference
creative and
tomor row
of
of
today may
The creative ideas of
interpretation and thought.
become
rules
used
the habitually
In
to both,
distinguished from
each other. (2.09)
Suppose that the transforma tion rules have replicat ed the
land use des ign
alone it is impossible
rule set
conventional
or
replicated many
called
desi gns.
land use
land use
them most closely
some land
use design a new
generated
better
land
to infer that
Sup pose
creative.
conventional for
adhered to
By
w-th reasonable fidelity.
use
looking at the
tho se
that
ru1 es are
the
rule s
t hen be
rules might
The
design
and des i gns
designs
than
which
f from
conve ntiona 1 des igns.
set of rules were
have
inferred which
had
been
done
CHAPTER II: REPLICATION
previously,
PAGE 31
then they
might be
termed
design termed "creative" from which
brand of creativity
the
"creative", and
they were inferred.
might better be termed a
This
very high level
of competence.
IV.
STRATEGY FOR REPLICATORS
Exploring the "conventional" and the "creative" implies a
different kind
above.
of development
Instead
of
designer's design,
even
working
the
are,
the main of
experiment described
bet ter"fit"
It should
rules
with
wh ich
be observed
were
an activity's adjacency
the seeking of flat land, and
the
simplicity and
working tow lard
it implies
they
as
consisted in
toward
general applicability.
the promise of
that,
from the
utilized
to itself,
the seeki ng of central position
with respect to the whole development. I t w'ould be tempting to
try various formulations
do on
several
land
of these rules to see
use designs,
as we 11
rules which are involved
kinds of
designs from
designers'
experiments would
be to
that
how well they
as to
explore the
in i mpr oving the
point on .
find the
fit with
goal in
The
these
of rules
simp lest set
to
account for the largest part of the desi gn.
But,
in
worki ng toward rules
would expect some
replicate any
struck
of g reater
loss in the accu racy with
part icular design.
between rules
which are
A
so
gene rali ty one
which they might
ba lance will have
specific as
application to other situations and those
to be
to have
no
which do a very bad
job of replication or generate foolish designs. (2.10)
CHAPTER II: REPLICATION
There may
with
the
another
be designs which
land
suggested.
use
of
the
rules.
the point
different criteria
but
of
for
(for
view
land use
The
the consequences
of his
work in
is not
the
what
extent
designers
that
Of course,
to satisfy criteria
it
should
for the
a question
the
of how
its own terms.
in
example,
(2.11)
to
are
working
ground level).
but
realms;
designs,
have been
that a good design ought
one could argue
to replicate
explanations
primarily
other realm, not understandable in
in many
Two
may
design
from
seen at
simply be
are impossible
designer
his design
environment is
design may
design
First, the
aspect
generating
PAGE 32
satisfy
replicator of
The
themselves.
rules
governing the design may be derived primarily from a realm the
may simply have been looking
replicator is not looking at, he
at the wrong variables.
has to
second explanation
The
between the representation
representation may
with the
of the design which
difference
is the object
intent of the designer. (2.12)
of the replication and the full
The
do
to a
depend
large
extent upon
planning and design;
conventions of the field of city
the
a land
use plan, often has the same legend which limits the choice of
variables; it is course in detail;
time only;
and puts only one
the replication
intent
may fall short
because the
rules
it represents one point in
land use at one location. Thus,
of including the
governing the
expressed in the conventional media
design
full design
may not
be
of design. The replicator
CHAPTER II: REPLICATION
may
fail
because
PAGE 33
the
rules
governing
the
design
were
unrepresented or unrepresentable, perhaps not becoming evident
until the design is realized.
V.
THE SHIFT FROM REPLICATION TO DESIGN AID
There are
two major reasons
least temporarily,
why, at
the project shifted at this point from attempting to replicate
designs
to
designer.
the building
The first
of
a
system
to assist
has to do with the
the
city
substance of design,
and the second with the formal structure of the replications.
Subs tance
Substantively, the
replication could be
looked at
as a
way of
o btaining the consequences
form of
a set of rule s about how that form should evolve.
(2.07)
rules, figure
Guayana
in terms
could
of environmental
be interpreted
effort to generate a "least cos t" city form:
construe tion because of the ut i lization
cost"
to
channel;
industry
because
"least cost"
of urb an
as
an
"least cost" for
of flat land;
f proximity
The
to
services, at
the
"least
deep
sea
least in
the
early st ages, because of the pO ssibility of extending existing
services
to
adjacent
transpor tation because
as possi ble.
develo pment;
purpose of producing
"least
of kee p ing the development
It is not a
to obser ve that the
and
cost"
of
as compact
cha 1lenge to this line of reasoning
rules may not have been the
a "least
cost" city form.
they stand were conceived from the
best for the
The rules as
point of view of trying to
replicate what the designers actually did.
The point is
that,
PAGE 34
CHAPTER II: REPLICATION
that the
having recognized
once
"least
on to
then go
designer could
the
some fashion,
in
other criterion
any
cost" or
express
seem to
rules did
that
rules for
design better
purpose, test them, and observe their consequences.
Structure
The
the
structure of
variable
the
characters,
transformation rules;
names:
industry,
function
and
the
for
state"
of the
the "design state"
theorem to be proven:
environment; and a
and
the "initial
axiom:
an
the
postulates:
a list of
description
the
for
needed
acceptable
adjacent;
urbanization, nearest,
functions
replication contains:
of the designers.
In building the replication one could work forward to and
to
cut down
replication.
which would help
design state, a procedure
backward from the
the time
develop the
needed to
the
set up the replication, the
Once he has
(2.13)
rules for
replicator's job is to find the rules of interpolation between
the initial and
final states. A designer, on
the other hand,
can be thought of as constantly inventing and re-inventing all
the
structure
the
of
components
studying
components and
different
outlined,
just
the
trying
consequences of
his
choices. (2.14)
A
system
replication
implications
which
could
from
would
accommodate
also accommodate
design
experiments
assumptions.
effort to determine what kinds
experiments
Chapter
in
Two
in
drawing
is
an
of city design assumptions can
CHAPTER II: REPLICATION
be made clear.
PAGE 35
CHAPTER II: REPLICATION
PAGE 36
CIUDAD GUAYANA in 1961
1
123456789012345678
1
.*.
2
3
4
5
6
+.+.*. 2 1 ....
*.+.+.+.+. 2+. * 1 . . . .
*.*.* .... 1 +.+ . . . . .
*+. . . . . . 1 1 .+.+. . ..
7
8
*.*.*0
9 *.*.
10
11.
12
.
+.*.* . ..
+.*.*. 1 1 . .
......
2
*. .
.
... .
13
14
15
.
.
.
.
1
.
.
. +.+.
2 . . .
4
5
6
7
8
9
1
+.
11
12
+.+.+.+.
. . . . . .+.+.+.
0 00 0 a0 +.
. . . . . +.
.
2
3
. +.+ .
.............
.
.+.+.
1
.
.
13
14
15
1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS -
-
'1' = URB1961(20);
'2'
IND1961(21)
= NON EMPTY LAND
= RIVER(8); '-'
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+'
CHANNEL(12);
'*'
= CHANNEL & RIVER; ' ' = NO REF ATTS
=
Figure (2.01) Ciudad Guayana as
it was in 1961 when the
Venezuelan Development Corporation was formed.
"urb1961"
means "urbanization" in 1961; "ind1961" means "industry" in
1961.
CHAPTER
PAGE 37
II: REPLICATION
AN ALTERNATIVE DESIGN FOR CIUDAD GUAYANA
BY THE CITY DESIGNERS IN THE DEVELOPMENT CORPORATION
123456789012345678
.
1
+.** * 2 .
+.*.*.
1 1 1 .
2
3
.*.
1
2
3
4
+.*.+.2 1 1 1 ..
5
*.++. ++.+.
6
7
9
10
11
12
13
14
15
4
5
*.*.*.2 2 2 1 .+.+. 1 ....
6
*.
7
*.*.*.
8
2+. 1 1 1 2 2 .
. 2 2 1 1 . 1 1 .+.+
2 1 1 3 3 .++.
*.*.
*. 2
. 2
2 2
2 2 .
2 2
2 2 2
2 2 1
*.
.. .
. . . . .
. .
1 . . .
8
2 3 3
1 1+.+. . . . .
. . .
1 1 1 . 1+.+. .
1 1 . 1 1+..
1+.+.+.+. .
+.+.+. .
.+.
9
10
11
12
13
14
2
2
2
1
1
. . . . . .+.
15
123456789012345678
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS -
IDESIGN(2);
-
'1' = UDESIGN(3);
'2' =
'3' = CDESIGN(4)
= NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+' = RIVER(8); '-'
= CHANNEL & RIVER; '
' = NO REF ATTS
'*'
CHANNEL(12);
=
An alternative design for Ciudad Guayana
Figure (2.02)
created by the city designers in the Development Corporation.
"udesign" means their design for "urbanization", "idesign" for
"industry" and "cdesign" for "center".
CHAPTER II: REPLICATION
PAGE 38
PATTERN FOR CIUDAD GUAYANA GENERATED BY DECISION RULES
1
3 4 5 6 7 8
123456789012
1
2
3
+.*.*.
4
+.+.*.2 1 2.
5
6
7
8
9
10
11
12
13
14
15
+.*.*
*.+.+.+.+.
*
*.
2
1
2
2
1
2 1 .
2 . . 1 .+.+
*.*.*.
*.*.
2+.
1
..
2 2 1 B . 1 1 .+.+.
*.
2*0*.2 . 1 B 1 . B .+.+. 1
2
2 2
2 .
1 2
2 1
1 2
S2
2.
2
.
1
2
1
1
.
1
.
1 1
.
B+.+.
.+.+
22 1 1 .
. +.
1 2 1 .
1 1 .+.+.+.+. .
2 .+.++.+
2 .+.
. .+
.
.
.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
123456789012345678
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS -
-
'1' = URBANION(6);
'2'
INDUSTRY(7); '3' = CENTER(5)
'B' = 1, 3
'.'
= NON EMPTY LAND
= RIVER(8); '-'
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+'
'*' = CHANNEL & RIVER; ' ' = NO REF ATTS
CHANNEL(12);
Pattern generated for Ciudad
Figure (2.03)
decision rules specified in this chapter.
=
Guayana by
the
PAGE 39
CHAPTER I I: REPLICATION
COMPARISON OF GENERATED PATTERN (URBANIZATION)
WITH DESIGNER'S PATTERN (UDESIGN)
123456789012345678
.*.
1
.
1
2
+.*.*. . .
2
3
4
+.**. A AA .
+.+.*. . A 2 2 .
3
4
*.+.+.+.+.
5
**.*.
6
*A . . .
7
8
9
..
*.*.*.
* . .
...
12
1 1
13
.
14
.
15
..
2 . . . .
A A . A A .+.+..
. . 1 AAA . 1 .+.+.
1 1 1 .
....
*.*1
10
11
.+. 2 A 2 .. .
. A .+.+.
2 A+.+
1 A A 2+.+.+.+.
1 .
1 .
1 .
6
7
A . . .
8
. ...
9
. . . . .
A A 2 . 2+.+.
1 2 A 2 . 2 2+..
.
1
. .
5
10
11
12
•
*+.+.+. .
13
.+.
0+.
14
15
45673
1234567890123
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS - -
'1'
= URBANION(6);
'2'
=
UDESIGN(3)
'A'
= 1,
2
'.'
= NON EMPTY LAND
= RIVER(8); '-'=
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+'
= CHANNEL & RIVER; ' ' = NO REF ATTS
'*'
CHANNEL(12);
Comparison of the pattern of urbanization
Figure (2.04)
generated by the decision rules ("urbanization") with the
pattern of urbanization of the designers ("udesign").
CHAPTER
I1: REPLICATION
PAGE 40
COMPARISON OF GENERATED PATTERN (INDUSTRY)
WITH DESIGNER'S PATTERN (IDESIGN)
123456789012345678
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1
0*, .
+.*.*. A1
+.*
*.
.
+.+.*. A . 1 .
.
*.++.+0.+. A+. 1 . . 2 2 .
*.*.*0 A 2 2
.++o
*
......
*.
A A . . . . . 0+.+0 . . .
*.*.*. A 2 2 . .
. .+.+.
.....
**.
A AA . 2 2 ..
. .+0+ . ...
*. A A A AA
A
...
. .++
.. ..
1 A 2 2 21 ....
+. .
. 2 A 2 . . .+.+ +.+. .
1 1 . 1 1 .+.+.+..
1 . I . 1 .+.
. . 1 . . .+.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
123456789012345678
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS - -
'1' = INDUSTRY(7);
'2' =
IDESIGN(2)
'A' = 1,
2
'.'
= NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+' = RIVER(8); '-'
= CHANNEL & RIVER; ' ' = NO REF ATTS
'*'
CHANNEL(12);
=
Figure (2.05) Comparison of the pattern of industry generated
("industry") with the pattern of
by the decision rules
("idesign")/
designers
industry of the
CHAPTER II: REPLICATION
PAGE 41
COMPARISON OF GENERATED PATTERN (CENTER)
WITH DESIGNER'S PATTERN (CDESIGN)
1
123456789012345678
1
2
.*.
+.*.*.
3
+.*.*
4
5
+.+.*0
*.+.+.+.+.
.+.
6
*.*.*
7
*
8
9
10
11
12
13
14
.
.
.
*.*.*. ..
**. . . . .
*.
.. . .
. . . . .
. . . . .
. . . .
. . . .
15
.
.
.
.
+.+
.
1
.0.
. .
.
.
.
.
.
.
. .
.
.
.
.
.
.
.
.
.
.+.+. .0.
.
1
2
.
.
45
3
.
.
6
.
7
1 . 2 A .+.+0 . .
..
. 2 2 . * 1+ +. . . . .
. . .
. .+.+.
..
. .....
. .+.
.0
. .+.+.+.+.
.
.+.+.+. .
*+.
.
.
8
9
10
11
12
13
14
15
. . . . . .+.
123456789012345678
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS - -
'1' = CENTER(5);
'2'
=
CDESIGN(4)
'A' = 1i, 2
'.' = NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+'
= CHANNEL & RIVER; '
'*'
CHANNEL(12);
'
= RIVER(8); '-'
= NO REF ATTS
Figure (2.06) Comparison of the pattern of center generated
by the decision rules ("center") with the pattern of center of
the designers ("cdesign").
CHAPTER II: REPLICATION
PAGE 42
1al or a2 and not a6,
a7 or a8
2 b7
3 al not a2
(5 b4
(10 read next
(9 c9d7
b4
8)c9d7
(7 c9d7
6)c9d7
RULE FOR ASSIGNING URBANIZATION
nd not a6,a7 or a8
2 b5
7 al not a2
(14
5 -
3 a1
not a2
6 b6
5 b6
9d6 12)c9d6 11)c9d6 10)c9d6 (9 c9d6
8)c9d6
RULE FOR ASSIGNING INDUSTRY
1 al or a2 and not a8
2 b8
3 b7
(7 read next
(6 c9d8
(5 c9d8
T4)c9d8
RULE FOR ASSIGNING CENTER
LEGEND: 1=slope0-5 2=slope6-15 3=unacceptable land 4=river
5=channe I 6=industry 7=urbanization 8=center
9=center of gravity of 6,7,and8; a=at the location
ore locations)
b=beside the location c=closest to (of 2 or mn
(1)=node number of the decision tree
d=assi gn ;
Figure (2.07)
Decision rules which generated Figure (2.03).
PAGE 43
CHAPTER III
FUNCTIONAL ANALYSIS OF CITY DESIGN
A function
is the
specification of
a rule
which, when
applied to values falling within its domain of definition, may
be
evaluated. Computers
functions which
demand
can be
algorithms, realizations
used to
compute their
is programmable on some computer.
to assist
consider
designers with
of city
whether any
values. If
a
form of an algorithm, then it
function can be realized in the
how
of
(3.01) In order to consider
computers it
is necessary
to
thought of
in
can be
design
functional terms.
The replication of the last chapter was an algorithm but,
it might
that was
be argued,
general application. This
do
formalized.
to
and trying
There
case, not
chapter is an effort
extract
sections
are
to rebut that
might
be
functions
of
part which
the
discussing
"Further Description", of "Transformation",
of any
designer does or
into some of what the
assessment by looking
might
an isolated
and of "Testing"
These three different kinds of functions have been arbitrarily
constructed
in
order
to
thinking, and to help reveal
the
explore
designers'
ways
of
the sorts of capabilities needed
for computer assistance.
At any instant in time during
designer can be
environment
the process of design, the
thought of as having data
and rules
of
inference
which describe the
drawn from
information
CHAPTER III:
about
and
environments
environment
combination wi th
rules, in
PAGE 44
FUNCTIONAL ANALYSIS
related
his data,
The
matters.
permit him
to further
the description, or to
describe the environment, to transform
test the description agai nst the achievement of some goal. The
data
which
"primitive",
our
"basic"
are
to inferences
or true by definition;
For example,
(3.02)
space, and time.
of
two
types:
and "derived", or true by
or "laws"
conventional notions
are
of logic,
if
mathematics,
"primitive"
data were
the locations of houses, then the total number of them, or the
average distance between
them would be "derived"
data.
Data
resulting from the combin ation of of "basic" data and rules of
inference are "inf erred" data.
primitive
data gi ven
The
above that
conclusion drawn from the
"the
houses consti tuted
a
neighborhood" woul d be an "inferred" datum.
An inference
can take
many forms,
common are: the combination of
statement:
but
two of
the most
a universal and an existential
"all drivers will be alerted to possible action by
any sign more than one square
foot in area within twenty feet
of the road and which carries a message of danger"; "this sign
satisfies the previously stated
sign will alert
conditions";
drivers. The universal statement
phrased as a conditional statement,
form. And the second is
is
the
value
("signs
evaluating the function
in
therefore, this
can also be
a variation of this first
the functional form, t = f(x) where t
which
alert
drivers")
f(x) where "f" is
the universal statement and
obtained
by
the rule expressed
"x" is the needed information
PAGE 45
FUNCTIONAL ANALYSIS
CHAPTER III:
from the environment, the arguments
distance from the
the sign,
"evaluating"
of the function, (size of
content); and
road, and message
the rule
means applying
to
the arguments
and
obtaining the value.
If the function is continuous there
of the
arguments which can be
found to satisfy
-
within some standard, k. (t + or
"size of
example both
were
road"
expressed
is a range of values
k = f(x))
In the previous
the "distance
the sign" and
as intervals
the function
of
discrete values, but the message content
from the
than
rather
values
had to be the single
This function was continuous with respect to
value, "danger".
two out of three of its arguments.
The discussion which follows is organized under the three
functional
of
functions
identified above:
types
"further
description", "transformation of the description" and "testing
The examples
are intended to illustrate
types, not to serve
as a comprehensive list of
of the description".
the functional
what designers do.
I.
Functions
types:
data;
FUNCTIONS OF FURTHER DESCRIPTION
of further
"measurement
description consist
functions" which
result
of two
in
major
"derived"
and "extension" functions which are defined to be those
which result in "inferred"
data.
Measurement
Derived data
which
receive
as
are obtained
arguments
from "measurement"
the
primitive
data
functions
of
the
r
result in
manipulations t o
Many of the
values in
those realms.
They are a name attributed
to a place
(3.03)
there and
or not
category having the same or a
can be
a s k wher e
.He can
c:oincid
they
or where
the
of a
designer can
n this area?", or
transit stop
transit stops?"
a member
simi lar name. Assuming that the
descript on exi st,
of the
"where is the
last observation,
make this
each of
geometric
are of the
be there
points" are
and
pr imitive data which the designer has
type "nominal".
categories
mathematical
apply logi cal,
and
description
which can
PAGE 46
FUNCTIONAL ANALYSIS
CHAPTER III:
e
"where are the
the "visually
strategic
with transit
he ne eds to
ask
stops.
look at
To
each place
where one category occurred and se e whether the other occurred
If
as well.
thought of
each occurrence
as a
member of a
members
two
category at a
set by the
the same
share
place is
same name,
is all e xis t ing locations,
universe o f discourse
if
of t he
is
location
and the
then to see
to
logically
intersect the two sets.
This same operation is accomplished almost literally when
a designer overlays
one map on another and makes
he is interested in at
of the overlaps between the categories
the
moment.
The
interpretations other
stops
and
visually
a third map
have
empirical
than being the "coincidence
of transit
new
may
category
strategic points",
for
example,
those
locations may be "ideal spots for public information centers".
That shift
realm
of
of meaning
thought.
connects the
The
new category
functions which
do
that
to another
kind
of
CHAPTER III:
FUNCTIONAL ANALYSIS
connecting involve a
PAGE 47.
rule of inference and are
dealt with in
the next part of this chapter.
Observation
regarding
practice, almost avoid
are.
location
can,
in
the issue of precisely
conventional
"where" things
The reason, perhaps, is that because the map is both the
medium of
study and of
communication to
never or seldom has to be
answer
to "where
point."
There
is
it?"
is
"there, on
are, of course,
as well as by street, street
"stone fence",
"location"
described in any other medium.
the
map, at
references to areas
The
that
by ward,
renewal area, for examples,
precinct, census tract, and urban
tree",
others;
corner, and address, and by "oak
"stream
bed"
or "abandonned
loggi ng
road" to mention just a few fixed and not-so-fixed elements in
the environment.
Location in
relation to other
locations rather
than in
relation to some larger geographic reference system, designers
are quite explicit about.
number
of
connectedness,
Distance
occurrences
continuity;
per
away, nearest of several,
unit
area;
adjacency,
direction, orientation,
and
and
relative position in three dimensions plus functions which can
be written on them computing,
example,
for
the rate of change
of any of the elements per unit distance are some of the major
kinds
of locational
manipulations
which
can be
explicitly
formulated.
The designer may have numbers
with
his map
description. The
and other names associated
numbers may
take ordinal
or
FUNCTIONAL ANALYSIS
CHAPTER III:
cardinal
values,
manipulations:
He
manipulations.
an
values
ordinal
cardinal
for
and
magnitude;
implying
each
for
PAGE 48
allowable
values,
"are there
can ask,
residents in this neighborhood?"
of
relations
ranking,
4
set
the
of
arithmetic
than
more employees
can record the answer or
He
not;
this sort of question like tho se above derives more data
from
the description
through manipiulation
information ; it
itself add new empirical
but
does not
by
does not "extend"
the
description.
Extension
The
designer
has
he
as
ways
different
describes
description which informs
the
category of
him along the lines
of the
as
in
at
of h is
but does not
is the
description then it
many
The
it.
hen
its usefulness. W1
the description extends
existing data
looking
of
ways
information because of
is design
environment
inq uiry
a new
change the
result of
what is called here a function of "extension".
A function
data of
of "extension" receives as
the description and
arguments "basic"
information about
some "system"
related to the environment embodied in a rule of inference. It
is characteristic of functions of "extension" that they can be
empirically tested, at least in principle.
For example,
"playgrounds
easy to
environment related
to be
a description
get
to by
of an
environment might
be
ages 6-9".
The
children
system from which information
drawn would be the
behavior of children of
would have
those ages
CHAPTER III:
FUNCTIONAL ANALYSIS
with respect
to moving
rule of
which
variable would
residence.
the
about their
inference would
a playground
was
suffice:
Then the
state the
conditions under
easy to
get
Assume that
distance
to.
of 0-500
feet from
would be
easy to
given to
one
their
would be
playgrounds.
name "playgrounds
6-9" and
The
have to
residences and
be the
children ages
local environment.
measurement of the environment
distances between
category would
PAGE 49
The
get to
new
by
those playgrounds
which satisfied the condition, "within 500 feet of a residence
of children ages
6-9".
could be verified,
This new category
of the description
possibly, by well designed
interviews and
observation procedures.
The
discussion
organized under
thinking
functions
several different
of
"extension"
ways the
environment: its
about the
development
of
be
designer has
of
relation
its perception
over time;
will
and
to use;
its
and
form;
its
ordering patterns.
Environmental Use.
designer may need to
The
know about
soil bearing capacities in an area.
He may have access to the
detailed data, he needs to connect
them to something he wants
to
do
they
before
capacities could,
then,
become
be related to what kinds of activities
or construction might be placed on
bearing
capacity
locations
furthermore,
of
were to
tall
that the
Bearing
information.
design
be
buildings
map of
the land. Suppose that the
of
sole determinant
the
in
bearing
some
city.
the
Suppose,
capacities is
highly
F
CHAPTER III:
FUNCTIONAL ANALYSIS
accurate as
field.
PAGE 50
verified repeatedly
by new
Use of locally available
measurements in
materials and labor prohibit
other than conventional foundation types.
bearing
capacities
permissible
above
some
tall building
the
Then a map showing
amount would
locations and
a
be
a
map
solution to
of
the
problem of where to put them.
If the map
assuming
were also to be used for
that the
bearing capacity
design determinant
The map
or above
at
category
of the
this
the designer
bearing capacity.
of soils having
new
value
desc ription which
was
critical
still the
the variable,
dictated by the placement
capacity
was
(a doubtful assumption) then
another va lue of
would like
low buildings, and,
a bearing
would become
a
new
isomorphic with
the
"allowable location of low buildings".
The designer migh t want to see what the implications were
of assuming
that he
had a way
of treating
the soil
or the
foundations to permit tall buildings on soils presently having
too low
a bearing cap acity.
He
could then see
what pattern
emerged for areas for tall buildings by mapping the areas with
lower bearing capaciti es.
why
one of
the
new
There might be enough other reasons
maps was
to
desirable
recommending that a br eakthrough be
make it
worth
sought which would permit
the lower bearing capa city soils to be used.
In arriving
connected
at
two realms
his recommendations
of thought:
human purpose of city building.
the
designer
has
environment and
the
the
He has isolated (by a process
FUNCTIONAL ANALYSIS
CHAPTER III:
not
single
the
discussed)
a range
has specified
the
in
variable
of values
environment,
rule of inference
in city building, a
"bearing capacity" and
which
PAGE 51
capacity
of bearing
which will accommodate specified values of buildings.
Soils bear
a close
relationship to
climate in
like the
characteristics
other environmental
the
they exhibit
that
are imperfectly known and,
effects of natural processes which
therefore, call for expert opinion.
Although one can measure
climate -
its temperature, humidity,
the effects of the local
wind velocities and direction, one cannot always estimate with
great
those same
accuracy what
will be
measurements
under
different circumstances. Nevertheless there are rules of thumb
which help
out the
to fill
buildings are
descriptions of
have high
likely to
tall
an area:
their bases
winds around
during the winter.
Other rules for the use of the natural environment can be
the
indicating
found
activities, for the use of land
land development based
land
of
suitability
various
for
in highway corridors, and for
on sewerage potential of
the soil and
other factors relating more directly to health. (3.04)
Another type
space or
of rule
vice versa.
If a
links human
activity to
designer needed to
enclosed
find enclosed
places for meetings of 400 to 1000 people, he might look for a
drawing
"meeting
showing
drawing showed too
could
state
(or
places"
but,
few, he might have to
look
up)
criteria
that
assuming
look elsewhere.
for
a
the
He
space's
the task:
appropria teness to
not
areas of
PAGE 52
FUNCTI ONAL ANALYSI
CHAPTER I I1:
than 3500 square
less
than 20 feet;
clear spans
feet;
some district;
have such
order to be able to construct his new category.
mentioned data
above
of
the
description are
measurements themselves,
the
make
available
use
places
meeting
the
as data
to
the category "where
environment in
miss ing
further
from
they have
them
has
might go, he
where the meetings
determination of
Many of these
infer
designer
the
Once
descriptions.
or
and use
standards in
city designe rs;
descriptions normally available to
to
and many
a space
a set of
the environment and such
inventory of
times
e ither the
the space to
private establishm ent;
need to
He would
other factors.
code s;
location in th e city to
owner of
a cooperative
government or
not less
free to compare with the times
needed for the meetings in question;
within
least 40';
hei ghts
to sa tisfy local
means of egress
at which those spaces would be
be
of at
from
made
then might
describe
mass transi t
a
the
stops might
go" .
Other
space-use
turnover
of
measurements helpful
environment;
pupils
sales
in making
per
committee in the country;
others.
employees
per square
foot
an economic
school
facilities
by
and
been
have
standards
use-space
recreation facilities
developed for
foot,
or
per square
other
many
analysis of
almost
every
the
school
use standards, and many
(3.05)
Another
type of
explicit
formulation of
environmental
PAGE 53
FUNCTIONAL ANALYSIS
CHAPTER II
traffic
relationsh
and
transportation
models,
other ac tivities an d environmenta 11
activities to
upon human t rip behavior
is heavily dependent
links
factors and
usually either
expressed as an inverse ratio of d istance raised to some power
from home
to work
curve to represent
uses
or
behavior. And
based on
facilit es
establishe ,d mathematical
to an
exi ting trip
relationship i
type of
land
itted
or
and
on
based
is
standards feor
which iimplicitly
are based on
characteristics.
And there
the
have been
time.
cannot be verified
In both
time to see
the sense
would
what
The form of the
testing is
a
way of
use"
seeing
the predicted data, on the
because they refer
cases, the experts are generating
means of a body of
explicitly
for the "environmental
whether the new data are accurate;
other hand,
efforts to
consi dering.
prediction function is similar to
field
highways
and comfort
Predicti on is used in
happen to the environment he is
the latter,
desig ýn of
( 3.06)
designer's looking forward in
functions:
The highway
diffe rent facto rs which might
influence urban locational patterns.
Development Over Time.
nuisance
T"hese standards
hufman performanc :e
formulate more complex mixes of
of the
the
human comfiort and safety .
are intended to insure
have their
proximity of
adjacency or
characteristics of each activity o r facility.
engineers
still another
to future
new data by
generalizations which describe some system
they know about.
The "system of environmental change" can be very complex.
FUNCTIONAL ANALYSIS
CHAPTER III:
one
Taking
a
example:
PAGE 54
to
decision
developer's
build
a
commercial office building will be, very likely, a function of
he has
price which
the
buildi ng site,
for the
to pay
the
interest rates, the zoning regulations pertaining to the site,
the marke t
schedule for building the
plans and
others'
a nd
city's
type and
space, both
for office
for
plans
amount, others'
types
other
of
space
and
His decision
many other factors.
facilitie s in the area, and
space, the
same type of
to reques t a zoning variance may depend upon whether he thinks
he
can
make
an
considering the
rents he
potenti
development costs,
and upo
from
resistance
people
of
prices, and
those
kinds of developmen t woul
he
When
does that
the
zoning
plus ot her
to charge
and
building,
other
or
his
lot of
groups
and
e city council.
to
repr esent the developer,
mi ght
zoni ng,
land
mentio ned
hen do
so on) and
under
he anticipates a
the ar ea
i
The designer, in order
factors like
it
the
whether
acceptance on the parts of
map
of
11 be a ble
producing
income
"go"
economic
tak
take
final
map,
above (the
a map showing
where cer tain
place under tho se conditi ons.
the designer
simulates
the
decision function of the developers.
Perception and
of the
natural and
Form. The designer records
the qualities
It is
usually these
man-made landscape.
qualities and the processes which are associated with changing
these
qualities with
which
the
designer is
most
directly
concerned. If the landscape were already described in terms of
CHAPTER III:
FUNCTIONAL ANALYSIS
and direction of slope
its slope
dimensional nat ural
them.
for finding
landscape s,
in great detail
forms might be
The
then three
computed by means of rules
c ategories "ridge",
"military crest",
persist because they are,
"isolated hill" , and even "hil I"
most
PAGE 55
identifiable. There
are
oth er
in
landscapes
where those cat egories are not a S read ily perceived as concave
as "saddles",
shapes such
The
the land
qualities of
would
of an d
designer's knowledge
h im
guide
sen sitivity
h ow people
and to
most
appropr iate
set
"description".
The
however, might
well be sufficie ntlv d
that
rules
desi gner would
the
and "bowls".
"val 1 eys", "passes",
used
be
to
visual
land
perceive the
categories
of
btain
forced
to the
the
of
categories,
fficult
to
externalize
to make
the
judgments
himself in the field rather than infer ring them from such data
as slope and di rection of slope.
In the obj ect world the nat ure of
permits an overlaying of
The envi ronmental
other ki nds of
be conveyed
forms and bui iding of form upon form.
medium both
al ows
when the perceiver is
non-verbal permits
similar messages to
in a variety
of different
it permits economical phra sing and a
distinctive grammar where,
for example, the suj ect
object, and the predicate words.
components of
and
verba
overlapping and it
positions and sequences;
The
the vocabulary of forms
can be an
(3.07)
experience
which
func tion as
si gn,
object or symbol may be the same, functioning fi rst as one and
then as the
other, action and conception
fused together i nto
the
probably highly
difficult to
meaning are
process
the
there
description
may b e
CIould
whi
are
sensitivity to
be
a
form of the
on the designer for both
and
"perception and f orm" expl i cit. One
rule of inference for the b ehavior
by
manipulations.
the
lirni ted
played
the
part of
on the
the connotations
environment, it is necessary to rely
Neverthel ess
specify, and
aspects of the perception and
perceiver. In most
of
the
to the
variable with respect
knowledge of
data
in
to
which contribute
environment
the
data of
communication of
the
adventure
an
(3.08)
understanding".
form and
life
human
makes
"that
experience
The
PAGE 56
FUNCTIONAL ANALYSIS
CHAPTER III:
role
in
making
the
design
functions
of
might try to substitute a
of the subj ect which would
produce the same response.
For example, if
a des igner wanted to
in his "corridor" zone which were
of one of the maj or streets
percei ved as
generally
identify the
compute his
the most
fol lowing dat a
new category:
of the building;
walking
whom
exposed,
centrality to their view; and
groups:
from the
Once having
he
might
form in
to
the local
against the background, proximity
of signs, surface
number of people
of
duration
viewer t ype,
newcomers, local employees
by. (3.09)
forms"
environment needed
size of bu ilding, presence
and texture
to
"dominant
singularity of
context, clarity of outline
to observer and
know which aspects
driving, and
exposure
and
the
perhaps only three
a nid shoppers, and passers
classified the
environment in
those
PAGE 57
FUNCTIONAL ANALYSIS
CHAPTER III:
categories, the designer would have to complement the existing
field observation by trained
description where necessary with
computing some categories from
designers and, where possible,
might
example,
for
outline,
of
clarity
data:
other
be
measured by projecting lines from street level at the edges of
their
of
hypothesis
"esthetic" kind
A
response.
and
personal
seemingly
be made
in part,
can, at least
of judgment
verify his
groups, to
of the different
interviewing samples
try, through
And he could
three groups identified.
for the
with other
would be dominant
could evaluate which forms
He then
forms.
then intersected
if they
to determine
the forms
explicit.
Order
"Functions of order" differ from "functions of extension"
function to the designer is the
in that the importance of the
organizational scheme
than
the specific
which
example
may
which it provides
data which
to
help
sometimes arrange
designers
are
for the
part of
clarify
elements of
this
data rather
An
the scheme.
idea
follows;
the environment
in
linear patterns, densely clustered along the line with density
off
falling
away
extending out normal
following
the lines
expressed as
description
might
fall
line
and
to the major line
This
out.
a function.
which come
into
the
from
such
a
to
pattern.
lines
smaller
with higher densities
particular pattern
Its arguments
even close
with
can
of the
are any data
suggesting that
Its
value
be
is
they
the
PAGE 58
FUNCTIONAL ANALYSIS
CHAPTER IiI:
specification of data not present
specified pattern.
gaps of the
to fill in the
data required
in the existing description,
(3.10)
cited the designer
form example just
when applied to
things.
to put
provide places
they
design:
heuri sti c
powerful
order" have a
"Functions of
In
fi 11 it
in
the linear
a rule which
introduced
open questions
the e xisting environment left
as to what would
role
points and what the
in at the empty
empirical correlat es w'ould be of its various parts.
intersections
a re
links
a nd
connecting
organize
often
des igners
st rongly geome tric pattern, a
environment in a
the
u rban
example
For
"neighborhoods", a nd
someti mes
left
called
each membe r of
grid perha ps:
"centers",
the
"roads"
and
areas,
over
the
roads, and
the centers,
neighborhoods is arran ged in a h ierarchy corresponding to s ome
"level"
movement, or activity. (3.11)
of importance,
addition,
the grid
i tself
is assumed
to
rectilinear structural "framework" for the
which is supposed
to afford flexibility of
be
a strong
a good
and
growth over time.
are helpful
conventions.
the designer's
many of
in
growth of the city
Admittedly somewhat out of date, these "megaforms"
in illustrating
Often,
(3.12)
Other
bes i de the
are the
and,
organizing schemes
for the
grid and the linear
radial, circumferentia 1,
analogies
with
trees,
form wi th arms
star,
witt h
environment
man made
extending out
galaxy,
organisms
and network;
and
with
PAGE 59
FUNCTIONAL ANALYSIS
CHAPTER III:
rivers.(3.13) Some elements which bespeak their formal origins
the conventional vocabulary of
have become permanent parts of
radials,
belts,
green
form:
city
and
circumferentials,
satellite centers.
The designer, as he works with these "functions of order"
may tug and pull against them resisting the full extent of the
on him by making ad
regularity they force
them, some
to be
some local building
explicable in terms of
important natural
some particularly
saved, or
hoc adjustments to
feature which
may cause a bend or even a break in the pattern, some in terms
of the
inexplicable either to the
pattern, and others
into the
to introduce variety
designer's wish simply
analyst or to
the designer. (3.14)
"functions of
Some
the visual form and
event and
of each
their or der and
perceived) play a major role
each event
empirical
instead
on
on patterns of
the communicated mean ing
timing of
occurrence (as
in determinining the strength
success of
and the
based
of being
plan view, may be based
geometric patterns in
events in which
order"
the pattern
01
whole. The
as a
correlates of the elements of these patterns may be
the doorways, paving, signs, and lights of the designer of the
public or popular world. (3.15)
Other "functions
single class:
small,
for example,
medium and
activities and
of order"
large,
"role"
in
include "seriation"
three sub-categories
each
with its
the city'.
withi n a
of cente rs:
own composition
Dichotomization
is
of
the
FUNCTIONAL ANALYSIS
CHAPTER III:
division
into
two
often
PAGE 60
opposed
cl asses,
accompanie d by the establishment of
oppo sites:
the
"communi ty " and "privacy",
the descri ption as
to
expected
play
organizati on, the
for some
group.
or
new class
to
create
connection and
are
environment,
new
ogical
designer
possibilities which
description
of the
of
hypotheses. A
his
the
the
other
n creates gaps
And diagr am s
to be challenged.
clustering,
are
bubble diagrams,
to
to fit
ex isting
the
fit
other "functio ns
some
its specification.
of
(3. 17)
FUNCTIONS OF TRANSFORMATION
engages in
descriptions
he looks at
and the ways of
design.
He
may
design alternative is a
existing description of
of
hypoth et i cal
wi th the
and examines along
environment
existing
possibilities for change
purposes
part of
Like
organizatio
order" more strongly geometric in
The
datum.
distorted
then
or, perhaps,
IfI.
some existing
studying essential relati onships in the
often used as ways of
which
serves to
the intersections of each of
to
to be filled and regula rities
design
"Morpho- logi cal"
to afford the possibility either
"functions of order", m orpho-
exhibiting
(3.16)
data of
which it is
cross classification of classes,
assign ing that
often
between
gradations
well as with some social role
the classe s in the table and
environmen t
five
each associated with some
draw the designer's attention to
of
is
a scale of values between
example,
for
and
ou t
the
satisfying th e
m any
to study
investigate
many
set of changes
the environment in which
such
to the
the cha nges
CHAPTER III:
FUNCTIONAL ANALYSIS
are mutually
consistent
A
PAGE 61
design alternative
the designers
the isolated hypotheses tried out by
can
be analysed
and
new
in that it
description of
the
Moreover, it presumably achieves some entire set
environment.
of purposes
an entire
as
differs from
which has
been formulated
to solve
some design
prob em.
The city designer
control over the course of
which he can exercise on ly partial
events.
The world gets designed and built even if he does not
intervene.
He
often
can
i ntervene
preconditions upon which the
has
an environment in
most often works on
altering
the
env rironment related "systems " he
or
operate ;
envisioned
by
by
altering
the
rules
w hich
describe how the system he envisi ons operates.
A function of "tran sformatio in" receives as arguments, the
"primitive"
data
of
the
env ironmental
desdription
and
functions of "further descriptionI". When evaluated, it yi elds
a
changed description.
"primitive" data or
It
does
so by
altering either
the rules of inference
the
in the "functions
of extension".
Transformingf the
as looking at
Data. The designer can
a world which is best
imagine himself
described by envisioning
"systems of environmental change".
For example, the designer
might change the "environment"
within which a developer acts in such
a way as to change what
he does and where he does it but not to change the criteria by
which he evaluates the choices open to him.
CHAPTER III:
The
FUNCTIONAL ANALYSIS
designer
PAGE 62
intervening
is
alter
to
system in
environmental
conditions upon
which the
operates.
assumes that he
can control
He
specifies the changes, simulates the
the changed
environment and
of
some
the
question
these conditions,
action of the systems on
then might check
to see
if the
environment so obtained is satisfactory.
A variation on this theme would
be if the designer chose
to influence the environment related
constraints on the agent
industrial development through
zoning ordinances to encourage
available for
the areas
increasing
lower income people
as these actions
activities,
by increasing the lot
as
discriminatory behavior
or to
such use
exclude
size, as misguided
combination with other
may be they have, in
such
changes to the
At the local level,
of environmental change.
active
promotion
on the
part of
industry
of
real estate
or
agents,
achieved their stated purposes.
Changing
developer's
the
which he
environment within
actions
acts is
basic
transforming the
data of the function of "development over time".
could also alter
the
changing
by
the basic data of other
The designer
types of "functions
of extension": "functions of environmental use" and "functions
of perception and form". Illustrating
example cited above
interest
in
"form
the latter type:
guided to his
the designer may have been
dominance"
attention of as much of the
by
wishing
in the
to
direct
the
population as possible to certain
kinds of facilities on the premice that "form dominance" would
CHAPTER III:
FUNCTIONAL ANALYSIS
increase awareness
and use
PAGE 63
of those
facilities.
He
could,
perceived the environment, try to
with his rule of how people
rearrange the environment to achieve his goal.
Transforming the
designer can
the
designer
have
could
the "system"
rules of perception.
and,
A second
the rule
to
the
representing by
the
have trained people to observe
He might
their
changed
hence,
of inference
change
chosen
he was
mode the
In the previous
basic data of the functions.
characteristics of
better
is to change
operate in
instead of the
example,
Rules _of Inference.
of
perceptions
the
environment. Inasmuchas the perceived "form dominance" is part
of the description of the environment, the designer would have
transformed the description.
A second example of this mode of work might be the design
of a new type of developer which would respond to the needs of
low income residents without a
For
example,
in
houses
settlement
have
into
gotten
existing institution with motives
development, an
well
very high financial incentive.
housing
which seem
suited to the purpose but which previously did not engage
that sort
of
activity.
Cooperatives
composed of the affected individuals,
existing model,
for the
same reasons.
been set
have
up
a new institution on an
In these
cases, the
designer tries to bring together a group whose motivations are
in line with the designer's convictions about how things ought
to
be done.
developer, the
Once
again,
new group
as in
may be
the
case
described as
of the
private
acting freely
FUNCTIONAL ANALYS IS
CHAPTER III:
area (3.18),
some opportunity
within
existing
for
he
after
change, he
environmental
of roads and
land ownership,
and existing
im portant
operating in
the new
which
system
for
has
imagi ned
some
will
probab ly
want to
might change the environment or
mode of
environment.
the
particula rly
are
considerations
perhaps ,
order to
be
locations
to
pertaining
legislation
designer,
patterns of
which may
an area
considerations as the
bounded by such
utilities
PAGE 64
after
the
city
agent
of
see how
it
new
1
revert to the first
conditions within
change the
acts, and,
These
that,
see how
the
changed environment affect the quality of life in a variety of
different ways.
Thirdly,
the
wisdom about
a geographic
health,
safety,
based o n
development rules
environmental
experience as
comfort,
build
and
other
tall buildings.
associated with
the
of
"system", one which would provide
a
of
expert
which pertains
dimensions
of
to
human
and political
An example of the latter
of development
institutions which could enforce
invention
set
where to
recommendation about
set
This
a
body of
well as technological, economic,
the soils engineer's
constitute
a
region, wisdom
feasibility of environmental change.
would be
es tab l ish
could
designer
new
rules,
if
them, would
environment
related
new constraints on what the
other agents of environmental change actually did.
III.
During
the design
FUNCTIONS OF TESTING
process
the
designer is
constantly
neither
the
alternative
comprehensive
special conceptual
activities pose
These
alternatives.
and
with elsewhere.
are dealt
many of which
procedural problems
many
of
evaluation
comparative
the
nor
design
single
a
of
evaluation
is
to here
is referred
"testing" which
criteri a. The
many
explicitly, against
implicitly or
ideas, either
testing his
PAGE 65
FUNCTIONAL ANALYSIS
CHAPTER III:
(3.19) "Testi ng", as used here, refers to what the designer is
doing as he i s working toward a design alternative.
three kinds of arguments:
A "testi ng" functions receives
of the
data
other sources ;
o
true
"testin.g" fu
descri ption;
sources;
is,
he
the value of
they match within the standard,
sources and, if
function
with the data from other
compares the data of the description
the "test"
The
comparison.
standard of
and a
from
data obtained
description;
environmental
or
,
or
can
he
To
change the
change
the
data
t
e
of
the
other
comparison, that
two types
between the
o
from
obtained
data
the standard of
the "fit"
value
the
transform
can
he can change
can relax
.
designer
the
:tion
passed
of data.
(3.20)
"Testing" functions
internal
Internal
consistency
(eg. that
acres of house
data of
refers
description:
environmental
and
consistency
types:
environment
in city design
tc
the
eit :her
there were
can be of
external
no more
lots in one hundrE!d acres of
the description and
rulEýs of
consistency.
consistency
among
two basic
data
than one
of
of
the
the
hundred
land) or between
inference representing
that
(eg.
related "systems",
environment
were
the shadows
and
the longitude
position considering
the proper
drawn in
PAGE 66
FUNCTIONAL ANALYSIS
CHAPTER III:
latitude of the environment and the time of day and year being
recreational
engaged in
people
these
range for
acceptable
at this place is within an
that the noise index
(eg.
the specification of
of the description and
between the data
purpose,
to the consistency
External consistency refers
represented).
activities).
of
"description"
of
type;
highway, and vehicle
function
compares
then
"desired value"
source which
measured
expresses some group's
environment matches
If the
The
noise
the
The "testing"
index
a value obtained
tolerance of
with
a
from a
noise when
measured noise index in the
value" within
the "desired
test is passed.
of the
convert
could
noise index.
of the noise index,
engaged in some activity.
then the
the
of
kinds
three
a
environment related
and from the
environmental measurements into a
being
obtained from
received
which
coeficients which
acoustics:
of
turn
was in
environment: highway position, use
arguments from the
system
index
The noise
considered.
the environment are
the environment
in
index
a noise
of
the values
function
The measurements of
example.
used as an
be
noise will
tolerance of
and people's
Highway noise
the standard
"desired value" could
as well
refer to the designer's preferred level or the average of many
groups. The values and standards
in other
ways.
And
for comparison can be stated
the "desired
value"
may
change
with
the new environment,
exposure
able to stand more (or
a
for example,
embodie d
or
people
may be
less ) noise after they have lived there
consid erati on could
while. This
example,
PAGE 67
FUNCTIONAL ANALYSIS
CHAPTER III:
n
be built
into the
above
which
might
example
anothe r
demonstrate the long term adaptation effects. (3.21)
The "funct ions" o f cit y design as spelled out here are of
three types:
or
infer
" functio r
data
new
transformation"
directly, or
and
which
i ndi rect 1
"functions of
further description"
ot
the
Iter
desc ription;
the
by altering
testing"
data
of
"functions
the
the rules
which compare
which derive
of
description
of inference;
the
description with data obtained from other sources.
data of
the
PAGE 68
CHAPTER IV
THE DISCOURSE LANGUAGE
The
language of
DISCOURSE will
be
presented in
three
sections: the first dealing with LOCATIONS and ATTRIBUTES, the
basic
the
of
features
second,
the
language;
the
with
elaboration of the data structure to include numerical values;
and the
third a
short description of
of the
the components
DISCOURSE system.
of DISCOURSE has
The structure
as possible
transparent
clear understanding
and
in order
chooses.
The
that
operations have been
it, and
are
world as
the
developed to be
he
close to
event, though he may not always
what the designer does in any
realize
to depict
able
a
and changing
is manipulating
of what he
he be
designer have
that the
in order
simple and
been made as
be so
intended to
avoid
as to
basic
commitment of the designer to an uncomfortable or unwanted way
of working.
The
basic
features
accommodate the types
functions
chapter:
with values
the
language
of the
derivation and
to
the previous
inference of
data, and the
obtained from
it
allow
of functions developed in
transformation of the
data, the
the data
of
new
comparison of
other sources
which may
represent standards or norms.
As a
analysing,
designer works he creates
and
of
generating
new ways of
environmental
working, of
forms.
The
CHAPTER IV:
PAGE 69
LANGUAGE
of a design exercise may
techniques he uses toward the finish
No
be entirely different than those he used at the beginning.
want
to
make.
features
which
to
the designer
permit
special requests, allowing him to build
special
his
to make
requests.
and to modify his own
as he
But he
must program.
still
appears
There
that the
requests which respond
better to his needs,
to be
order to
designer, in
unresolvable dilemma:
an
make
will demand more
he will have
language and, as a consequence,
features of the
his
eases
DISCOURSE
in
much of the bookkeeping for
programming job by taking care of
him.
own
his
The designer must, therefore, do some "programming"
works.
order
construct
design proposals
well as his
design as
techniques of
additional
has
therefore,
language,
The
will
a designer
the requests
anticipate all
programmer can
to become proficient in an increasingly complex language. Ease
of use and subtlety of expression
is taken up in Chapter VI,
features
LOCATIONS)
(or
and
data structure
of the
limitations of the
of
on future development.
BASIC STRUCTURE OF THE LANGUAGE
I.
Places
This issue
may be opposed.
structure will
ATTRIBUTES
explored later
the
Some of
of DISCOURSE.
system which have resulted
be
are
the
from this kind
this chapter,
in
basic
but
principally in Chapter VI.
usually mean
LOCATIONS
space which can
that space.
a
position
in some
be distinguished from all
The space is
two dimensional;
geographic
other positions in
two coordinates
PAGE 70
LANGUAGE
CHAPTER IV:
LOCATION uniquely. It should be
are necessary to describe any
single dimensional array.
space to a
the two
that
coordinate addressing of
three
dimens ional
or
data structure
the
list 'I
information
cannot deduce from
i n DISCOURSE
with the
is inconsisten
nconsistent with its
particular
the mom ent.
can be thought of as the name for a
c oordinate s
A pair of
One
the LOCATIONS
pace
is
rather it
;
impl ementation a t
LOCATION
N dimensional
computer ultimately reduces any
noted that the
abo ut
mem bers
of
the LOCATION
is
at
that
what
Coordinates refe r to an a rea rather
list
particu lar
contain
LOCATION.
the
than a sing le point;
LOCATION may be the size of a city block or, as in the example
II,
in Chapter
ATTRIBUTE at
a LOCATION
within
LOCATION.
that
The presence or absence
a square mile.
Thus the
"resolving
designer's lense is limited in major
LOCATION he uses .
Other ways of
to its
clue as
gives no
of an
placement
power"
of
the
part by the size of each
increasing the detail of the
description will be discussed below.
ATTRIBUTES
An ATTRIBUTE
an ATTRIBUTE
is descriptive of LOCATIONS.
is the name of
In DISCOURSE
a list of LOCATIONS
which share
it. Examples of ATTRIBUTE names are "water", "tidal flats" and
"middle
income residents".
LOCation is called
An ATTribute
which
eXists at
a
be thought of
an "ATTXLOC". Most maps can
as a display of ATTXLOCS.
The
data structure
of
DISCOURSE
I is
a
particularly
CHAPTER IV:
PAGE 71
LANGUAGE
eloquent representation of the logical structure of ATTRIBUTES
and
figure
LOCATIONS,
created:
for
one
(4.0 1).
LOCATIONS
LOCATION consisted of its two
Two
in which
"dictionaries"
were
the
each
coordinates;
name
for
and the other for
ATTRIBUTES.
Organization of
data by
l ists means
organization of the data ins ide
the logical organization of
geographic location
that the
physical
the computer is divorced from
One logical record (the
the data.
of indu stry , for
example) is
located in
computer storage by storage allo cation features which are part
of the computer language in
(4.01)
whi ch DISCOURSE I was programmed.
In order to find that rec ord again, it is pointed at by
the previous member
of the same list
and it,
in turn, points
to the next member of the li st. The values of the pointers are
determined once the storage Ioca tion has been determined.
organization the
With list
shortened
havir ig had
to
lists can
predic t only
the
be lengthened
or
total amount
of
working space re!quired for all lists plus expansion instead of
having had to re!serve a specific
the expansion
airea of any
place in computer storage for
parti cular list.
For
problems in
which the data aire seldom changed or come in predictably sized
feature is
packages, this
meani ngless.
design where thE designer is addi ng,
his data constaritly, it is
has disadvantage!s as
organization
in
well.
computer
But
in problems
of
subtracting, and changing
signi ficant. But list
organization
The pointers which
permit random
storage take
up space
themselves;
PAGE 72
LANGUAGE
CHAPTER IV:
time is required for moving around through storage which is in
computer's core and
part in
has to be constantly
used as
in the
it, leaving part
memory device if it
some other
it is expensive
each list is continuous
records; and because
to cut
sequentially organized
that required for searching
excess of
part of the search procedure, for it would have to be read in
core.
of the computer's
and out
This
consideration becomes
crucial for problems in which such search in the categories of
the description is
Both
be extensive.
to
are likely
data
in which the
commonplace and for problems
are
city
true of
design.
list
The
organization
contrasted with an
be
logical records
phone book.
like a
in sequence,
must
storage
computer
organization in which the
placed physically
are
of
One
common attribute
of the records is
selected by which
to key
them in sequence;
in the case of
the phone book, it
is the
name.
If
address
one were to look
it
would
be
number or
for a particular phone
difficult. If one
lengthened either all subsequent or
record
is
to
be
all previous records must
be moved to leave space.
These are limitations to sequential organization when the
problems to be dealt with
as
well
as
continuous
involve records of changing lengths
modification
of
the
records.
But
sequentially organized records can be stored redundantly, that
is, one could
name; a
make three phone books:
second ordered
one alphabetically by
by telephone number;
and a
third by
CHAPTER IV:
PAGE 73
LANGUAGE
the same
for the operation of either the
overhead which one has to pay
list or sequential organization will
depend upon many factors
of the
the core-resident memory
are the size of
The
multiple list structures.
advantages of record access as
among which
many of
Redundant storage affords
street address.
computer plus the amount of the data that the programmers must
of logical structure of
software being used. The transparency
version of
in the first
list organization utilized
the pure
and
hard-
the particular
of
features
and the
with,
deal
DISCOURSE gave way to a structure which took advantage of both
list and array features of the computer. (4.02)
In
the
version of
first
ATTRIBUTE names
were associa ted
wi th
"permanent extens ion"
"initial",
The paradigm
typ es of
three
and "temporary
on the
of behavior
LOCATION
the
DISCOURSE
part of
and
lists:
extensi on" .
the designer
this
earlier arrangement of lists sugges ted was roughly as follows:
a designer makes an "initial"
desc ription of the environment,
indicating where things are;
d uri n g the course of his work he
looks
thin gs are ,
where
around at
combination, creating categor ies of
of
"temporary"
"permanent"
deleting)
to
use
change
of the
a category
him
of the
The
designer
looking around and
in
they are
where
the description which are
a nd,
he
finally,
descrip tion either
LOCATIONS or by creating a new
LOCATION.
and
desc ription to
makes
by adding
(or from)
a
(or
some
cat egory and adding it to some
persists in
this
then changing things unt il
way,
repeatedly
he reaches
the
CHAPTER IV:
end of
LANGUAGE
PAGE 74
his working session. At
that point he stores
what he
has done. At the next working sessions he again starts with an
initial
this time,
description, but,
it can
the
of the previous (or any
or the terminal description
"initial"
be either
other) working session.
In the latest version of DISCOURSE there is only one type
of list
but the designer
can
create his own
variety through
the ways in whi h he names and manipulates them.
ine trees, not
good drainage,
by
"where are the
ike
Questions
too steep, ... "
on ATTR IBUTES
logical ope ations
goo d view,
places with a
and
can be a nswered
A
LOCATIONS.
category of the description is a new ATTRI BUTE lis t;
new
a desi gn
change is an alte ration of the ATTRIBUTES of LOCAT IONS.
ATTRIBUT E lists exi st which
In DISCOURSE two
to
in
the designer
various
ways:
Universe is
an ATTRIBUTE which
null exists
at no LOCATIONS.
less
obviously
useful
description
whic h
LOCATIONS.
(Eg.
operators of
(4.10).
at
"new
for
that
LOCAT ION and
Universe is essent ial
(Eg. "dry land *=* unive rse
water.
"null".
and
exists at every
fi nding all LOCA TIONS
operations as
"uni verse"
are useful
exce pt those
keepi ng
min
oment does
popula tion
not.
.and.
*=*
DIS COURSE are illustrated
a
in such
whi ch have
wat er" ) Null
catego ry
not
nul11")
ex ist
The
in figures
is
of
the
at
any
logical
(4.02)
to
CHAPTER IV
PA E 75
•Z
L---
4
o
LU4
z
0
U
C/)
09
3
z
O
0
C
U
0
(ri
"U
tIS
ti·J
0%
C)
·u
14
3
hif
rao
%a,
PAGE 76
CHAPTER IV
map* (a b) 1-7 1-9
12 3 4 56789
. aaa
. bbb
Saat
tbbb
.aattbbb.
.aaa
.
Symbol
Computer generated
ATTRIBUTE number
referring to
ATTRIBUTE "A" (or
"a" in the Figure
titles).
123456789
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS 't'
= a,
in the map
bbb.
-
'a'
= A(2);
'b' = B(3)
b
= NON EMPTY LAND (contains ATTRIBUTES other than "a"
-----------------------------------------------map*
(a
b universe)
1-7 1-9
1234
5 6 7 8 9
0
0
0
0
0
0
0
0 0
B
B
S B
B B
0 0
0C
0 0 0 0
AAA 0
AA TT
A A TT
AAA 0
0 0 0 0
0 0 0 0
or "b" )
0
B
B
3
0
0
0
0
0
0 0
0 0
123456789
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS - ' = UNIVERSE(1)
'A'
=
a,
'B'
=
b,
0
'T'
=
a,
b,
'a'
= A(2);
'b' = B(3);
'0
0
!j
S.' = NON EMPTY LAND
Figure (4.02)
Positions of ATTRIBUTES
"a"
and "b",
the basic data
------------------------------------
Figure (4.03) "a", "b" and "universe"
In figures (4.03) to (4.10) the desired results are always
in capital letters.
PAGE 77
zt
0
cV)
(I
iU
o
3I
0
u
II
r-4r
F-O
L.
CnL
.0
C -
(o
_O
-•
1
*
4
F-CIO U
*
*
*F-
*
V-
CL"
C
<
4-•
-<w
C14
-
*z0CO
LCC
%
--
•
0-
cN
L..
r L
-
LoU
.
*
N
-
z
<
..j
r).
II
-
S-
.-
0•
>D
L
3
1
-
L
PAGE 78
N-
0(
O
<:
o
0
-o
O
C,
4C-,<
I
-W
a < 41 W<
Or,CI
.o
-
.•
N-
re
*
I*
*, o
I
*o
*O
C
,
* .0
CN
*
.i
z
*
.4 .
-E
*
*
-n
*<<
O
.
0 *m
Or
CI
o
Ln
%
C-C
wo
<<*
L0
H
4-
2-
-
LUO
SH
C
LU
rDE
C
-J
-
OU
Q-
.4J<
---
*
PAGE 79
co
LU
H
"t'
*4
aa
"II
CU
U-1
0
-o
O
I
C
.*0203
L..
4J
*C*
*
*
*
I)
h
-H
.
.«
*
-
-c
(- N
<
Cu
0
u
.
In
CO
C
S0
• -
03
L
.2
C..
III II
II
<
-
CC
I
*
PAGE 80
0000000
01
e4
0
-J
0
I-
(O 0r
a ILl
I.- 4
.V4M
LOnL
,
r-
0
11 -II
Sc.
E
0
---
C0
PAGE 81
S
CN
_-
·
a9
r-
(O
Ln
fY3 r~tl
·
·
·
·
CS
·
·
s~3
·
o
r-4
r3D i~fi
O,
01-
· rsD ~32 p~
C
&
4.J 4j
*
Z
LU-
4- 4-J<
OO
C•
*
•
Of
D
H
H-
r>-
0-
<
r-
u
x
1wE
C
N
-f
Lfn (0
o-
On
L-U
-j
Z
>-
*
*
I
rr
4-J
I II
*
*
II= i!
r
PAGE 82
ri
4
r%-
LA
N-1
Z
0000000
b0
.,c
'.
,,0
.O
.. .0
'~
r..P
0.
O r%
,)
0
C
S'.
ON
o
•
O
O•
O
c co
>-
r 00O
0o
II 0
. 9
*C
O
-z
---I.-.
ccn
r,,
r- 1N 11-tr U-1*
Lo\I-.
CZ
d,
<1l
*... >.
m
I
II
II
0m
)
oc
,
,a-
*
PAGE 83
C
L.
E
0
II
v
-0
O
40O
CD
J
3 to
O9
c'
Co
LU
CZ
CO
H
4-
O
t Coc-
L-
.C C
C,
d
•
C.
C
·
'-4
u
Co
OC
z1
*-rCo
*-4
*
*
*
*
*
*•
**
**
**
*
*o
•
•
*
•
• "
•
LL=
I-.
C
-J
>-
u
Le
CE
z
C
E
C
mI z
o
cm
-- E
r-4C-4 n -: LC%
QO r-~
w
-
<D
-
-2
C_
0.
*--
-,
o
CHAPTER IV:
PAGE 84
LANGUAGE
In the
logical operations described above,
the computer
had to look down the ATTRIBUTE lists member by member in order
to determine, for example, whether or not some ot her ATTRIBUTE
was present at any of the LOCATIONS on the list.
down a list member by member
oper ation may
is called "looping" . The looping
to remember if
be easier
This looking
each ATTRIBUTE list can also be thought
it is
realized that
of as a ring in which
the last member points back to the first.
A second
kind of looping
for geometrical
requ ired because each LOCATION is
two
numbers:
its
coordinates.
lang uage were necessary
functions was
represented in DISCOURSE as
Additional
fea tures of
to permit the designer
the
to manipulate
them (and to manipulate numerical data which will be discussed
below).
Both
arithmetic
and
relational
features
were
required.
The arithmetic operators are:
"+" = plus
"-"
"1*11
= minus
= times
"/" = divided by
".power." = raised to the power
The relational operators are:
".grt."
".geq."
".]eq."
".les."
".eql."
".neq."
=
=
=
=
=
=
Additional objects of
greater than
greater than or equal to
less than or equal to
less than
equal to
not equal to
the description were required
as well:
CHAPTER IV:
PAGE 85
LANGUAGE
which would
temporary variables
store a
number in
order to
The rules
keep track of the results of numerical computation.
specifying the precise
The features
manuals. (4.03)
and "Reference"
the "Notation"
are included in
use of these features
are illustrated in the several examples in this chapter and in
Chapter V.
Three
examples
geometric
of
the
LOCATION.
nearest member
of
are
"nearest" which
ATTRIBUTE list
some
which
to
some
These three have been included as basic requests in
and others which
DISCOURSE. These
which are probably not
want but
"ci rcl e"
LOCATION;
of specifi ed dimension; and
creates a circle
finds
some
size around
a spec ified
which
creates a rectangular block
frequently used are "block" which
of
functions
a particular
designer may
of general use can
still be
written in DISCOURSE.
To il lustrate
LOCATIONS
the feat ures of
and ATTRIBUTES
Commands in DISCOURSE
spaced
except
separated from
where
the system
left margin, single
will appear on the
demands
the preceding and
an
and
space
extra
following text by
space. Commands given by the designer
made.
will be
data file
a simple
clarity
pertaining to
an extra
are in lower case type;
the computer responds in upper case.
Step 1.
LOCATION and,
The designer
therefore, the size
command establishes
columns
selects the size
numbered 1
the grid
to 11,
of the
of the grid.
with rows numbered
space
for 32
map and
The "setup"
1 to
8 and
ATTRIBUTES and
32
CHAPTER IV:
LANGUAGE
PAGE 86
temporary variables.
comment step 1.
setup 1 8 1 11 32 32
He decides upon the initial set of categories of
Step 2.
his description.
comment step 2.
land *=* null
water *=* null
lowres *=* null
midres *=* null
tidal *=* null
highpoint *=*
3.
Step
null
The
ATTRIBUTE
can
be
"1owres"
on
lists
expanded
or
contracted by saying:
put lowres 6,7
delete lowres 6,7
in
order to
put
or delete
all the
high
po i nts:
put lowres for each loc on highpoint
delete lowres for each loc on highpoint
and,
LOCATIONS,
finally, to add or subtract "lowres"
one gives
the
following
corners of the block are identified:
from a block of
commands in which
the
CHAPTER IV:
LANGUAGE
PAGE 87
put lowres(6,7)... lowres(7,9)
delete lowres(6,7)...lowres(7,9)
The following commands pertain to
the data file which
being built:
comment step 3. Locating the ATTRIBUTES
comment or describing the LOCATIONS.
put land(1,1)...land(4,5) land(1,6)... land(1,7)
land(1,8)... land(5,11) land(5,1)...land(8,1)
put water for each loc on uni verse .and.. not. land
put lowres(l,1)...lowres(2,3) lowres(2,3) . .. lowres(3,4)
put midres(2,8)...midres(3, 8) midres(2,9) ... midres(5,11)
put tidal (2,2). .. tidal(4,5)
delete tidal(2,2)
put highpoint 1,6 4,8 5,11 6, 1
comment to store this file for future use:
store mapp p
Step 4.
He may
wish to
display the
check them
for errors and in
Display in
DISCOURSE is of two
display and tabular display.
in tabular form;
comment step 4
show lowres
LOWRES(
LOWRES(
LOWRES(
LOWRES(
LOWRES(
LOWRES(
LOWRES(
data in order to
order to begin
major types at
look around.
present:
map
The command "show" yields values
"map" makes a map-like table.
CHAPTER IV:
PAGE 88
LANGUAGE
LOWRES(3, 3)
LOWRES(3,4)
map (lowres
midres; highpoint land)
1-8 1-11
1
12345678901
- - - -.-...
0 0 0
-1-1-1-1-. . .-2-2-2-2
.1-1-. . .-2-2-2-2
.
. . . . . . .*.-2-2-2
-. . . 0 . 0 0 .-2-2*2
i,
·
·
·
·
·
•
12345678901
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS -
-
'1'
= LOWRES(7);
'2' = MI
DRES(8)
I.,
= NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+'
= HIGHPONT(10);
= LAND(5);
'*'
= LAND & HIGHPONT; ' ' = NO REF ATTS
Step 5.
He could "fill"
the
land where the
'-
low income
residents lived:
comment step 5 altering the data
put land for each loc on tidal .and. lowres
delete tidal for each loc on tidal .and. lowres
Step
6.
He
development subject
might decide
to the
to
assemble a
following criteria:
adjacent to existing residential development;
zone for
new
that it
be
that it
not be
CHAPTER IV:
on water,
that
PAGE 89
LANGUAGE
high point, and
not on a
the LOCATION
on that
closest to
list
flats; and
not on tidal
a highpoint
be
assigned to low income residents.
The sequence of requests might look as follows:
1.
This
"tlistl *=*
creates a
new
LOCATIONS on "lowres"
2.
This
.or.
lowres
midres"
ATTRIBUTE
which includes
all
the
or "midres".
"ex beside (tlistl)"
creates
a
new
ATTRIBUTE
includes all the LOCATIONS which
named
"beside"
which
are beside "tlistl", but not
on "tlistl".
3.
"tlist2 *=* beside .andn. water .and..not. highpoint
.andn. tidal"
This creates
"beside" which
a new ATTRIBUTE
also contain
in which the
LOCATIONS of
"water", "highpoint"
or "tidal"
have been eliminated.
4.
"ex dnearest (highpoint tlist2)"
This finds the
LOCATION on "tlist2" which
any LOCATION on "hignpoint",
its
column coordinate
in
puts
is closest to
its row coordinate in "qi",
"qj", and
its
distance from
the
closest highpoint in "q".
5.
"put lowres qi,qj"
This puts "lowres"
6.
in that LOCATION found by "dnearest".
"show loc qi,qj"
This simply displays to the
LOCATION which has been changed.
designer the contents of the
CHAPTER IV:
LANGUAGE
PAGE 90
If the designer wanted to
he could make them a Request
call
It by
choice open
residence
a name.
when he
he would
COMbination file (RCOM file) and
Additionally, he might want
called that
RCOM file
assign. Therefore,
substituting "argl" for
then read "put
repeat that series of requests
he
"lowres" in line 5
argl qi,qj".
The first named
of what
type of
might rewrite
it
above which would
argument in the
RCOM call
would be substituted for
executed.
The execution of the RCOM file follows:
"argl" when the
comment step 6
oldzone *=*
to leave the
Iowres .or. midres
rd newzone p(lowres)
LOWRES HAS BEEN ASSIGNED TO LOCATION 1,7
AND IS 500 FEET AWAY FROM THE NEAREST HIGHPOINT
rd newzone p(lowres)
LOWRES HAS BEEN ASSIGNED TO LOCATION 1,5
AND IS 500 FEET AWAY FROM THE NEAREST HIGHPOINT
rd newzone p(lowres)
LOWRES HAS BEEN ASSIGNED TO LOCATION 1,8
AND IS 1000 FEET AWAY FROM THE NEAREST HIGHPOINT
rd newzone p(lowres)
LOWRES HAS BEEN ASSIGNED TO LOCATION 1,4
AND IS 1000 FEET AWAY FROM THE NEAREST HIGHPOINT
rd newzone p(lowres.)
LOWRES HAS BEEN ASSIGNED TO LOCATION 3,1
AND IS 1500 FEET AWAY FROM THE NEAREST HIGHPOINT
rd newzone p(lowres)
LOWRES HAS BEEN ASSIGNED TO LOCATION 4,1
AND IS 1000 FEET AWAY FROM THE NEAREST HIGHPOINT
rd newzone p(midres)
MIDRES HAS BEEN ASSIGNED TO LOCATION 5,1
AND IS 500 FEET AWAY FROM THE NEAREST HIGHPOINT
rd newzone p(mldres)
MIDRES HAS BEEN ASSIGNED TO LOCATION 1,9
AND IS 1500 FEET AWAY FROM THE NEAREST HIGHPOINT
rd newzone p(midres)
MIDRES HAS BEEN ASSIGNED TO LOCATION 1,10
AND IS 1802 FEET AWAY FROM THE NEAREST HIGHPOINT
file was
CHAPTER IV:
LANGUAGE
PAGE 91
comment the results of the RCOM file
comment can then be mapped:
newzone *=* lowres .or.
midres .and..not. oldzone
mapset standard (6;+) red
map* (newzone oldzone;highpoint land) 1-8 1-11
1
12345678901
1
-2-2-2-1-1+.-1-1-1-1-.
1
2
-2-2-2-2-.
. .- 2-2-2-2
2
3
-1-.-2-2-.
.- 2-2-2-2
3
4
5
6
-1-.-. .
.
.* .+.-2-2-2
-1 . . . . . . .- 2-2+2
+ . . . . . . . . . .
4
5
6
7
8
-. . . . . . . . . . .
-. . . . . . . . . . .
7
8
5
2
3
4
12345678901
1
6
7
8
9
0
1
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS -
-
'1'
= NEWZONE(16);
'2'
=
OLDZONE(12)
.
NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+'
'+'
= LAND & HIGHPONT; '
= LAND(5);
'
= HIGHPONT(10);
= NO REF ATTS
'-
comment the RCOM file, "newzone p", which generated
comment the new pattern follows:
printf newzone p
TLIST1 *=*
LOWRES
.OR.
MIDRES
EX BESIDE (TLIST1)
TLIST2 *=* BESIDE .AND..NOT. WATER .AND..NOT. HIGH'
POINT.AND..NOT. TIDAL
EX DNEAREST (HIGHPOINT TLIST2)
PUT ARG1 QI, QJ
EXPAND +$2ARG1 HAS BEEN ASSIGNED TO LOCATION $3QI,$3QJ/+
EXPAND +AND IS $1Q FEET AWAY FROM THE NEAREST HIGHPOINT/+
READCONSOLE
CHAPTER IV:
comment
comment
comment
comment
comment
comment
comment
comment
comment
comment
comment
comment
LANGUAGE
PAGE 92
newzone p requires three other small files
which are geometric "Functions of Measurement".
(see Chapter III)
Ibeside p results in an ATTRIBUTE list which
is beside the named ATTRIBUTE list but not on I t.
Idnearest p find the nearest member of the
second named list to any member of the
first
named list.
and Inearest p is called by
dnearest p in the course of its work, finding t he
closest member of the list named to the
LOCATION specified.
printf Ibeside p
BEGIN_tMACRO BESIDE
BESIDE *=*
NULL
FOR 1= EACH LOC ON ARG1 BLOCK 1. BESIDE(I)
BESIDE *=*
BESIDE .AND..NOT. ARG1
END_MACRO
RETURN
printf Idnearest p
BEGIN_MACRO DNEAREST
Q = 999999.
THROUGH LABEL FOR I=, J= EACH LOC ON ARG1
EX NEAREST (I J ARG2)
IF Q .LES. MINDIST GOT O LABEL
Q = MINDIST
QI = BESTI
QJ = BESTJ
LABEL$ CONTINUE
END_MACRO
RETURN
printf Inearest p
BEGIN_MACRO NEAREST
MINDIST = 999999.
THROUGH NE2 FOR 11=, JJ= EACH LOC ON ARG3
SUM1 = ARG1 - II .POWER. 2.
2. + SUM1 .POWER. 0.5
DIST = ARG2 - JJ .POWER.
IF DIST .GRT.
MINDIST GOTO NE2
MINDIST = DIST
BESTI = II
BESTJ = JJ
NE2$ CONTINUE
MINDIST = MINDIST * GRIDSIZE
CHAPTER IV:
PAGE 93
LANGUAGE
RETURN
ENDMACRO
READCONSOLE
Through
of
the Introduction
the
to name a
statement) plus the capacity
(the "if"
with that name (to make RCOM
series of requests and call them
files), most of the sorts
and
"for"
("goto" statement), and the
"through" statements), the branch
conditional
loop (the
of manipulations which the designer
can make on his description are possible in DISCOURSE.
II.
CHARACTERISTICS OF ATTRIBUTES
CHARVARS AND CHARCONS
An ATTRIBUTE
kinds of
at
LOCATIONS;
different
CHARacteristic
whose
and
"number" would
different LOCATIONS;
value
everywhere.
calculations of total
have
a
land
"residence",
use
(low
of
income
vary
its
with
its CHARCON "income" would have the same
The
names
CHARCON
and
values
permit
income or income per person in locations
where more than one income level
doing
regardless
which could
a
is
and a CHARCON "income",
values
a
value may
CHARCON
ATTRIBUTE "lowres"
residents) had a CHARVAR "number"
CHARVAR
a
CONstant
value remains
For example, if the
LOCATION.
characteristics:
an ATTRIBUTE whose
a CHARacteristic of
CHARVAR is
VARy
can have two
survey he
"industry",
is found. If a designer were
could
"commerce",
LOCATION (or some finer breakdown).
assign
and
the
ATTRIBUTES
"open space"
to that
Probably he would be well
CHAPTER IV:
PAGE 94
LANGUAGE
advised to make a small grid size. The smaller grid size would
permit mapping of the uses in
as many of the
Of course,
LOCATIONS
if he were
distinct areas or at least with
as possible in distinct categories.
involved in a
study and all he needed to know
regional transportation
at the time was the number of
origins and destinations by type and
in that LOCATION, then the grid
quantity and time of day
size of 500 feet square might
be satisfactory.
The
numbers
ATTRIBUTE "residence"
of residents
could
broken down
have
by age
as CHARVARS
and sex
the
and as
a
CHARCON the average income per family. A school could have the
number
of
students,
teachers,
administrators
and
staff;
industry and commerce could be elaborated in similar ways.
In
DISCOURSE,
the
new
ATTRIBUTE
plus
CHARVARS
and
CHARCONS would be defined:
"define residence with 3 charvars:
men women children '
and with 1 charcon: averageincome = 9000"
and it
would be placed in LOCATION i,j:
"put residence i,j(22. 30. 32.)"
indicating
children. Note
that there
were
that the average
22 men,
alone then
to state
been:
32
refers to
residence occurs. If it had
the average
"residence" would have to
CHARVAR rather than a CHARCON.
women, and
income per family
all LOCATIONS where the ATTRIBUTE
been desired
30
income for
this LOCATION
have had "income"
as a
The two statements would have
CHAPTER IV:
LANGUAGE
PAGE 95
"define residence with 4 charvars:
men women children '
averageincome*"
"put residence i,j(22. 30. 32. 9000.)
Another kind of legend might be
spaces and
of human
kept separate as
might like
some
the two
much as possible. For
the
purpose might be
categories were
example, the designer
of loft spaces of
average areas
of
more than
those recorded.
to find under-used spaces
of a certain size.
of use of
activity where
to record the number
area, and
an inventory of types of
His
for public events
He might al so like to know the time tables
the spaces.
gotten much of
(Whereas he might have
the data for the previo us legen d from census and other surveys
already completed, he might we11 have had to
collect most of
the data for this legen d by his own field surveys.)
One way of recordi ng this information would be:
"define loftspace with 4 c harvars: number averages ize
use6to8 use8tol0* and with 1 charcon: minsize = 5000."
In "number" would go the
greater than or
would go
number of loftspaces whi ch were
equal to 5000 square
their average size
feet;
in "avera ges Ize"
for this LOCATION;
in "use6to8"
and "use8tol0" would go the number of those spaces in use from
6 to 8 and from 8 to 10 PM respectively.
If the designer
similar
different
ways,
and
could describe other types
he
activities, he
could
describe
could
environment as space and activity.
build up
people
a
of spaces in
engaged
picture of
For example:
in
the
CHAPTER IV:
LANGUAGE
PAGE 96
"define people with 4 charvars: working residing '
shopping playing*"
in which
way of knowing
however, there would be no
In this case,
type of
space
were working.
they
Instead,
the
designer could make a finer breakdown of types of people:
"define workingpeople with 2 charvars: inloftspace
insmallspace*"
Similarly there
space Industrial
would be
no way
to know
establishments were in
what kind
unless there
of
were a
further specification under that category:
"define industry with 2 charvars:
inloftspace
insmall space*"
Nor would
there be any way
to know whether
the working
people were working at industry or retail, for example, unless
there were the further specification:
"define industry with 4 charvars:
inloftspace
insmallspace workingloft workingsmall*"
categories of
With
the description
of the
environment
referring to classes of individuals rather than to individuals
enrichment of the
description
beyond
the
description can be done
of each
category, even
point) of
redundancy
to the
as
by elaborating the
point (and
has been
done
well
above.
(Enrichment of the description is further discussed in Chapter
VI.)
To
illustrate
pertaining
the
display
particularly to
features
data which
of
the
language
include CHARVARS
and
CHAPTER IV:
PAGE 97
LANGUAGE
CHARCONS, the data file used In
has
been elaborated
figures (4.11)
and will
to (4.21).
read as part of the text.
the Section I. of the chapter
be
shown in
Figures (4.11)
various ways
to (4.15)
in
can be
comment step 1: finding out what
is
in the data file.
PAGE 98
list attributes
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
0 -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 --
NULL
UNIVERSE
ROW
HIGHRISE
DETACHED
LAND
WATER
LOWRES
MIDRES
TIDAL
HIGHPONT
RENTS
list row
CHARVAR
CHARVAR
CHARVAR
CHARVAR
CHARCON
I
2
3
4
1
ROW
ROW
ROW
ROW
ROW
CALLED
CALLED
CALLED
CALLED
CALLED
RENT
UNITSOCC
UNITSVAC
DUSPACE
COST
list land
LAND HAS NO CHARVARS
LAND HAS NO CHARCONS
list lowres
CHARVAR 1 OF LOWRES IS CALLED NUMBER
CHARCON 1 OF LOWRES IS CALLED INCOME
list tidal
CHARVAR 1 OF TIDAL IS CALLED AREA
TIDAL HAS NO CHARCONS
list highpoint
CHARVAR 1 OF HIGHPONT IS CALLED HEIGHT
HIGHPONT HAS NO CHARCONS
list rents
CHARVAR 1
CHARVAR 2
CHARVAR 3
RENTS HAS
OF
OF
OF
NO
RENTS IS CALLED AVERAGE
RENTS IS CALLED TOTALD
RENTS IS CALLED TOTALP
CHARCONS
Figure (4.11)
PAGE 99
list temporary variables
GRIDSIZE = 500.
I = 0.
T = 9.
S = 975.
Q = 1.
R = 11.
N = 1.
P = 8.
comment step 2: continuing the search, now using
comment "Functions of Measurement"
show count lowres
LOWRES EXISTS AT 9 LOCATIONS
comment this also could have been sritten as a loop:
for i = each loc on lowres t#z = z + 1.
show z
Z = 9.
show total lowres(number)
TOTAL OF LOWRES(NUMBER) = 975.
comment similarly the last command could have been:
for i = each loc on lowres y = y + lowres(number i)
show y
Y = 975.
show total lowres(number) with tidal
TOTAL OF LOWRES(NUMBER) = 175.
comment this totals up the number of "lowres" living in
comment locations where there is also "tidal", and
comment would have needed:
tlist *=* lowres .and. tidal
for i= each loc on tlist u = u + lowres(number i)
show u
U = 175.
Figure (4.12)
PAGE 100
show lowres
LOWRES (1,1
LOWRES (1,2
LOWRES (1,3
LOWRES (2,1
LOWRES (2,2
LOWRES (2,3
LOWRES (2,4
LOWRES (3,3
LOW RES (3,4
show lowres with tidal
LOWRES(3,
LOWRES(3,
show row with lowres .or.
midres
ROW(1,1)
ROW(1,3)
ROW(2,2)
ROW(2,4)
ROW(3,3)
ROW(3,4)
ROW(3,9)
ROW(3,10)
ROW (3, 11)
comment this shows al 1 the locations which contain row
comment houses and ei ther "lowres" or "midres".
show row(rent)
ROW( RENT 1,1) = 100.
ROW( RENT 1,3) = 100.
ROW( RENT 2,2) = 90.
ROW( RENT 2,4) = 120.
ROW( RENT 3,3) = 100.
ROW( RENT 3,4) = 90.
ROW( RENT 3,9) = 250.
ROW( RENT 3,10) = 250.
ROW( RENT 3,11) = 300.
show row(rent 1,1)
ROW(RENT 1,1) = 100.
Figure (4.13)
PAGE 101
show highpoint(height)
HIGHPONT(HEIGHT
HIGHPONT(HEIGHT
HIGHPONT(HEIGHT
HIGHPONT(HEIGHT
1,6) = 30.
4,8) = 55.
5,11) = 65.
6,1) = 25.
show loc 1,1
LOCATION 1,1 CONTAINS
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
1 -2 -3 -5 -7 -11 --
UNIVERSE
ROW
HIGHRISE
LAND
LOWRES
RENTS
show gridsize
GRIDSIZE = 500.
comment any DISCOURSE element can be renamed.
rename lowres lowincome
list attributes
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
0 -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 -12 --
NULL
UNIVERSE
ROW
HIGHRISE
DETACHED
LAND
WATER
LOWINCME
MIDRES
TIDAL
HIGHPONT
RENTS
TLIST
rename lowincome lowres
delete tlist
Figure (4.14)
comment step 3: generating new values, a "Functions
comment of Further Description"
rd rents p
AVERAGE RENT, TOTAL RENT, AND NUMBER OF DWELLING
UNITS PAYING RENT HAVE BEEN COMPUTED.
THEY ARE ADDRESSED: RENTS(AVERAGE),
RENTS(TOTALD), AND RENTS(TOTALP).
PAGE 102
ALL NON-MACRO LABELS REMOVED
comment the RCOM file which did that is printed below:
printf rents p
SUBSTITUTION ON
THROUGH R3 FOR I = EACH LOC ON RENTS
FOR Z = 1. STEP 1. UNTIL 3. RENTS(Z
I) = 0.
R3$ CONTINUE
THROUGH R2 FOR QQ = 2. STEP 1. UNTIL 4.
THROUGH RI FOR I = EACH LOC ON QQ
RENTS(TOTALD I) = QQ(RENT I) * QQ(UNITSOCC I)
RENTS(TOTALP 1) = QQ(UNITSOCC I)
R1$ CONTINUE
R2$ CONTINUE
THROUGH R4 FOR I = EACH LOC ON RENTS
RENTS(AVERAGE I) = RENTS(TOTALD I)/RENTS(TOTALP 1)
R4$ CONTINUE
EXPAND +AVERAGE RENT, TOTAL RENT, AND NUMBER OF DWELLING/+
EXPAND +UNITS PAYING RENT HAVE BEEN COMPUTED'./+
EXPAND +THEY ARE ADDRESSED: $2RENTS*($2RENTS(AVERAGE)*),/+
/+
EXPAND +$2RRENTS*($2RENTS(TOTALD)*),
EXPAND +AND $2RENTS*($2RENTS(TOTALP)*)'. //+
SUBSTITUTION OFF
READ_CONSOLE
R; T=0.04/O0.87 23.32.55
Figure (4.15)
PAGE 103
comment step 4: display
rd tabmap2 p(rents average rents totalp)
1
2
3
4
5
6
7
8
1
80
80
80
200
85
30
2
80
100
90
20
140
10
3
B
B
%4
B
8
5
B
B
6
B
B
7
B
B
8
B
B
B
B
B
B
120
40
B
B
B
B
B
B
B
B
9
10
11
200
30
200
40
210
50
210
30
200
25
175
10
150
15
150
10
B
B
B
B
190 180 190
30
40
30
280
25
310
40
250
40
1
2
3
4
5
6
7
8
9
10
11
RENTS(AVERAGE) AND RENTS(TOTALP)
HAVE BEEN GIVEN A VALUE WHERE THEY EXIST
"B" INDICATES THAT THEY DO NOT EXIST AT THAT LOCATION
ALL NON-MACRO LABELS REMOVED
commen t
comment
comment
comment
The top values in each LOCATION are the average
rent paid in that LOCATION; the bottom values are
the numbers of swelling units in each which
pay rent.
Figure (4.16)
comment the designer-written RCOM' file which
comment created the display
printf tabmap2 p
PAGE 104
SET CARRIAGE WIDTH 120
TABSET 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80
85 90 95 100 105 110 115 120
THROUGH P1 FOR J= Q STEP 1. UNTIL R
EXPAND +.$3J+
P1$ CONTINUE
EXPAND +///+
THROUGH P2 FOR I = N STEP 1. UNTIL P
EXPAND +$31+
THROUGH P3 FOR J= Q STEP 1. UNTIL R
GOTO P5
IF ARG1(ARG2 IJ)
EXPAND +.B+
GOTO P3
P5$ CONTINUE
EXPAND +.$3ARG1(ARG2 I,J)+
P3$ CONTINUE
EXPAND +.$31/+
THROUGH P6 FOR J= Q STEP 1. UNTIL R
IF ARG3(ARG4 1,J) GOTO P7
EXPAND +.B+
GOTO P6
P7$ CONTINUE
EXPAND +.$3ARG3(ARG4 I,J)+
P6$ CONTINUE
EXPAND +//+
P2$ CONTINUE
THROUGH P4 FOR J= Q STEP 1. UNTIL R
EXPAND +.$3J+
P4$ CONTINUE
EXPAND +/$2ARG1*($2ARG1(ARG2)*) AND $2ARG3*($2ARG3(ARG4)*)/+
EXPAND +HAVE BEEN GIVEN A VALUE WHERE THEY EXIST/+
EXPAND +"B" INDICATES THAT THEY DO NOT EXIST AT THAT LOCATION///+
RETURN
R; T=0.05/0.47 00.59.43
Figure (4.17)
PAGE 105
comment
comment
comment
comment
showing the LOCATIONS of row housing with
step 5:
respect to low and middle income residents.
The "mapset" command allows the designer to pick
his own character set for the map if he chooses.
red (6;+)
mapset standard (1;a)
red
map* (row lowres midres; highpoint land) 1-8 1-11
1
12345678901
•. • •-.
1
1
2 -2-A-2-A-
-A-2-A-.-•+.-.
..-3-3-3-3
3
-.-.- A-A.
4
5
-.
6
+.
...........
7
8
-.
-.
.
.
2
.
-3-A-A-A
3
...
. . .
.
+.-3-3-3
. .-3-3+3
4
5
.
.
.
6
.
.
.
7
8
. . .
..........
12345678901
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS WRES(7); '3' = MIDRES(8)
'A' = A,
2 'A' = A, 3
'.'
-
'A'
(4.,8)
'2' = LO
= NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '-'
= LAND & HIGHPONT; '
'+'
' = LAND(5);
Figure
= ROW(2);
= HIGHPONT(10);
' = NO REF ATTS
'-
PAGE 106
comment alternatively one could differentiate between the
comment combination of "lowres" and "row" and of "midres"
comment and "row".
mapset standard (1 2;a) red only (1 3;b) red only (6;+) red
map* (row lowres midres; highpoint land) 1-8 1-11
1
12345678901
1
-A-2-A-.-.+.-.-.-.-.-.
1
2
3
4
-2-A-2-A-.
-.-.- A-A-.
. .- 3-3-3-3
. .- 3-B-B-B
*....
+.-3-3-3
2
3
4
5
6
7
8
-. .......
+. .
. .
-. . . .
. .
.-3-3+3
. . . .
. . . .
. .. ..
5
6
7
8
.
.
.
.
.
.
.
.
.
123456789O1
)
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS 7); '3' = MIDRES(8)
'A'
= 1,
-
'1'
= ROW(2);
'2' = LOWRES(
2
'B' = 1, 3
= NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+' = HIGHPONT(10);
' = LAND(5);
'+'
= LAND & HIGHPONT; ' ' = NO REF ATTS
Figure (4.19)
'-
PAGE 107
* C;
4
04
.0
O
4
C-400
4
O*
CD 00
*
0
LNOO
44
C)00 r-l O
at'%
00
*CaJ
NO
C••
"-C'4
0N*
N Ln
0.
0
LU
CD
000
L 1n
000
-I
La
4 I-4v-Ie'%
0.-
--0Iq l 0 4
000
000
*
O
0•
C
" 00
LU
v C-
0
C
*
00
Lno
**
*N
40LAL
Cw
=a:
00
m
*LAv-
v-i
C;
A
C)
La0
v-
h*n
or- Cn
o
N>uor
a
CN
O-
4
U.J C%
4 ACN
__
?4
r-
a
*W~
*nC)
<wCD*CD
H*
vI
Q.LUL
-- = CDJ·ovU
sC
4
O
C4 v-I0LA a
C*
Z
-LLJ
)OU
CN
vI
0 J--- mCDO
C:) 0Lunnw
Uln U-1Iv-400
oCD
-l E al
*LL)-*L
tn r-4 " c
Oi
%cn4 *
r-4 C%4
M000
%00
v-C'
4
0O
-
zWv-4
C)4
C-4
4 LO
LO%00
r4
0ULO%
lA0X
C00
4
4E- C'
NV0
4
4**1-4
04
0 4-@
eNJ,4M-
r-I-
.0
W4
v-r4
N- f
rl%
4
0
v--I
0
0
-..t-m
C)
U%
<
LUw
r- 4
C4 r-4 cn
er
CD
oo
0I'*0
v-I
%
tN% =) LAC
.-- LO tC
LO
a OM0
MLC
4
1-1-
a)
-4 0 LA
f%- A
c)
0
C
LU M
S41% a4
00
=
% `ý Ln
4
0n P
W
L
0%
r4LO t4
0
Co0oA
02:
400
w
a
r-4tLnm
* CD C
D
ON
N
4
0
00A
cJ ''AO4** M IA
0
L100L
U)
0
0M 4 C 14*,
*) C_v-r
OLOZ:
*-
MO0
v-4.:I-LA
ONN
14viC-)
00
a
LA
%
CD r-. I*-
AN
v--
00
•
-
c) n 4= H
z:
H-o
0
HO
4
0
444W
LA
4
*
rl
v ir
S4
*
*
0
0o0
C)4
CN
t00
as
LA?
O
_L 4a
.004
r-
at
•
Z,
444
0 %
4%@vi
V-4~
M rl-
M
C
CD LOv4#
nLAK*cn
V%0
0
COO
*
0
0002:
T 00D
MCA
A
AN M
a
G)M
04rGc 4
-4 rLn
MO
NV)
i-
Cq
400
4
cn
-q 0
aC *Ln
C)
WC%4CD
L00
(UVN
v
CD
.CD
ONur
0) Cý4 h-lv
LA
i-
%-C:)rf
-41
v-4Srr-.
4 4
<0
L*
Q0 o *OWN
0C/)7J
eq* oL 0
A *LA
Co
00u*
C-
L
v-I 0
0) Ln C4~
4
0
-4
U%
mbH
O
000
000
N
pLn
o
0
*
00t
C
I- COLA000
-.t IC
0
4
4
C)C*v-C)4
v-I
II
..
ri
Q
II
0) %..,
•
1•
%L
CnVMM-4
m 4*CJL
O *CD
%
**C
W
4
Ln
LnrC
J
%
%< rrLD
0 uj
00 ; t4 -:0 LOtU v-<LA
C=
v-I4
4
UA4
HZ o-
-
LLHAMC-1LL
0UD 4 r4LUD
%
M
40.D-0O
PAGE
Ma
0
*.
.0
0
S0
~t
h
Vs
hr
v
DL
*0
*00
M000004
(NA
00
0'LA?
*
LOr'=)-4e
C*-
LA
C)
(N(N0
C', C.
SLn
~
Ul
0%00
C.
'
C
)
C
v
un a *
LOC
%-
--
oo
V
I.
t
04r
%
-
(/)
%-o
r cr.
o.
OLU
a
Va
r-4
OM
0
-j<
0
Lo
<
=
LU--
LL
-zl-
C,4 a CL r--4v
0
-
Z
LU L
00LLZIT- L
-LL
LL w3 0*W
.. LA 0
0
-
--Lo
La
0. 0
-C
*. %.
t
*
0
% eC
*r
c
=
0
-J
0C
U)
rZl C%4-.tN a
0=
C
0C.
%C
A7)
C · L
*
V)
*002
-:rL
eC)
*
u-4I
=00
6
10
CCCN
oC
000
006
C
ZuC)'-o,
LA
IC
.*
*C)
f%
WAr4=
Ln**LC14
r-lP
W
CD %4
%C)
-- r**4
aLAn
%
C10)0
C*
LLO
N'*
%*.
r-1
00
0O
(N 0U
%
Ln
w
e0-T
0rI%C=)-%%.00
*
,* *q-%%-In
**
C *
C1.4
%-*
<
_
04
4LAZ
%.
00
00(4C)
n 1ý' tn
C
C%
*IAC
e J 0 u-I CO
rL40
LA<"%CDO
A 4.
*
Mtr r
LA V
ti
*
VOO
wo
-*
* t-IC-D
0
*C
"
.. C
rl(N CD (NC) C,4r
C%4 r.C=
r-l < 4=
pN %
r
O
CN
C) 141)0
L
*
CI r-
CHN
O
*
eu-IC
CUN
4004
Om
·
LD
CLA
CO-I
^ 000P.t
*
C.
M
O
04O
IA
-
Ln
%-r-l r-4 nU)
L
*
r%-
CD
%--O.D4'j.?LA
r.N
4
0
*
I
O
*
*
.0
*
NC
L000
-IC
*
*
(N
7
L?(
*
*
· CN
V*
Z
um
CD r-l
ofC
P')
LA
Lh
#
C Ph u4uI
-I#
V *'%(%i-I
00CDC
*
00 CN
0
O.0
*
%C1
-C
C)
u-I C) C)
C)I --I
%t
04 LA
C* C)C:)*
e000% *
NcI
m LU
%N
LA-
.
*C''0
*-C
0
*L
C)
00
l
C
LAt
a--r"COo
LA (N -M tAC
D '
0
--q 00
II
II
II
C N
II
,.rr-
II
II
...
,
IIt <
az
u
Zacr
U
108
V
CHAPTER IV:
PAGE 109
LANGUAGE
I II.
SYSTEM COMPONENTS
The components of the DISCOURSE system are the designer's
console at which
his
output;
he issues his requests and
core,
computer, which
the
central
is the active
receives most of
processing
part of the
unit
of
the
computer's memory
where the computation
takes place and where
of DISCOURSE resides;
and "disk files" which contain the rest
of
the DISCOURSE
system and
the main program
the designer's
data files
and
special RCOM files.
The path followed
at the
during a typical request
console, interpreted
branches to
read into
by the
is initiated
main program
which then
the appropriate pre-compiled subprogram
core from the
disk if it
which is
is not already
also reading in the appropriate parts
in core,
of the data if they are
not in there already, executes the requests and either returns
the reply to the designer at his
printer in
want to
the computation
wait for
console, prints it on a line
center if
what he has
the designer
requested to
does not
be typed
by the
console, or writes the reply on a disk file.
The examples in the previou s two sections of this chapter
were all
involved work in
utilized
to
comfortable as
the
within i DISCOURSE,
generated from
designer
but some
another progra m called EDIT
create
and
to moSdify
disk
must be
fluent
in
(4.04) which is
files.
presently possib ,le in working
of theim
To
be
as
with DISCOURSE,
both DISCOURSE
and
EDIT.
CHAPTER IV:
DISCOURSE permits
called a
EDIT.
course
the making of a
having to
MACRO without
MAC ROS
which the
PAGE 110
LANGUAGE
have been useful
user has
of this
which perm it
RCOM file
leave DISCOURSE
and enter
for short series
wanted to repeat
work, but
of requests
quite often
DISCOURSE does
editing MACROS once
ones which are
special type of
during the
not have
created;
not wanted can be erased and
instead,
features
the old
new ones created
in their places.
The f iles
used in connection
with the examples
in this
chapter we re all written in EDIT, tried out in DISCOURSE which
returned a n erroi r
found some hing
message
and EDIT
RCOM fil
worked.
i s eas y:
Communication
one need
D ISCOURSE,
working on In
return to DISCOU RSE recall
could not
agai n in D SCOURS E, and so on
interpret, ed i te d i n EDIT,
until the
it
only "store"
enter EDIT,
the
create
"stored"
between DISCOURSE
what he
has been
the needed file,
data and resume work,
his newly created RCOM file. The communication
able to ut i lize
will
become even easier because soon
call
EDIT from w ithin DISCOURSE.
the use r will be able to
When a desi gner commences to work on a des ign exercise he
has to spend a
does have
great
deal of time entering the
about the environment
this file is
into a data
easily modified, there is,
data which he
file.
Although
nevertheless, apt to
be more work in the early than in later stages associated with
the
building of
however, that
the
data file.
one of the major
It
should be
emphasized,
virtues of DISCOURSE
is that
CHAPTER
new data
The
IV:
PAGE 111
LANGUAGE
can be easily added
data file
which is stored on
designer constructed
files, only a
could be made from the console.
in the
and then stored for
system between
the disk
series of
later use.
is, like
other
requests which
Thus, there is no distinction
the initial
data and
new data.
The
designer can, of course, make that distinction if he chooses.
PAGE 112
CHAPTER V
SIMULATION OF ENVIRONMENTAL USE
facilities at a place
likelihood of people using recreational
in an
would
urban area
vary
the distance
with
more
look
to
perhaps adding
places of employment were,
facilities and the
recreational
the
where
at
closely
those
of
The designer might then
facilities from places of employment.
begin
that the
given designer
to a
seem reasonable
It might
recreational facilities at different locations and making some
guesses about how the use pattern would be affected.
This experiment
effort to
is an
extend the
designer's
ability to consider this one aspect of his design problem.
the designer left himself the
order to set up the experiment,
of
consequences
the
explore
to
opportunity
In
differing
assumptions about how people use environments.
DESCRIPTION OF THE EXPERIMENT
I.
In the
highly simplified world
designer has
where
has
included
average income
categories of
experiment, the
recreation;
a little
extra
and educational
employees.
He has
places
work;
bluecollar employees
and
of employees in each category
and the numbers
He
places of
described only:
managerial, clerical,
of this
information,
level achieved
at each place.
namely,
by the
described the piece
the
three
of the
urban area he is working on as a collection of 400 blocks, 200
feet square.
or "RCOM
A "file,"
Files.
of
PAGE 113
SIMULATION
CHAPTER V:
basic
to the
designer-written supplement
or a
the data
is either a description
file,"
language of DISCOURSE, adapting DISCOURSE to the special needs
of
the
experiment. (5.01)
experiment are included
files needed
the
All
the
part of the
later in this chapter as
file in which
The
explained where they appear.
text and are
for
the designer has recorded the initial state of the environment
is entitled RECREAT P.
people will
himself
environment in
use the
the opportunity
at one location
the basic
in which
a file
DISCOURSE Requests
of Chapter
Further
Description,"
and,
is a
III, RECUSE P
a
specifically,
Extension" because it generates new
will use
are COMbined,
is entitled RECUSE P. In
which actually makes the computation
the terminology
namely
The RCOM file,
distance.
recreational facilities within what
how
left
way but
a general
certain aspects:
to change
of people employed
what proportion
assumptions about
his
expressed
designer has
The
"Function of
"Function
of
data by means of existing
• I
data
plus
a rule
of
inference
how people
about
use
the
environment.
The designer wants
to be able to get
he has generated with RECUSE P:
at the information
to find out where people are
using the recreational facilities, PWHERE P; where those using
some recreational
the locations
facility come from,
which are used by
RWHERE P;
more than some
the designer will decide at the time, MAP P.
and
to map
number which
Again, according
CHAPTER V:
PAGE 114
SIMULATION
these are "Functions of
to Chapter III,
but simply represent the data which
they do not add new data,
are there, or
Measurement" in that
of arithmetic, logic,
derive data by the rules
or geometry.
wish to
generate and
interrogate data
those easily handl ed by the basic
the
files he
designer works,
that as the
be noted
It should
reques
writt n.
has alrea dy
in many
he may
ways besides
in DISCOURSE or by
ill,
He
therefore,
be
engaged in intermi ttent "programm ng."
II.
USES OF THE EXPERIMENT
The most obvious use to be
made of this experiment is to
try new locations
for recreational facilities and
how they are used
by whom and from where. Note
simulates how the environment will
about
the
environment,
data
the
that RECUSE P
be used, given information
in
RECREAT
procedure, or rule of inference, in RECUSE P.
Given some disposition of
to observe
and
P,
the
(5.02)
the employment and recreation,
the designer can establish new values for the parameters which
accessible to
are directly
parameters
the type
are:
environment he
him in the
of
the
wishes to simulate;
they will walk to recreational
P. These
file RECUSE
use
employees whose
the distances
of
he thinks
facilities; and the proportion
of employees at some workplace which will use the recreational
facilities at those distances.
There are several
he
might neither
be
reasons why he might
sure how
far people
do this.
would
First,
go to
use
recreati onal
walking,
nor
at all.
If
he could s tate
see
the
extreme but reasonable
range
the
of
consequences
h is assumptions.
to
one
from
distances
several
ke
go
limits, he
in
He might,
f rom 0.5 percent to
proporti on of users
mode,
employed woul
what proportion of those
know
recreati onal use of
vary
given some transportation
facilities
t hen
could
PAGE 115
SIMULATION
CHAPTER V:
terms
of
for example,
and
blocks
50 percent.
the
Second,
he might try to establ ish reasonable assumptions for different
groups
where
then see
and
recreati onal facilitie s.
then
the
the distance which
the differences
observe
same
want to assume a new
in order to extend
go and
would
be using
might
Third, he might
mode of transportation
people
they
in
the
patterns of use.
If he found
to vari ations of the param, eters
of use due
radically
imply
field survey or
a literature search i n order
best
about
assumptions
and hi s
environment
best
guesses
recreational
facil ities, better
environments
would
development
through a
program.
problem in
of
He might,
make
The
this way
about
in
capacities
how people
that
the
of
use
recreation
the
consequence
should be
use
people
how
a bout the
data
second
of
to generate new
course, discover
differe nce
no
location
uncertainty about
help to reduc e the
parameter valuies.
his
the
night want to initiate a
the designer
information which m ight
making
was sufficient to
for
programs
d ifferent
recreational facili ties,
the
int ensity and pattern
that the range in the
of
that the
working
designer
T
CHAPTER V:
SIMULATION
becomes more
PAGE 116
aware of
critical to
the environmental
his description
and the
elements which
relationships which
are
he
believes to exist among them.
The uses which the designer would make of this experiment
might reflect
might
the range of
be placed,
his ideas about
how people
use the
where recreation
environment, and
what
transportation facilities were imaginable. He might change the
experiment in response to its
to his own
in response
not being interesting
growing understanding of
to him;
the problem
and changing perceptions of what was important to observe;
to new information
in response
people
use
becoming available
or
environment
the
about
new
or
about how
modes
of
transportation.
of interrogating the data, for
He might develop new ways
compute
to
example,
available
or to
their annual income;
users.
power
purchasing
any recreation location as
recreation users at
by recreation
the
of
a function of
compute average distance traveled
assign a cost to
If he could
transportation system as well as
some new
to new recreation locations,
one of which depended
he could compare alternative proposals,
heavily upon making better use of existing parks and the other
of
which did
modes
change the
not
of
travel or
distance
traveled by people, but provided more recreation areas.
The designer
change his
RECUSE
could, after
data, or
P. He
his procedure
might decide
some time,
experimenting for
that he
for generating
needed to
new data,
discriminate
PAGE 117
SIMULATION
CHAPTER V:
among more types of environmental
day of use;
RECUSE P
users or introduce times of
or he might decide that the procedure embodied in
which selected the
location was
nearest recreation
not spreading people out enough to approximate reality so that
he could get
a sense of where
to distribute people to
therefore, rewrite it
locations which fell
He might,
meet.
people would
all recreation
their places of
within some radius from
work.
III.
The experiment
THE EXPERIMENT
on the following
cursory look at the data file,
contents and
The
mapping "office,"
experiment
second
following
within it.
itself
page.
pages is preceded
by a
RECREAT P, listing some of its
"industry,"
commences with
Explanatory
and "recreation."
"design2
comments
p"
are
on
the
embedded
PAGE 118
comment recreat p is a data file containing information
comment about places of employment and recreation. One
comment can find out about it by reading "irecreat p"
comment which is included later in this section of the
comment chapter.
list attributes
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
0 -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 -12 -13 -14 --
NULL
UNIVERSE
MANAGEAL
CLERICAL
BLUECOAR
INDUSTRY
OFFICE
RECREAON
TLIST
TLIST1
TLIST2
ALLWORK
ALLREC
SUBREC
SUBWORK
list recreation
CHARVAR 1 OF
CHARVAR 2 OF
CHARVAR 3 OF
RECREAON HAS
RECREAON IS CALLED CLERICAL
RECREAON IS CALLED MANAGEAL
RECREAON IS CALLED BLUCOLAR
NO CHARCONS
list managerial
CHARVAR
CHARVAR
CHARVAR
CHARVAR
CHARCON
CHARCON
1
2
3
4
1
2
OF
OF
OF
OF
OF
OF
MANAGEAL
MANAGEAL
MANAGEAL
MANAGEAL
MANAGEAL
MANAGEAL
IS
IS
IS
IS
IS
IS
CALLED
CALLED
CALLED
CALLED
CALLED
CALLED
NUMBER
RECLOCI
RECLOCJ
RECNUMB
EDUCATON
INCOME
list office
CHARVAR 1 OF OFFICE IS CALLED EMPLOYES
OFFICE HAS NO CHARCONS
PAGE 119
map (office industry; recreation) 1-20 1-20
1
2
1234567890
1234567890
+.
+.+.+.
2 2
2222
222
1 1
111
111
2111
1
1 1
1
11
12
13
+o+,+,+,+o+,+,+,+.
+0
+0
+0
12345678901
2345
6 7 8 9 0
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS -
-
'1' = OFFICE(6);
'2' =
STRY(5)
'.
= NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - ' = NO REF ATTS
comment each grid location
comment is 200 feet square:
show gridsize
GRIDSIZE = 200.
'+'
= RECREAON(7);
INDU
r
PAGE 120
design2 p
IF DESIGN SYSTEM CRASHES GIVE THE DEBUG COMMAND:
DUMP NAME 12000 3D000
R 0.03/0.03
rd recreat p
R 0.69/0.72
rd execrecr p
comment this file has a few convenience commands for this
comment particular experiment.
printf execrecr p
READY OFF
MESSAGE OFF
SET CARRIAGE WIDTH 62
RD LNEAREST P
RETURN
rd recuse p(managerial 300. 600. 0.3 0.1)
USE OF RECREATIONAL FACILITI ES BY MANAGEAL EMPLOYEES HAS BEEN
COMPUTED
rd pwhere p(managerial)
RECREATIONAL
NUMBER OF MANAGEAL
FACILITIES AT
EMPLOYEES USING
comment the distances of 300 feet and 600 feet were too short
rd recuse p(managerial 500. 1000. 0.2 0.1)
USE OF RECREATIONAL FACILITI ES BY MANAGEAL EMPLOYEES HAS BEEN
COMPUTED
rd pwhere p(managerial)
RECREATIONAL
NUMBER OF MANAGEAL
FACILITIES AT
EMPLOYEES USING
3,11
30.00
3,13
20.00
17,12
80.00
17,19
40.00
17,20
20.00
18,12
10.00
19,12
10.00
20,12
10.00
rd rwhere p(managerial 17 12)
MANAGEAL EMPLOYEES FROM THE FOLLOW ING LOCATIONS
USE THE RECREATION AT LOCATION 17, 12
40.00
FROM 13,9
10.00
FROM 14,8
10.00
FROM 15,8
10.00
FROM 16,8
10.00C
FROM 17,8
80.00
TOTAL AT 17,12
7
PAGE 121
rd recuse p(clerica1 700. 1500. 0.4 0.2)
USE OF RECREATIONAL FACILITIES BY CLERICAL EMPLOYEES HAS BEEN
COMPUTED
rd pwhere p(clerical)
NUMBER OF CLERICAL
RECREATIONAL
EMPLOYEES USING
FACILITIES AT
400.00
3,11
3,13
320.00
5040.00
17,12
680.00
17,19
480.00
17,20
160.00
18,12
160.00
19,12
160.00
20,12
rd rwhere p(clerical 17 12)
CLERICAL EMPLOYEES FROM THE FOLLOVII NG LOCATIONS
USE THE RECREATION AT LOCATI ON 17,12
160.00
FROM 11,8
680.00
FROM 11,9
680.00
FROM 12, 7
680.00
12,8
FROM
680.00
FROM 12,9
680.00
FROM 13, 7
160.00
FROM 13,8
680.00
FROM 13,9
160.00
FROM 14,8
160.00
FROM 15,8
160.00
FROM 16,8
160.00
FROM 17,8
5040.00
TOTAL AT 17,12
rd rwhere p(clerical 17 19)
CLERICAL EMPLOYEES FROM THE FOLLOWING LOCATIONS
USE THE RECREATION AT LOCATI ON 17,19
680.00
FROM 12, 19
680.00
TOTAL AT 17,19
PAGE 122
comment map p maps the most used locations.
rd imap p
THIS FILE MAPS THE SUBSET OF ALL RECREATION LOCATIONS
IN WHICH MORE THAN "ARGI"
FACILITI ES.
EMPLOYEES USE THE RECREATIONAL
IT MAPS THE LOCATIONS OF EMPLOYMENT WHERE
THE USER CAME FROM.
AND IT REFERENCES THESE AGAINST ALL
IT IS CALLED BY
EMPLOYMENT AND RECREATION PLACES.
SAYING "MAP P (ARG1)" WHERE ARG1 IS A CONSTANT.
rd map p(150.)
2
1
6 7 8 9 0 1 2 3 4 5 6 7 8 9 0
12345
1
2
+
+1+ .+1
-2-2
-.-
2-2-2
-2-2
-.-.-.
-2
-2-2
-2
-.-2-2
-2-2-2
11
12
13
14
15
16
-2-2-2
+1+.+.+* .++
.+*1+1
+1
+1
123456
7 8 9 0 1234567890
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS - -
'1' = SUBREC(13);
'2' = SUB
WORK(14)
I.' = NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+'
'*' = ALLWORK & ALLREC; '
ALLWORK(11);
= ALLREC(12); '-'
' = NO REF ATTS
=
PAGE 123
CHAPTER V: SIMULATION
IV.
All
of the
FILES FOR THE EXPERIMENT
files
necessary to
do
included in the remainder of this chapter.
the experiment
are
All files in which
each line is preceded by the word SAY simply print the rest of
line
the
for the
designer
when
he
calls that
file.
In
RECFILES P the designer is keeping track of what files he does
have.
The other
files follow in the order in which they are
listed in RECFILES P.
RECFILES P
SAY THERE ARE THE FOLLOWING FILES:
SAY RECREAT P, RECUSE P, NEAREST P, PWHERE P,
SAY RWHERE P, AND MAP P.
SAY THE DESIGNER CAN GET INFORMATION ABOUT EACH
SAY BY CALLING ANY OF THEM WITH THE PREFIX "I",
SAY EG. "RD IRECREAT P".
READ_CONSOLE
IRECREAT P
SAY
SAY
SAY
SAY
SAY
SAY
SAY
SAY
SAY
SAY
SAY
SAY
SAY
SAY
SAY
SAY
SAY
SAY
THIS FILE CONTAINS THE INITIAL DATA DESCRIBING THE
ENVIRONMENT. THERE ARE SIX DIFFERENT "ATTRIBUTES",
MANAGERIAL, CLERICAL AND BLUECOLLAR (WHICH IDENTIFY TYPES
INDUSTRY, OFFICE AND RECREATION (WHICH
OF EMPLOYEES);
EACH TYPE OF EMPLOYEE
IDENTIFY TYPES OF ACTIVITIES).
HAS ASSOCIATED THE NUMBER AT ANY LOCATION, THE ROW AND
COLUMN NUMBERS OF THE RECREATION LOCATION
THEY USE AND THE NUMBER USING THE RECREATION THERE, THEIR
AVERAGE YEARS OF EDUCATION AND AVERAGE INCOME.
OFFICE AND INDUSTRY HAVE THE NUBMER OF EMPLOYEES IN ALL
CATEGORIES; AND RECREATION HAS THE NUMBER
OF EMPLOYEES IN EACH CATEGORY. "PUT CLERICAL 5,6(400.
0. 0. 0.)" MEANS PUT 400. CLERICAL EMPLOYEES AT ROW
5 COLUMN 7 AND SET THE VALUES OF THE 2ND, 3RD AND 4TH
VARIABLES (THE ROW AND COLUMN NUMBERS OF THE LOCATION
OF RECREATION AND THE NUMBER USING IT) TO 0. INITIALLY.
THE COMMAND "SETUP 1 20 1 20 40 40 0"
ESTABLISHES A GRID OF LOCATIONS 20 BY 20 NUMBERED 1 TO 20
SAY IN EACH DIRECTION. ALSO ITCREATES SPACE FOR 40 "ATTRI
SAY BUTES",40 TEMPORARY VARIABLES AND 0 ALPHANUMERIC VARIABLES.
SAY THIS FILE ALSO DEFINES THE "GRIDSIZE" OF THE BLOCKS (OR
SAY INDIVIDUAL LOCATIONS) AS 200 FEET SQUARE.
PAGE 124
CHAPTER V: SIMULATION
READ_CONSOLE
RECREAT P
SETUP 1 20 1 20 40 40 0
DEFINE MANAGEAL 4: NUMBER RECLOCI RECLOCJ RECNUMB*
2: EDUCATON
= 14.
INCOME = 15000.
0. 0.) 5,8(100. 0. 0. 0.) 6,4(100.
PUT MANAGEAL 5,7(100.
) 6,6(100.
0. 0. 0. 6,7 (100. 0. 0.
.
0.
0. 0.) 6,5(100. 0. 0
100. 0. 0. 0.)
0.
,2(100.
0.
0.) 7,
.)
7
0.) 7,1(100. 0. 0. 0
11,
0.)
0.
(
100.
0.
400. 0. 0. 0.)
7,14(100. 0. 0. 0.) 7,15
11,8(100. 0. 0. 0.) 1 1,9( 400. 0. 0.
0. 0. 0.) 12,8(400.
0.) 11,20(100. 0. 0. 0.) 12,7(400.
0.
0. 0.) 12,20(100.
0.) 12,9 (400. 0. 0. 0.) 12,19(400.
0.) 13,7(400. 0.
) 13,20(100.
0. 0.) 13,8(100. 0. 0. 0 .) 13,9(400. 0. 0.
. 0. 16,8(100. 0.
0. 0.) 14,8(100. 0. 0 . 0 ) 15,8(100. 0
0.) 17,8(100. 0. 0. 0.) 18,8(100. 0. 0. 0.) 19,8(100. 0.
0.) 20,8(100. 0. 0. 0
DEFINE CLERICAL 4: NU MBE R RECLOCI RECLOCJ RECNUMB* 2: EDUCATON
= 11.
INCOME = 6000.
I
PUT CLERICAL 5,7 (400. 0. 0. 0.) 5,8( 400.
0.
0. 0.) 6,5(400. 0. 0 . 0 .) 6,6(400.
.0
0.
7,2(400.
0
0.
0.) 7,1(400. 0.
.) ( 800. 0.
0.
7
7,14(800. 0. 0. 0.) ,15
0.
0
11,8(800. 0. 0. 0.) 11,9(3400.
0.
O
.
12,7(3400. 0. 0. 0. ) 12,8(3400.
0. 0.
0.) 12,20(80
12,19(3400. 0. 0
0.) 13,8(800. 0. 0. 0.) 13,9(340 0 . Vn AV
0. 0. 0.) 15, 8(80
0. 0.) 14,8(800
0. 0. 0.) 18, 8(80
0. 0.) 17 ,8(800
0. 0. 0.) 6,4(400.
0.) 6,7(400.
0.
) 7,3(400.
0. 0.
11,7(3400. 0. 0.
11,20(800. 0. 0.
12,9(3400. 0. 0.
.) 13,7( 3400. 0.
0.) 13,20(800. 0.
0. 0.) 16,8(800.
0.
0.)
19,8(800.
0. 0.) 20,8(800.
0. 0.)
DEFINE BLUECOAR 4: NUMBER RECLOCI RECLOCJ RECNUMB * 2: EDUCATON
=
8.
INCOME = 8000.
0.) 6,4(1000.
PUT BLUECOAR 5,7(1000. 0. 0. 0.) 5,8(1000. 0.
6,7(1000. 0.
)
6,6(1000.
0.
0.
0.
0.
0.
0. 0. 0.) 6,5(1000.
0.
0.
7,3(1000. 0.
7,2(1000.
0.)
0.
0.
0. 0.) 7,1( 1000.
0. 0.) 7,14(50. 0. 0.
0. 0.)
.) 11,8(50.
0.) 7,15(50. 0. 0. 0.) 11,7 (200. 0. 0.
0. 0.)
0.
0.
50.
0.
12,7(200.
11,9( 200. 0. 0. 0.) 11,20(
12,8(200. 0. 0.
0.) 12,9(200. 0. 0. 0.) 12, 19(200. 0. 0. 0.) 12,20(50. 0. 0.
0. 0. 0.) 13,9(200. 0. 0.
0.) 13,7 (200. 0. 0. 0.) 13 ,8(50.
0.) 13,20(50. 0. 0.
0.) 14,8(50. 0. 0. 0.) 15,8 (50. 0. 0. 0.) 16,8(50. 0. 0. 0.)
17,8(50.
0. 0. 0.) 18,8(50. 0. 0. 0 .) 19,8(50. 0. 0. 0.) 20,8(50. 0.
CHAPTER V: SIMULATION
PAGE 125
0. 0.)
DEFINE INDUST RY 1: EMPLOYES*
6,4(1500.) 6,5(1500.)
5,8(1500.)
INDUSTRY 5,7(1500.)
PUT
6,6(1500.) 6, 7(1500.) 7,1(1 500.) 7,2(1500.) 7,3(1500.)
DEFINE OFFICE 1: EMPLOYES*
11,7(4000.) 11,8(950.)
7,15(950.)
7,14(950.)
PUT OFFICE
12,8(4000.) 12,9(4000.)
11,9(4000.) 1 1 ,20(950.) 12 ,7(4000.)
12,20(950.)
12,19(4000.)
13 ,9(4000. ) 13,20(950.) 14,8(950.)
13,8(950.)
13,7(4000.)
19,8(950.)
18,8(950.)
17 ,8(950.)
16,8(950.)
15,8(950.)
20,8(950.)
3: CLERI CAL MANGERAL BLUCOLAR*
DEFINE RECREA
2(0. 0. 0.)
PUT RECREAON
,12( 0. 0. 0.) 3,11(0. 0. 0.)
14(0 . 0. 0.)
3,13(0. 0. 0. 17,1 2(0. 0. 0.) 17,13(0. 0. 0.)
17, 18(0. 0.
17 ,16( 0. 0. 0.) 17,17(0. 0.
17,15(0. 0. 0
0.) 17,2 0(0.
0.) 17,19(0.
0. 0.) 19,12(0. 0. 0.) 20,12(0. 0. 0.)
0. 0.) 18,12
GRIDSIZE = 20
RD LNEAREST P
READ CONSOLE
I RECUSE P
SAY THIS FILE COMPUTES HOWMANY PEOPLE WILL USE RECREATION
SAY AS A FUNCTION OF DISTANCE FROM PLACE OF EMPLOYMENT.
SAY IT IS CALLED BY SAYING "RD RECUSE P (ARG1 ARG2
SAY ARG3 ARG4 ARG5)" WHERE ARG1 IS THE TYPE OF PEOPLE
ARG2 AND ARG3 ARE
SAY EMPLOYED, EG. "CLERICAL",
SAY TWO DISTANCES AWAY FROM PLACES OF EMPLOYMENT,
SAY AND ARG4 AND ARG5 ARE THE PERCENTAGES OF EMPLOYEES (OF
SAY TYPE ARG1) WHO WOU LD USE THE RECREATION IF IT
SAY FELL IN THE ZONES DELIMITED BY ARG2 AND ARG3.
SAY EG. "RD RECUSE P ( CLERICAL 400. 1000. 0.3 0.1)"
SAY WOULD MEAN THAT IF THE RECREATION
SAY WAS 500 FEET AWAY 10 PERCENT OF THE
SAY CLERICAL EMPLOYEES AT THAT LOCATION WOULD USE IT,
SAY OR IF THE RECREATION WAS 300 FEET AWAY 30 PERCENT
SAY WOULD USE IT.
READ CONSOLE
RECUSE P
TLIST *=*
NULL
FOR 1=, J= EACH LOC ON RECREATION
RECREATION(ARG1 I,J) =
Z = ARG3/GRIDSIZE
THROUGH RE1 FOR I=, J= EACH LOC ON ARG1
BLOCK Z TLIST(I,J) *
TLIST *=* TLIST .AND. RECREATION
EX NEAREST (I J TLIST)
PAGE 126
CHAPTER V: SIMULATION
IF MINDIST .GRT. ARG3 GOTO RE1
IF MINDIST .GRT. ARG2 GOTO RE4
PCT = ARG4 * ARG1(NUMBER
I,J)
GOTO RE5
RE4$
I,J)
PCT = ARG5 * ARGl(NUMBER
RE5$ ARG1(RECNUMB I,J) = PCT
RECREATION(ARG1 BESTI,BESTJ)
= RECREATION(ARG1 '
BESTI,BESTJ) + PCT
ARG1(RECLOCI IJ) = BESTI
ARG1(RECLOCJ I,J) = BESTJ
RE1$ CONTINUE
EXPAND +USE OF RECREATIONAL FACILITIES BY $2ARG1 EMPLOYEES+
EXPAND + HAS BEEN COMPUTED/+
READ_CONSOLE
INEAREST P
SAY THIS FILE IS CALLED BY "RECUSE P" TO COMPUTE
SAY THE NEAREST LOCATION WHICH CONTAINS ATTRIBUTE "ARG3"
SAY TO THE LOCATION IN QUESTION, "ARG1 ARG2" WHERE "ARG1"
SAY IS THE ROW NUMBER AND "ARG2" IS THE
SAY COLUMN. THIS FILE CAN ALSO BE USED BY THE DESIGNER IN ORDER,
SAY FOR EXAMPLE, TO DETERMINE THE NEAREST "INDUSTRY"
SAY TO SOME LOCATION, EG. 12,18, CONTAINING "OFFICE".
SAY THAT COMMAND WOULD BE "EX NEAREST (12 18 INDUSTRY)".
SAY THIS FILE GENERATES VALUES FOR THREE VARIABLES:
SAY MINDIST, BESTI, AND BESTJ WHICH ARE, RESPECTIVELY,
SAY THE DISTANCE AWAY OF THE NEAREST MEMBER OF ARG3, AND
SAY THE I AND J COORDINATES OF THAT MEMBER.
READ_CONSOLE
NEAREST P
BEGIN_MACRO NEAREST
MINDIST = 999999.
THROUGH NE2 FOR 11=, JJ= EACH LOC ON ARG3
SUM1 = ARG1 - II .POWER.
DIST = ARG2 - JJ .POWER.
2.
2.
+ SUM1
IF DIST .GRT. MINDIST GOTO NE2
MINDIST = DIST
BESTI = II
BESTJ = JJ
NE2$ CONTINUE
MINDIST = MINDIST * GRIDSIZE
RETURN
END _MACRO
READ CONSOLE
.POWER.
0.5
CHAPTER V: SIMULATION
PAGE 127
IPWHERE P
SAY THIS FILE IS USED WITH THE DATAFILE "RECREAT P"
IT COMPUTES
SAY AND AFTER CALLING "RECUSE P".
THE
USING
ARE
EMPLOYEES
ARG1
MANY
HOW
SAY
LOCATIONS.
WHICH
AT
AND
FACILITIES
SAY RECREATIONAL
EG. "MANAGERIAL".
EMPLOYED,
PEOPLE
OF
TYPE
THE
IS
SAY ARG1
ARGi".
P
PWHERE
"RD
SAYING
SAY IT IS CALLED BY
READ_CONSOLE
PWHERE P
TABSET 10 40
EXPAND +.NUMBER OF $2ARG1.RECREATIONAL/+
EXPAND +.EMPLOYEES USING.FACILITIES AT/+
THROUGH PW1 FOR 1=, J= EACH LOC ON RECREATION
IF RECREATION (ARG1 I,J) .EQL. 0. GOTO PW1
EXPAND +.$1RECREATION(ARG1 1,J).$31,$3J/+
PW1$ CONTINUE
READ_CONSOLE
IRWHERE P
SAY THIS FILE IS USED WITH THE DATAFILE "RECREAT P",
SAY AND AFTER CALLING "RECUSE P".
SAY IT COMPUTES HOW MANY "ARG1" ARE USING RECREATION
SAY AT "ARG2, ARG3" WHERE "ARG1" IS THE TYPE OF PEOPLE
SAY EMPLOYED, EG. "CLERICAL", AND "ARG2" AND "ARG3" ARE
SAY THE COORDINATES OF THE LOCATION OF THE RECREATION.
SAY IT IS CALLED BY SAYING "RD RWHERE P (ARG1 ARG2 ARG3)".
READ_CONSOLE
RWHERE P
TABSET 30
Q = 0.
EXPAND +$2ARG1 ** EMPLOYEES FROM THE FOLLOWING LOCATIONS/+
EXPAND +USE THE RECREATION AT LOCATION $3ARG2,$3ARG3/+
THROUGH WH1 FOR I=, J= EACH LOC ON ARG1
IF ARG1(RECLOCI I,J) .EQL. ARG2 IF ARG1(RECLOCJ I,J) .EQL. ARG3 '
GOTO WH2
GOTO WH1
WH2$ CONTINUE
Q = Q + ARG1(RECNUMB I,J)
EXPAND +FROM $31*,$3J.$1ARG1(RECNUMB I,J)/+
WH1$ CONTINUE
EXPAND +TOTAL AT $3ARG2,$3ARG3.$1Q/+
READ_CONSOLE
PAGE 128
CHAPTER V: SIMULATION
IMAP P
SAY THIS FILE MAPS THE SUBSET OF ALL RECREATION LOCATIONS
SAY IN WHICH MORE THAN "ARG1" EMPLOYEES USE THE RECREATIONAL
IT MAPS THE LOCATIONS OF EMPLOYMENT WHERE
SAY FACILITIES.
SAY THE USER CAME FROM. AND IT REFERENCES THESE AGAINST ALL
SAY EMPLOYMENT AND RECREATION PLACES. IT IS CALLED BY
SAY SAYING "MAP P (ARG1)" WHERE ARG1 IS A CONSTANT.
READ_CONSOLE
MA P P
SUBSTITUTION ON
TLIST1 *=*
TLIST2 *=*
NULL
NULL
THROUGH MAl FOR I=,
SUM = 0.
FOR Q = 1.
J= EACH LOC ON RECREATION
STEP 1. UNTIL 3. SUM = SUM + RECREATION(Q I,J)
IF SUM .GRT. ARGI PUT TLIST1 I,J
MA1$ CONTINUE
THROUGH MA2 FOR R=2. STEP 1. UNTIL 4.
THROUGH MA3 FOR 1=, J= EACH LOC ON R
F = R(RECLOCI I,J)
G = R(RECLOCJ I,J)
IF R(RECNUMB I,J) .GRT. 0.
MA3$ CONTINUE
MA2$ CONTINUE
ALLWORK *=* OFFICE .OR.
ALLREC *=* RECREATION
SUBREC *=* TLIST1
SUBWORK *=* TLIST2
IF TLIST1(F,G)
PUT TLIST2 I,J
INDUSTRY
MAP (SUBREC SUBWORK; ALLREC ALLWORK) 1-20 1-20
READ CONSOLE
PAGE 129
CHAPTER VI
FUTURE DEVELOPMENT OF DISCOURSE
As it stands right now, DISCOURSE
offers the user a data
structure which can be easily interpreted as a geographic grid
of LOCATIONS with
and
LOCATIONS
example,
to the
properties (or ATTRIBUTES) attached
numbers
to
attached
(for
ATTRIBUTES,
the
400 residents at location i,j).
An ATTRIBUTE, ("residents" in this
example),
is the name
by it.
ATTRIBUTES
of LOCATIONS which are described
of a list
can have numerical
or nominal values associated
two
which vary
those
types:
with them of
different LOCATIONS
with
and
those which remain constant regardless of LOCATION. The system
also accommodates
and
temporary variables which can
alphanumeric values.
All elements
have
have numeric
both names
and
values.
There are three dictionaries:
temporary variables.
accessed through
Any datum in
LOCATIONS, ATTRIBUTES and
the
these dictionaries by
system can and must be
means of the
name of
the element in which the datum resides.
Reflecting the designer's need to add new data and modify
the existing data, DISCOURSE has a
this purpose,
data at any
hoc and to
of the
and for storing
time.
full range of commands for
the different "states"
Reflecting his need to
manipulate data ad
manipulate according to rules all
description, DISCOURSE
of the
of the elements
contains logical,
relational,
including
conditional
The last
mentioned are
list and
reflecting the
array structures:
the values
iteration
of two
types
one which
looks
and the other
member of ATTRIBUTE lists
successively at each
which increments
and
branching,
(looping) statements.
of control
a full range
operators as well as
and arithmetic
statements
PAGE 130
DEVELOPMENT
CHAPTER VI:
and which
of a variable
can be
used to examine specific parts of an array.
DISCOURSE features a flexible
which
"FORMAT"
statements.
designer to
The
specify the way
requests
the
for
which
designer,
output facility called "EXPAND"
FORTRAN's
of
combination
a
resembles
"map"
displays
map-like
and
tabular
produce
and
its "show"
to
In addition
statement
EXPAND
he wants
(see figure (4.16), page 103) EXPAND,
"WRITE"
and
permits
the
information displayed.
also, is the first step
in the direction of putting information contained in DISCOURSE
in such a form that it can be used as input by other systems.
And, finally, reflecting the designer's
language
for
himself
simpler than
basic
requests
of
designer, chooses it to be,
DISCOURSE, but exactly as he, the
DISCOURSE contains
the
need to create a
facilities for
writing subroutines
(RCOM
files) using the basic requests of the language.
In
the first
version
command "center" executed a
in a language
DISCOURSE, for
it is
the
found the "center
a geographically distributed set
II
example,
particular algorithm, implemented
invisible to the user, which
of gravity" of
In DISCOURSE
of
possible to
of elements.
write and name
one's own
CHAPTER V
DEVELOPMENT
"center"
gorithm.
language,
FORTRAN.
like
A
the
It
PAGE 131
can
then be
writing
of
similar increase
used
as
subroutine
in the
part of
subprograms
richness of
the
in
facilities
available to the system user has occurred in the other aspects
of the
system, including more and
general purpose
has increased
required
computer language.
the possible
because the
more features
more of the features
richness of
users have
be accessible to
increase of complexity,
This line
of a
of development
usage;
it
has been
consistently demanded
their use. But,
that
despite this
the language remains possible
to use
for some purposes without having to "program" in the language.
The language of DISCOURSE has changed from being somewhat
specific to
its initial uses to
being much more
general but
still allowing the individual user to develop his own specific
style.
Any
which may not be best for what
may
into a way of
language forces the user
be difficult
to
subtlety of expression;
use, but
using it
he is doing; a richer language
in it
an awkward
the
user may
achieve
language may prevent self
expression altogether.
Much of the
time using DISCOURSE is
files, that is, building a higher level
specific to
the design
problem than
spent building RCOM
language which is more
the basic
requests and
which is more specific to what the designer wants to do at the
moment. The dilemma posed by the apparent conflict between two
of the goals for the development of the language - richness of
the language
permitting subtle
expression and
simplicity of
CHAPTER VI:
DEVELOPMENT
PAGE 132
the language permitting ease of usage - may be resolved in the
future through
the development
of problem
oriented features
for the system; and by developing prototypes of the use of the
system.
I.
PROBLEM ORIENTED FEATURES
features are
Problem oriented
functions whic h
the designe r can
ready made
techniques or
use in connection
with his
work, techni qu es requiring a minimum of alteration to suit his
ne eds. Although
particular
system to-da te
built by
the experience
of
limited (6.01) there seem to
is
se;ve ral users
whi ch are
usage of
the
be RCOM files
sufficiently similar
that
they or elem ients of them cou Id have been provided in the basic
language.
T*he result would have
without loss
features wi
0
f subtlety of
been increased ease
expression.
of use
The problem oriented
range from tho se which have a clear reference to
particular typ es of problems to those which do not.
Examples of
intensity ratio"
one end of
the range
might be a
"density of
function or a
"land
use
persons residing
per square foot of residential construction." Although many of
these
are easy
to
presence ready made
user
would help
useful to
write in
the language
in the list of features
him to
begin to
him during the course
use the
as
it is,
their
available to the
system, might
of his designing,
be
and would
serve as models for his own making of techniques better suited
to his own needs. Examples at the
a
function which
computes the
other end of the range are:
equivalence of
two lists
of
CHAPTER VI:
DEVELOPMENT
LOCATIONS v,
ch share commo n
an average
stance function, and
they can be written
PAGE 133
ATTRIBUTES, weighting functions,
in the
many others which, although
anguage , the wri ting of t:hem does
desi gner with respect
fruits o f insi ght to the
not yield ainy
to his desi gn problem. On the other hand, the "compact:ness" of
a
pattern
or the
implemented in
"mi x"
of
activ itie s might
two
a varie ty of
well
be
would s tand
differ en t ways, and
for many di fferent desi gner-d efined functions. The tactics for
developing "problem ori ented feature s"i will be:
DesiLr
Problems :
considerati ons
problems,
which
look ing
come
up
the
in
connection
kinds
the
for example,
at
associated with design for ve ry
kinds
specific
design
w ith
f accounting
of
procedures
den se urban areas: floor area
ratios, parking ratios and so on.
looking at the functions us ed over and
Design Functions:
over again
system.
the
by designer s
As can be seen in the
geometric
designers.
"nearest
in connection
request s
Many
of
of several"
which would have
ha ve
these :,
could be
with their
using the
examples in Chapter s IV and V,
be
had
to
like
"average
written
written
by
distance,"
in generalized
serve d in every situation.
the
or
form
Some implemented
already in DISCOURSE as RCOM files should probably be included
as
basic commands
applicability will
in the
language; others
probably be left
of less
general
in the DISCOURSE language
with instructions for use.
Data
Structure:
looking
at the
data
structure as
it
PAGE 134
DEVELOPMENT
CHAPTER VI:
there are more types
presently stands and determining whether
of
briefl ,
structure
the data
Reviewing
easil
could
which
manipulations
on
be
written
it
features :
it.
two
a
dimensional array (the GRID) which can have members ( ATTX LOCS)
vectors; and temporary
be examined
made available
to array compari sons.
can be
the results can be
of lists
are the
according
to
and list
Many other examp les can be created from
fe atur es.
different combin ationis of these
then be looked
if it
comp aris on of vectors
CHARVARS,
values contained in
features can
ate r use. Examples
orderi ng
rank
if
and
for immediate or
of 1 ists,
comparison
be man i pulated,
easil
features
other
with
compared
like
be treated
variables. Eac h of these
if it can
to see
as well
which can
and CHARVARS)
(CHARCONS
assoc iated
have values
ATTXLOCS can
and the
the ATTRIBUTES
the GRID;
associated with LOCATIONS in
of lists (ATTRIBUTES)
at fo r possible emp iric al
Each example would
interpretations and
for its usefulnetss to the designer.
Techni ques to
been useful
candidates
are
constru(=tion and
matrix
may
the techniques
mathema tical
techniques
which have
survey ing the kinds of design
in c)ther fie lds and
to whi ch
problems
looking at
Borr ow:
optimization
manipulation. Some
techniques exisit alread y in CHOICE
Obv ious
be ap plied.
and
techniques,
of these
(6.02), a system
la tter
to help
the designer in the eval uation of several desi gn alternati ves.
the
technique s
may
manipulations
facilitie s
other
Many
of
require
data
than those
structures
which
exist
and
at
CHAPTER VI:
present in
whether new
issue of
The
DISCOURSE.
facilities
DISCOURSE
data from
or the
into DISCOURSE
be built
should
PAGE 135
DEVELOPMENT
system and returned is one
taken out, processed in some other
which will have to be resolved on an individual basis and will
to the particular implementation
depend on factors pertaining
of
such as
system
the computer
central processing
in the
storage available
programming
time and
free
other
unit for
programs.
finding ways
Interface:
terms of
used;
with data
from
from various kinds
example, the census or
other sources, for
of surveys:
interface
to
characterizing each of the sources identified, in
and developing techniques,
programs and, where needed,
will help to ease the
utilizing other computer hardware which
large quantities of data
burden of entering
to get
are easy
of data
Small quantities
least conventionally
are at
and how they
its data
into the system.
into the
system:
witness the several examples in Chapters IV and V.
II.
of difficulty inhibiting the
The second type
system to-date
has stemmed
implementation but from
its
uses.
use, examples
a lack of well
which strongly
parts
he
about.
These
nor its
explained examples of
suggest
whole classes
of
help the designer both to
design problem and to
thinks
use of the
its design
neither from
Such "use prototypes" should
think about his
the
USE PROTOTYPES
implement in DISCOURSE
"use
prototypes"
will
probably result from two approaches: from the use of DISCOURSE
CHAPTER VI:
by city
DEVELOPMENT
designers in
private firms;
PAGE 136
the field working
and from
in agencies
continuing efforts
and for
to analyse
the
process of design.
The functional analysis of city
a first
intended to b e
step toward building a
library of use
The 1 ibrary will probably contain some prototypes
prototypes.
which
design in Chapter III is
are close
"transformati on"
functiona I
to the
and "testing";
types of
and
"description,"
others which
describe
d uring the course of design.
patterns of work by the designer
Subcategories within each type wi
probably be developed from
different emp irical applications.
Many of the first type have
illustrations in Cha pter III.
been used for
to implement them on the system,
them
Some
where necessary.
What remains is
i mproving them and adding to
of t he
second
type should
be
developed in the context of a des ign project and will probably
range of techniques us ed to
result in a
t he time;
problem which was being faced at
should be developed in
second type
address a particular
and others of the
trying to
the context of
explore some specific idea like "adaptability" (6.03)
Ill.
The most
variety
needed improvement
of displays
changed by
time
GRAPHI CS AND MODES OF REPRESENTATI ON
and
the designer
has to
in his
head,
description
for dis plays
means of a graphic
system is
in the
which
input device.
carry
system, or wait frorr time to time
map
ca n be
At
easily
the present
changes to
the
independent of
the
the latest
update a
a richer
for the console to type out
the
PAGE 137
DEVELOPMENT
CHAPTER VI:
latest map.
display cathode
With
the
ray tubes
recently purchased
it will
the
through
sequences
help to protray "growth"
time,
over
envi ronment
be possible
to create
and "change"
patterns
activity
envi ronment,
the
(6.04)
a
ATTRIBUTES of the descri pt ion.
continuous display of selected
This feature should
ARDS
as
achieving
of
well
as
of
the
some
effects o f true "dynamic" dis play (6.05).
Addi tional features
ARDS
and which
display and
side the
firs
major
t,
"pointing" feature
the
areas, ATTRIBUTES,
map displa y
specified func tions;
for example,
readily be di
made availabl
the maps
to enl arge the grid
as well
Tlay
any element
or will be in
the very near future. A second
copy display
photography so that
maps can be
symbols. These
ar e either und er development
tures
hard
from
or to
da ta can
so t hat numerical
and related f
is
for
be shifted in scale,
pres en t map of
as the
which will be developed when
of the
and names listed along
third, outli nle and tone
I ayed i;
which
He can point a t,
refe r to
which may
second,
p resent
improvement over
to introduce data manually
LOCAT I ONS,
on the
direct his interrogation
the designer to
will perm it
example,
represent a
output are:
input and
whi ch are readily programmed
at present
kind of d isplay
the ha rdware has been identified
a
pri nter,
the designer ca n have
plotter and
or
by
a permanent record
of his work.
Other features which may
in the
system to stimulate
help through their availability
the development of
graphics are:
the
permutations
inversion,
transformations: mirror,
various symmetrical
change of
rotation, translation,
and
PAGE 138
DEVELOPMENT
CHAPTER VI:
of those
with
the combinations
scale and
and
to two
respect
three
and plan to
dimensional forms in two dimensional
projection;
section to persepctive viewing along
with a symbolic language
designer that he is looking at
which will make evident to the
the same form in the different views.
Not so clear, however, is
the development of support for
Some of
the less easily described activities of the designer.
the time spent designing is spent working with pencil on paper
making sketches, the
some of which may be in
section or perspective, and others of
diagram relations
which may
plan view, but
majority of which are in
which have
no direct
little or
spatial analogue.
A sketch
the other,
one into
elements, sometimes
consists of
but
a group
finely graded
members of
of elements the
which can generally be distinguished
one from the other.
The
elements are positioned in two dimensions or in three, as in a
rough
And,
model.
in the
variation
variations in
line, a
the introduction
with
same type
the intensity
of element,
of color,
still further richness
of
range
a
through
often done
tone, or
of
strength of
of the representation
can be
achieved.
Metric and Non-metric Representations
These
which, in
elements are
the case
represented in
of city designers,
a spatial
framework
is usually
metric or
PAGE 139
CHAPTER VI:
DEVELOPMENT
non-metric.
Each of these types of
rules.
spaces has its own set of
In metric space, distance is interpreted as the linear
(straight line)
ruler;
when
distance between points
non-metric
distance is measured
from the
other.
refers to a
space is
as measured
represented
by
by the number of nodes away
In metric space,
unit of distance on the
if the
by some
a
graph
one node is
unit of
distance
land, the representation
can be mapped point for point onto the land.
If the representation
is in a particular
space, for
example euclidean
space, then
euclidean
geometry
applicable.
representation
graph,
then
is in
all
of
become
non-metric space
axioms
the
of
kind of metric
all the
axioms of
(6.06)
as
If
represented by
graph
theory
the
a
become
applicable.
An
is usually
organization chart
space than metric space, though
Even so, some positional
better in
non-metric
the latter would be possible.
importance is usually ascribed to the
chart making it a mixture of the two types of representations:
organizationally;
top is up
in the organization,
sideways is indifferent
though often those in the
to rank
center of the
chart at the same level are more important because it is often
easier to draw the boxes with
more connections to them in the
center.
In the
of the
form is
metric representation, the possibilities
spaces represented into
more evident; in
a three
the non-metric
for fit
dimensional building
representation, the
PAGE 140
DEVELOPMENT
CHAPTER VI:
decision
organization's
an
officer in
some
of
centrality
than in metric space, particularly
making may be more evident
if the network of communication is very rich.
If
the
designer assigns
thought to the
elements and to the relations of
representation
then
may
he
conclusions
of
realm
in some
meaning
a
some mode of
draw
valid
from
inferences in the representation are true for the realm he has
been referring to.
organization to
For example, if he assigns positions in an
message to
sending a
graph and the
the nodes of a
receive
a message
through
the vice
the treasurer,
from
president;
message directly to the
then for
the links,
possibility of
it must first
the president
but
to
the president
pass
can send
a
treasurer. The further interpretation
of the diagram can be made that the president has more "power"
and the
vice president
more information
assumptions about the directional component
if the
appropriate
of the arrow have
been made.
Likewise in
may
be
a metric
associated
representation:
associated with
for
of reality
representation aspects
with
rules
the
example,
greater
greater likelihood
which
govern
proximity
of communication
may
the
be
between
two groups in the population.
Distortion of Metric Representation
Change in the function
mapping the metric representation
to a geographic region will result when distance in the metric
representation refers to time or, perhaps,
time and a mode of
like to
might
distance of
the same
map would look very
The resulting
was within five
He
might then
different from the
designer in
the
minutes
five
group.
map of
of geography may be
Distortion
geography would be distorted.
informative to
A
representation would be metric,
Although the
distance alone.
alone.
s pace the proximities so derived.
wish to represent in metric
very
was within
who
know
a certain group; and who
walking distance of
minutes driving
geographic distance
than to
transportation rather
designer
PAGE 141
DEVELOPMENT
CHAPTER VI:
reveal
it might
that
relationships which were not previously obvious.
Mixed Representations
need to have a geographic
some purposes, in that application,
reference.
The designer
and values
with each
nodes, its links, and
network by
of the
its faces;
nodes,
and he needs to
find
o r the longest
the shortest
the network:
elements of
see what is on or adjacent
want to
to associate names
needs to be able
links toward
along
move
and, for
represent road networks
be used to
Graphs can
his way
path.
to any of
its
be able to
through
the
also will
He
its elements.
The last requirement probab ly means that the network will have
to
be
overlayed
requirements imply
that employed
were added
used
for
graphical
t.he
onto
a data
so far
to handle
graphical
analysis it
geographic
structure somewhat
in D ISCOURSE II.
If a
netwrorks easily than
analysis.
If
is important
to
different from
data structure
it might
networks
first
The
grid.
are
also be
used
have flexibility
for
in
DEVELOPMENT
CHAPTER VI:
attaching
meanings
represent
a
PAGE 142
to
the
for
elements,
a
system,
communications
a
can
network
power
of
diagram
Each
relations, a family structure, or a logical proposition.
interpretation should
system
just as
it
be possible to
is now
possible
in the
make explicitly
to
expli cit
make such
interpretations of the data in DISCOURSE.
Three general principles for the
the experience of build
representation seem to emerge from
first,
DISCOURSE:
a
that
full
development any mode
manipulations
range of
needed which is proper to that mode.
With respect to a grid of LOCATIONS, the major categor ies
of
manipulations
adjacency,
seem
be:
or ientation
and
and
functions.
proximity,
res pect
With
connection, and
other manipulations
all
at,
existence
at, direction,
network: existence
functions
to
a
proximity
on
based
to
matrices.
(6.07)
The s econ d principle
distinguis habl e
possibilit y of
is
element
in three parts:
of
the
mode
first, that each
selected
have
the
having values associated with
being named, and
that name whic h are specific at the named elements, and having
values
wh ich
respective ly);
manipulabl e:
true
the
to
(analogous
(logical,
are
second,
ie.,
all
similarly
named
elements
CHARVAR,
and
CHARCON
ATTRIBUTE,
that
that
rel ational,
DISCOURSE be
of
usable in
and
the
all
names
parallel
set
arithmetic) which
the new mode;
and
values
of
exist
and third,
be
operators
now
in
that each
PAGE 143
DEVELOPMENT
CHAPTER VI:
such that its name calls
element should by individually named
element
name,
its
and
LOCATION
the
to
(analogous
particular
with that
associated
information
all the
forth
its
coordinates).
is that
A third principle, even more tenatively advanc ed,
possible
another
obtained
the information
such that
onto
one can
be
a
network
is
if
LOCATIONS on that
will be associated
one of the links passes
grid across which
user.
the
contained in
geographic grid, then the
overlayed onto a
and,
to
example,
For
other.
the
from
translation
one representation
means laying
"Communication," here,
made
accessible
be
communication,
possibly,
such
for
rules
the
that
and
shou ld be
between modes
and communication
translation
to that link such that when the link is queried for what falls
"along"
LOCATIONS will
the
information
the
it,
be available. And,
of
possibility
representation
matrix
is meant
by "translation"
for
passing from,
a
to
geographic
those
at
stored
example,
of
representation
a
network
the
same
material.
DESCRIPTION OF THE ENVIRONMENT
IV.
The present system seems to require constant reference to
two coordinates, and the
LOCATIONS by their
as
the
objects of
properties
inhibit
are
the use
the
description
attributed.
of
the
The
designers' traditional modes
to which
first
because of
use of LOCATIONS
its
of working;
qualities
characteristic
lack
and the
of fit
or
may
with
second may
DEVELOPMENT
CHAPTER VI:
PAGE 144
inhibit subtlety of description.
When a designer speaks of "places" he is usually speaking
of
Auburn
one
For example
alone.
necessary if
geographic
speaks of
not of
Street but
would be
combinations,
LOCATION
ATTRIBUTE and
one were
symbol of the whole,
used as the
act ivi ties and the people present
designer seldom
if ever has to
can usually des cribe
using measures of
Some coordinate
Square, or
the
label
in a
for
may be defined more
use a coordinate
scheme.
is
The
He
of common knowledge
somebody a map
to understand "where"
it is.
geogr aphic functions
However, this
not imply that the coordina te scheme
a
by the
than by the geography.
necessary if
are to be compl etely descri bed.
which
the place
p roximity, or he can "show"
sys tem is
Mt.
center, whichever
a place from points
the o ther person
which enables
is
or its
LOCATIONS
and longitude
to pinpoint
Square
Its limits
geographic region.
Harvard
the latitude
Harvard
sense.
not the
statement does
need be prominent
in the
language which the designer uses.
Ways
of
maak i ng
the coordinate
include developing graphic input
of the burden of typing data
sys tem
less
prominent
procedures to eliminate much
in according to coordinates, and
developing new techniques for describing the environment which
rely less on
places.
coordinates, but instead rely upon
To illustrate
this possibility
see
the names of
the example
in
figures (6.01) and (6.02).
Required is a special category of ATTRIBUTES which can be
CHAPTER VI:
PAGE 145
DEVELOPMENT
thought of
of the description,
as objects
specialty shops?"
"where are
designer asks
"Harvard Square"
so that
the answers
Street, " not
and "Mt. Aubu rn
4,5." Other operations need to be
when the
are
"location 3,4
thought through in the same
way and can probably also be simulated on the present system.
It
is
and
name
probably not only LOCATIONS to which designers give
The transpa rency
sense.
the
when one has to
structure clouds somewhat
and activities to LOCATIONS as
It
of
DISCOURSE
data
ascribe facilities
ATTR IBUTES of those LOCATIONS.
should be possible to more
chly descr ibe a physical world
DISCO URSE a building type
to which we then ascribe meani
SIn
is an ATTRIBUTE of a LOCATION;
the activ ity in that building
th e LOCATION .
must also be an ATTRIBUTE of
be
an
ATTRIBUTE of
buil ding.
the
Instead it should
There should
objects of the description whic h have
be
other
the same status in fhte
language as the present LOCATIO NS.
description
is to
solution
One possible
capable
havin g
of
these objects might
the existing
ATTRIBUTES
be ATTXACT and ATTXFAC,
ATTXLOCS.
assigned,
th e existence of an
ACTIVITIES and FACILITIES;
like
ATTRIBUTE at
corresponding to
these would have
And
of the
create new objects
to be
ways of
associating LOCATIONS ACTIVITIE S and FACILITIES in relation to
how the designer
thinking about
wanted the en vironment described.
this possibilit y is
for purposes which are non
to use some
A way of
of LOCATIONS
geo graphic and not consistent with
_
the rest of the representation
--
o
which would be geographic, for
CHAPTER VI:
PAGE 146
DEVELOPMENT
the LOCATIONS.
Again, to illustrate the idea but not to prove its worth,
the activity unit, a family named "Smith," might be equivalent
to a
"father" might be
could
have
location,"
"Smith"
family's
income residents,"
of "Smith."
two of the ATTRIBUTES
several
all
"Middle
"ij."
LOCATION called
CHARVARS:
"income,"
"age,"
specific
to
that
place
of
residence,
recreational facilities
particular
and shopping
and
habits of
or
"Father"
and
"work
The
family.
the
use
each of
of
its
members could be specified and manipulated.
Summary of Future Development Plans
The test of the DISCOURSE language is its capacity to
accommodate what the city designer does. The data structure
and manipulations
of DISCOURSE
- its
basic features
- have
been developed to the point where where they can accommodate
many aspects of the city designers' work, but there may be
many more, some of which may not be accommodated in the
language as it stands. The discovery of these can be greatly
hastened by having the system used by many designers in
been most involved in its
addition to those who have
Improvements of the present system are
development to-date.
These
increase its ease of usage.
needed, therefore, to
improvements include adding a wide variety of displays, adding
functions which can be written on the DISCOURSE data structure
which increase its convenience of usage, and developing clear
instructions for and examples of the use of the systeri.
two
major difficulties of the
There are, however,
language which are presently evident and which challenge its
it is hard to accommodate the designer's
basic features:
activity of "sketching;" and it is hard to describe "objects,"
from
distinguished
as
"activities"
or
"facilities,"
by
resolved
be
may
difficulties
These
"LOCATIONS."
the
of
objects
new
and
representation
of
modes
introducing new
description. Such changes may point toward the next stage of
development of DISCOURSE.
PAGE 147
design2 p
IF DESIGN SYSTEM CRASHES GIVE THE DEBUG COMMAND:
DUMP NAME 12000 3D000
R 0.03/0.03
comment That loaded the DISCOURSE system into core.
R 0.01/0.04
ready off
comment That turned off the time message.
rd attsdata p
comment That read in the data file.
list attributes
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
ATTR
0 -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 -12 -13 --
NULL
UNIVERSE
HARVSQ
MTAUBURN
A
B
MASSAVE
C
D
E
F
FAKENAME
SPESHOPS
BRICKFES
comment ATTRIBUTES 2-10 are place names.
rd ltryatts p
comment That read in the "macro"; now it's ready to use.
ex whereare (speshops)
SPESHOPS EXIST AT HARVSQ
MTAUBURN
ex whereare (brickfactories)
BRICKFES DO NOT EXIST AT THE NAMED PLACES
foodstores
*=*
harvsq
.or. mtauburn .or.
massave
ex whereare (foodstores)
FOODSTES
EXIST AT
Figure (6.01)
HARVSQ
MTAUBURN
MASSAVE
PLACES IDENTIFIED BY NAMES NOT COORDINATES
printf Itryatts p
PAGE 148
COMMENT THE REASON THIS WAS SET UP AS A "MACRO"
COMMENT WAS SO THAT IT COULD BE CALLED WITH A SINGLE NAME
COMMENT AND SO THAT IT COULD BE CALLED FROM OTHER RCOM FILES.
BEG IN MACRO WHEREARE
COMMENT THIS SETS TABS TO POSITION THE PRINTED OUTPUT.
TABSET 20
COMMENT THIS ALLOWS ME TO SUBSTITUTE ATTRIBUTE NUM.BERS
COMMENT FOR THEIR NAMES.
SUBSTITUTION ON
COMMENT Q HELPS ME TO KNOW WHERE I'VE BEEN IN THE PROGRAM.
Q = 0.
COMMENT THIS GOES THROUGH THE ATTRIBUTES WHICH ARE PLACE NAMES.
THROUGH WH1 FOR 1= 2. STEP 1. UNTIL 10.
COMMENT THIS INTERSECTS "ARGI", THE ATTRIBUTE I AM ASKING ABOUT
COMMENT WITH EACH OF THE PLACE NAMES.
I
TLIST *=* ARG1 .AND.
COMMENT IF THE INTERSECTION DOES NOT EXIST THEN IT BRANCHES.
0. GOTO WH1
IF TLIST .EQL.
COMMENT IF "ARG1" HAS BEEN FOUND TO EXIST AT SOME PLACE THEN
COMMENT THE PROGRAM SKIPS THE NEXT THREE LINES.
IF Q .EQL. 1. GOTO WH2
COMMENT THIS IS THE FIRST TYPED LINE IF "ARG1" IS ANY PLACE.
EXPAND +$2ARG1 EXIST AT $2FAKENAME(I)+
Q = 1.
GOTO WH1
WH2$ CONTINUE
EXPAND +**.$2FAKENAME(I)/+
WH1$ CONTINUE
COMMENT IF THE PROGRAM HAS GONE THROUGH ALL THE PLACE NAMES
COMMENT AND "ARG1" DOES EXIST AT ANY OF THEM THE PROGRAM SKIPS
COMMENT THE FOLLOWING LINE.
IF Q .EQL. 1. GOTO WH3
EXPAND +$2ARG1 DO NOT EXIST AT THE NAMED PLACES/+
WH3$ SUBSTITUTION OFF
RETURN
END_MACRO
READ_CONSOLE
R; T=0.06/0.84
Figure
(6.02)
13.34.27
FILE TO REFER TO PLACES WITH NAMJES
PAGE 149
APPENDIX
AN EXPERIMENT
DOCUMENTATION OF
IN THE REPLICATION OF A DESIGN
This Appendix contains some of the evidence of the work
the desinn for Ciudad Guavana.
The
on
"Fi rst" pattern which
s referred to in the illustrations is
the pattern exhibited i Chapter II. The "Second" pattern was
the chan nel
generated later with a ingle change in the data:
was moved from location 4,12 to location 4,11. The reason for
the change was that if the channel were in 4,12, then 5,13 was
"beside" i t according to the decision rules whereas in fact
location 5,13 is not nea r the channel.
This adj acency was unintended by the author and it was,
therefore, changed in order to improve the "accuracy" of the
data. The c hange, given the same assignment orderings and
decision rule s, was enormous.
Urbanization in the "Second"
the
pattern stay ed almost entirely on the eastern side of
than the
river, much more like Ciudad Guayana 1969 (p 152)
design. In th e "First" pattern, "industry" had been assigned
first to 5,1 3. If the single change in the data caused the
difference in the generated pattern, (further examination may
reveal other reasons) then it shows the importance of initial
judgments of the designers in describing the environments they
will design.
the
repl
of
ication
TABLE OF CONTENTS FOR THE APPENDIX
Ciudad Guayana,
1961
PAGE
151
Cludad Guayana,
1969
152
Design for Ciudad Guayana
by the Designers of the Development Corporation
153
Application of
to the Initial
154-157
the Decision Rules
State of the Environment,
1961
The Record of the Decision Paths
Taken through the Decision Trees (p42)
158-159
First Pattern for Ciudad Guayana
Generated by the Decision Rules
160
Comparison of the First Pattern Generated
by the Rul es with Designers' Pattern
for Urbani zation
161
Comparison of Urbanization by
Order of Assignment
162-163
APPEND I X
PAGE 150
Comparison of the First Pattern and
the Designers' Pattern for Industry
164
Comparison of industry for the Assignment Order
165-166
Comparison of the First Pattern and
the Designers' Pattern for Center
167
Application of the Decision Rules
Generating the Second Pattern
168-172
The Record of the Decision Paths
Taken through the Decision Trees (p42)
173
Second Pattern for Ciudad Guayana
Generated by Decision Rules
174
Comparison of Second Pattern with
Designers' for Urbanization
175
Comparison of the Second Pattern with
Designers' for Industry
176
Comparison of the Second Pattern with
Designers' for Center
177
Decision Rule for Urbanization
178-179
Decision Rule for Industry
180-181
Decision Rule for Center
182
Ciudad Guayana Data File
183-184
PAGE 151
APPENDIX
1
123456789012345678
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
+.*.*.
2
+.+.*.
*.+.+.+.+.
*.*.*.
*
.
*.*.*.
2
.
*0
.
.
.
.
.
1
.
.
.
.
.
.
0
..
.
.
.
0
.
.
0
.
.
0
0
0
0
0+.
.
.
.
.
.
.
•
•
•
·
•
o'l'
.+0
.+
.
.
.
.+0+-.
.
.
1
.0
.
.+.
.
.+.+0
+.+
.
1 .
.
.0
..
.
+.+.+.
.
1
.
1 .
. 0 . . . . 0 .+0+.
*. . 2 2 ..
1
2+. .
1
.
1
.0
+"
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
123456789012345678
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS - -
'1' = URB1961(20);
'2'
IND1961(21)
'.'
= NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+' = RIVER(8); '-'
CHANNEL(12);
'*'
= CHANNEL & RIVER; ' ' = NO REF ATTS
CIUDAD GUAYANA --
1961
SHOWING URBANIZATION AND INDUSTRY
T7
PAGE 152
APPENDIX
1
123456789012345678
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
*. .
1
2 1
2
+.*.*.
+.**.*. 1 1 1 1
+.+.*. 2 1 1 1 1 1
*.+.+.+*+. 2+. . 1 1 2 . 1
*.*.*. . . 2 1 .+.+
......
*. . . 2 . . . 1 1 .+.+. . . .
*.*.*. . 2 2 . . 1 .+.+. 1 . ..
*.*.
2 . . . . . . . . 1+.+0 . . . .
*. . 2 2 2 . . . .
0+.+6 .
. . .
.
. . . 2 . . . . . . +. .
. . . . . . .+.+.+ .+. .
.+.+.++ .
. . . . . +.
. . . . . +.
3
4
5
6
7
8
9
10
11
12
13
14
15
1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS -
-
'1' = URB1969(13);
'2'
IND1969(14)
'.'
= NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+'
= CHANNEL & RIVER; '
'*'
CHANNEL(12);
CIUDAD GUAYANA --
SHOWING URBANIZATION AND
= RIVER(8); '-'
' = NO REF ATTS
1969
INDUSTRY
=
APPENDIX
PAGE 153
1
1234567890123t5678
1
2
+0.*.*
3
4
5
6
7
1
+.*.*.
8
9
10
11
12
13
14
15
+.*.+. 2 1
*.+.+.+. +. 2+. 1 1
*.*.*. 2 2 2 1 .+.+. 1
*. . 2 2 1 1 . 1 1 .+.+.
*.*.*. 2 2 2 1 1 3 3 .+.+.
*.*. 2 2 . 2 2 3 3 .1 1+.+.
.
*. 2 2 2 2
1 1 1 . 1+.+
. 2 2
2 2 1 1
1
.
1
1+.
1
1 1
1 2
. .
.
1 .
. .
.
2 2 2 2 1 1 1+.+.+.+.
0
a
.0
f
.+.+.+.
.
123456789012345678
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS - IDESIGN(2); '3' = CDESIGN(4)
'.'
'1' = UDESIGN(3);
'2' =
= NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+' = RIVER(8); '-'
'*' = CHANNEL & RIVER; ' ' = NO REF ATTS
CHANNEL(12);
DESIGN FOR CIUDAD GUAYANA
BY THE DESIGNERS OF THE DEVELOPMENT CORPORATION
SHOWING URBANIZATION, INDUSTRY AND CENTER
design2 p
IF DESIGN SYSTEM CRASHES GIVE THE DEBUG
DUMP NAME 12000 3D000
R 0.02/0.02
rd dataguay p
R 0.78/0.80
rd macsguay p
rd assignu p
"URBANIZATION" HAS BEEN PLACED
DECISION PATH WAS Ul U2 U3 U4
3,17
d@rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 8,11
DECISION PATH WAS U1 U2 U3 U4 U7
assignu p
ERROR 1001
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 9,12
DECISION PATH WAS U1 U2 U3 U4 U6
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 5,13
DECISION PATH WAS 11 12 13 16 110
rd assign #i p
"INDUSTRY" HAS BEEN PLACED AT 9,4
DECISION PATH WAS 11 12 13 16 110
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 8,5
DECISION PATH WAS 11 12 13 16 110
rd assignc p
"CENTER" HAS BEEN PLACED AT 8,11
DECISION PATH WAS Cl C2 C6 C7
rd assignu p
"URBANIZATION" HAS BEEN PLACED
DECISION PATH WAS U1 U2 U3 U4
7,9
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 8,9
DECISION PATH WAS U1 U2 U3 U4 U7
rd assignu p
"URBANIZATION" HAS BEEN PLACED
DECISION PATH WAS Ul U2 U3 U4
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 7,6
12 13 16 110
DECISION PATH WAS 11
9,9
PAGE 154
T
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 7,7
12 13 16 110
DECISION PATH WAS 11
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 6,7
12 13 16 110
DECISION PATH WAS 11
rd assignu p
"URBANIZATION"
HAS BEEN PLACED AT 7,8
U2 U3 U4 U7
DECISION PATH WAS Ul
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 8,8
DECISION PATH WAS Ul U2 U3 U4 U7
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 9,8
U2 U3 U4 U7
DECISION PATH WAS Ul
rd assigni p
HAS BEEN PLACED AT 10,2
"INDUSTRY"
12 13 16 110
DECISION PATH WAS 11
rd assigni p
"INDUSTRY"
HAS BEEN PLACED AT 11,2
12 13 16 110
DECISION PATH WAS 11
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 11,1
12 13 16 110
DECISION PATH WAS 11
rd assignc p
"CENTER" HAS BEEN
PLACED AT 9, 12
C2 C3 C5 C7
DECISION PATH WAS C1
rd assignu p
"URBANIZATION"
HAS BEEN PLACED AT 8,7
DECISION PATH WAS U1
U2 U3 U4 U7
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 9,7
DECISION PATH WAS Ul
U2 U3 U4 U7
rd assicn##gnu p
"URBANIZATION" HAS BEEN PLACED AT 10,8
DECISION PATH WAS U1
U2 U3 U4 U7
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 4,15
DECISION PATH WAS 11 12 13 16 Ill
PAGE 155
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 2,17
12 13 16 111
DECISION PATH WAS 11
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 2,13
12 13 16 110
DECISION PATH WAS 11
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 11,7
DECISION PATH WAS Ul
U2 U3 U4 U6
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 12,6
DECISION PATH WAS U1
U2 U3 U4 U6
rd assignu p
HAS BEEN PLACED AT 10,7
"URBANIZATION"
U2 U3 U4 U7
DECISION PATH WAS Ul
rd assigni
p
"INDUSTRY" HAS BEEN PLACED AT 10,5
12 14 17 112
DECISION PATH WAS 11
rd assigni
p
"INDUSTRY"
HAS BEEN PLACED AT 10,6
12 14 17 112
DECISION PATH WAS 11
rd assigni
"INDUSTRY"
p
HAS BEEN
DECISION PATH WAS 11
PLACED AT 11,6
12 14 17 112
rd assignc p
"CENTER" HAS BEEN PLACED AT 7,9
DECISION PATH WAS Cl C2 C6 C7
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 11,5
U2 U3 U4 U7
DECISION PATH WAS U1
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 12,5
U2 U3 U4 U7
DECISION PATH WAS U1
rd assignu p
"URBANIZATION"
HAS
BEEN
U2 U3 U4 U7
DECISION PATH WAS U1
rd assigni p
"INDUSTRY" HAS BEEN
PLACED AT 12,4
PLACED AT 12,3
12 14 17 112
DECISION PATH WAS 11
rd assigni p
HAS BEEN PLACED AT 13,4
"INDUSTRY"
12 14 17 112
DECISION PATH WAS 11
PAGE 156
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 13,5
DECISION PATH WAS 11 12 14 17 112
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 13,3
DECISION PATH WAS UlJ
U2 U3 U4 U7
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 12,2
DECISION PATH WAS U1 U2 U3 Ul4 U7
rd assignu p
"URBANIZATION" HAS BEEN PLACEV AT
DECISION PATH WAS Ul U2 U3 U4 U7
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 1 L,5
DECISION PATH WAS 11 12 14 17 112
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 13,2
DECISION PATH WAS 11 12 14 17 112
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 14,3
DECISION PATH WAS 11 12 14 17 112
rd assignc p
"CENTER" HAS BEEN PLACED AT 8,8
DECISION PATH WAS Cl C2 C3 C5
rd assignu p
"URBANIZATION"' HAS B EEN PLACED AT
DECISION PATH WAS U1 U2 U3 U4 U7
rd assi-gnu p
"URBANIZATION" HAS BEEN PLACED AT 12,1
DECISION PATH WAS U1 U2 U3 U4 U7
rd assignu p
"URBANIZATION" HAS B EEN PLACED AT 14,2
DECISION PATH WAS Ul U2 U3 U4 U7
rd assigni p
"INDUSTRY"
HAS BEEN PLACED AT 15,3
DECISION PATH WAS 11
12 14 17 112
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 13,1
DECISION PATH WAS 11
12 14 17 112
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 14,1
12 14 17 112
DECISION PATH WAS 11
store d2guay p
PAGE 157
PAGE 158
list center
CHARVAR
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
1 OF CENTER
1 OF CENTER
2 OF CENTER
3 OF CENTER
4 OF CENTER
5 OF CENTER
6 OF CENTER
7 OF CENTER
8 OF CENTER
9 OF CENTER
10 OF CENTER
11 OF CENTER
12 OF CENTER
13 OF CENTER
14 OF CENTER
15 OF CENTER
IS CALLED ORDER
IS CALLED Cl
IS CALLED C2
IS CALLED C3
IS CALLED C4
IS CALLED C5
IS CALLED C6
IS CALLED C7
IS CALLED D8
IS CALLED D9
IS CALLED D10
IS CALLED D11
IS CALLED C12347
IS CALLED C12357
IS CALLED C1267
IS CALLED C17
for i= 12. 1. 15. show center(i)
CENTER(C12347) = 0.
CENTER(C12357) = 2.
CENTER(C1267) = 2.
CENTER(C17) = 0.
fo @for i=17. step 1.
until 22.
PAGE 159
show urbanization(i)
URBANION(U12346) = 4.
URBANION(U12347) = 20.
URBANION(U12358) = 0.
URBANION(U123359) = 0.
URBANION(U1210) = 0.
URBANION(U110) = 0.
for i=23. step 1. until 30. show industry(i)
INDUSTRY(112358) = 0.
INDUSTRY(112359) = 0.
INDUSTRY(1123610) = 10.
INDUSTRY(1123611) = 2.
INDUSTRY(112L4712) = 12.
INDUSTRY(1124713) = 0.
INDUSTRY(1I2414) = 0.
INDUSTRY(111I
•) = 0.
PAGE 160
APPENDIX
1
8 9 012345678
1234567
S*.
+.*.*.
.
2 2
1
2
+.*.*. 1 1
+.+.*.* 2 1 2.
*.+.+.+.+. 2+. 2 1 . .
*.*.*. 2 . . 1 .+.+ . ..
*. . 2 2 1 B . 1 1 .+.+.
*.*.*. 2 . 1 B 1 . B .++. 1
*.*.2 2 .. 1
1 . B+.+.
*.2
2 2 2 2 1 1. . +.+. .
2 2 .. 1 2 1 . . . ...
1 1 2 1 1 1 .+.+.+.+.
2 2 1 2 2 .+.+.+.
2 1 2 1 2 .+.
. . 2 1 . .+.
123456789012345678
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS = CENTER(5)
INDUSTRY(7); '3'
'B' = 1, 3
'.'
-
'1' = URBANION(6);
'2'
= NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+'
CHANNEL(12);
'*' = CHANNEL & RIVER; '
= RIVER(8); '-'
' = NO REF ATTS
FIRST PATTERN FOR CIUDAD GUAYANA
GENERATED BY DECISION RULES
SHOWING URBANIZATIONs INDUSTRY AND CENTER
=
PAGE 161
APPENDIX
1
123 4 567890123
4
5678
1
+.*.*. A A
2
.
.
.
.
.
+.+.*. . A 2
.+. 2 A 2
*+.+.+.+
*.*.*. . . . A .+.+.
2 .
* ..
.
A . AA .+.+.
*.*.*. . . 1 A A . 1 .+.+. A
*.* . . .
1 1 1 . 2 A+.+. .
A A 2 . 2+.+ ...
*. .....
. .
. 1 2 A 2 . 2 2+.
1 1 . 1 A A 2+.+.+.
.
+.
.
2
3
4
5
6
7
8
9
10
11
12
13
S
.
1
.
.
.
1
.
1
.
.+.
14
.
0
.
1
.
.+0
15
0+.+
123456789012345678
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS - UDESIGN(3)
'A' = 1, 2
'.'
'1'
= URBANION(6);
'2'
=
= NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+'
'*'
= CHANNEL & RIVER; '
CHANNEL(12);
'
= RIVER(8); '-'
= NO REF ATTS
COMPARISON OF FIRST URBANIZATION PATTERN
GENERATED BY DECISION RULES (URBANION)
WITH URBANIZATION PATTERN
GENERATED BY DESIGNERS (UDESIGN)
=
APPENDIX
PAGE 162
4
1
2
3
8
9
5
6
10
B
B
B
B
1
B
B
B
B
B
B
B
B
B
B
B
B
1
B
B
B
B
B
B
B
B
1
B
7
B
B
B
B
B
8
5
B
B
B
B
B
B
B
15
12
B
B
B
B
B
B
B
25
14
B
8
B
B
B
6
B
B
B
9
B
B
B
B
B
B
B
B
12
15
B
B
B
B
B
B
B
25
14
B
B
B
B
B
B
9
B
7
10
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
10
B
13
B
B
B
B
6
B
B
B
B
6
B
B
B
B
11
B
B
B
11
B
B
B
B
B
B
17
B
B
11
B
B
15
B
B
B
B
15
B
15
B
B
B
B
B
25
B
25
25
18
15
B
19
B
B
12
24
21
B
B
B
15
15
B
B
B
B
B
B
B
B
B
B
B
B
25
25
B
B
B
B
B
B
B
20
13
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
14 B
25
B
22
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
23
B
B
B
15
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
5
6
9
4
8
1
10
2
3
URBAN ION(ORDER), IUDES I GN(ORDERLO),
UDESI GN( ORDERHI)
HAVE BEEN GIVEN A VALUE WHERE THEY
IST
"B" INDICAT ES THAT THEY DO NOT EXI ST AT THAT LOCATION
TOP NUMBER IN EACH LOCATION IS THE ORDER ASSIGNED
BY THE DECI SION RULE
MIDDLE AND BOTTOM NUMBERS INDICATE TIHE DESIGNERS'
ACCEPTABLE RANGE OF ASSIGNMENT ORDER
6
COMPARISON OF FIRST URBANIZATION PATTERN
GENERATED BY DECISION RULES
WITH URBANIZATION PATTERN
GENERATED BY DESIGNERS
ROW G, COL 12 TO ROW 15, COL 10
PAGE 163
APPENDIX
1
11
B
B
B
2
3
4
5
6
7
8
9
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
1
1
1
3
B
B
B
2
10
5
B
5
13
B
B
B
B
B
B
B
B
B
B
B
B
B
12
14
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
14
B
B
B
B
B
B
B
B
B
1
1
1
1
1
1
B
6
11
B
B
B
B
B
B
B
B
B
B
B
B
14
15
B
B
B
B
B
B
1
1
B
2
5
B
6
11
B
B
B
B
B
B
1
1
1
B
B
B
B
B
B
15
16
B
B
B
B
B
B
I
1
1
B
2
5
B
B
B
B
B
B
B
B
B
B
B
17
18
B
B
B
B
B
B
B
B
2
6
11
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
18
B
B
B
B
B
B
B
B
B
B
B
B
B
5
6
7
8
9
10
B
10
16
17
13
11
URBANION(ORDER), UDESIGN(ORDERLO),AND UDESIGN(ORDERHI)
HAVE BEEN GIVEN A VALUE WHERE THEY EXIST
"B" INDICATES THAT THEY DO NOT EXIST AT THAT LOCATION
TOP NUMBER IN EACH LOCATION IS THE ORDER ASSIGNED
BY THE DECISION RULE
MIDDLE AND BOTTOM NUMBERS INDICATE THE DESIGNERS'
ACCEPTABLE RANGE OF ASSIGNMENT ORDER
COMPARISON OF FIRST URBANIZATION PATTERN
GENERATED BY DECISION RULES
WITH URBANIZATION PATTERN
GENERATED BY DESIGNERS
ROW 1, COL 11 TO ROW 10, COL 18
APPENDIX
PAGE 164
1
1 2 3 4 5 6 78 9 0 12 3 4 56 7 8
.
@
A
+.+.*. A . 1
*.+.+.+.+. A+. 1 . . 2 2
*.*.*. A 2 2 . .+.+.
*.
A A . .
+.+. .
*.*.*. A 2 2 . .0. . +.+. . .
*.*. A A . 2 2 . .
. .+ .+ .
*. A A A AA ....
.++ ..
1A 222 1.
...
+..
I
.
1
11
. 1
. +.+.+..
.+.
S
.
1
.
.+.
0
4
5
6
7
8
9
10
11
12
13
14
15
2 A 2 o. . .+.+.+.+.
1
1
1
2
3
123456789012345678
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS
IDESIGN(2)
'A'
= 1, 2
- -
'1' = INDUSTRY(7);
'2'
'.' = NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+' = RIVER(8); '-'
CHANNEL(12);
'*'
= CHANNEL & RIVER; ' ' = NO REF ATTS
COMPARISON OF FIRST INDUSTRY PATTERN
GENERATED BY DECISION RULES (INDUSTRY)
WITH INDUSTRY PATTERN
GENERATED BY DESIGNERS (IDESIGN)
=
PAGE 165
APPENDIX
1
2
4
5
6
3
7
8
9
10
B
B
B
B
B
B
7
B
B
B
B
B
B
B
B
B
8
15
8
B
B
B
B
B
B
B
14
25
14
B
B
B
B
5
7
B
B
B
6
B
B
B
B
B
B
8
B
15
B
B
B
B
B
14
B
B
B
25
B
B
B
8
B
B
B
B
4
B
B
B
B
B
B
8
15
B
B
B
2
B
B
B
B
B
B
14
25
7
B
B
B
B
9
B
B
B
1
3
B
B
B
B
B
B
B
15
1
8
B
2
B
B
B
1
B
B
25
B
14
B
7
B
B
B
1
10 B
8
B
15
14
1
B
B
B
1
B
B
15
1
2
15
B
B
B
1
B
B
25
1
7
25
B
B
B
B
B
B
B
B
B
16
11
10
9
B
8
B
8
B
15
15
B
B
B
B
B
B
25
14
25
B
14
B
B
B
12 B
B
17
B
B
B
B
B
B
B
15
B
15
B
B
B
B
15
B
B
B
25
B
B
B
25
25
B
B
B
B
B
18
19
B
B
B
13 24
21
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
22
B
B
B
B
14
25
B
20
B
B
B
B
B
B
B
B
B
B
B
B
B ISTB
B
B
B
B
B
B
B
B
8
B
23
B
B
B
15
B
B
B
B
B
IDE
B
B
B
B
B
B
B
B
B
B
IS
B
B
B
B
B
B
B
B
B
8
7
4
5
6
1
2
3
9
10
IDES
IDESIGN(ORDERL 0)), AND
INDUSTRY(ORDE R),
IGN(ORDERHI)
HAVE BEEN GIV EN A VALUE WHERE T HEY EX
"B" INDICATES THA T THEY DO NOT EXIST AT THAT LOCATION
TOP NUMBER IN EACI H LOCATION IS THE ORDER ASSIGNED
BY THE DECISI ON RULE
MIDDLE AND BO TTOM NUMBERS INDICATE T HE DESIGNERS'
ACCEPTABLE RA NGE OF ASSIGNMENT ORDER
6
COMPARISON OF FIRST INDUSTRY PATTERN
GENERATED BY DECISION RULES
WITH INDUSTRY PATTERN
GENERATED BY DESIGNERS
ROW 6, COL 1 TO ROW 15, COL 10
APPENDIX
PAGE 166
11
14
15
16
17
18
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
2
B
B
B
12
13
B
B
B
B
2
B
B
B
B
7
B
B
3
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
4
B
11
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
5
1
B
B
B
B
B
1
B
2
2
B
1
7
7
B
B
B
B
6
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
7
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
8
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
9
B
9
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
10
10 B
B
B
B
B
B
B
B
B
c
B
B
B
14
11
15
16
17
18
I NDU STRY( ORDER),
I DESI GN(ORDERLO),AND IDESIG N(ORDERHI)
HAVE BEEN GIVEN A VALUE WHERE THEY EXIST
"B" I ND I CATES THAT THEY
DO NOT EXIST AT THAT LOCATION
THEY
TOP NUMB ER IN EACH LOCATION IS THE ORDER ASSIGNED
BY THE DEC ISION RULE
MIDDLE AND BOTTOM NUMBERS INDICATE THE DESIGNERS'
ACCEPTABLE RANGE OF ASSIGNMENT ORDER
1
COMPARISON OF FIRST INDUSTRY PATTERN
GENERATED BY DECISION RULES
WITH INDUSTRY PATTERN
GENERATED BY DESIGNERS
ROW 1, COL 11 TO ROW 10, COL 18
PAGE 167
APPENDIX
1
123456789012345678
1
2
3
_*_
5
6
7
8
9
10
11
12
13
14
,÷o0
.
.**
.
15
0
.
.
.
*
0
0
•
.
.
0
1
.
.
.
.
1
.
2
A
.+.+.
.
.
.
.
.
2
2
.
.
I+.+.
.
.
.
0
0
+
.
0+0++
0
0
0+0
0
+
0 ++
0
0
0
÷
.
0
0
0 ,+0
.
.
.
*.*.*0
.
*..
+,--,
4
.+0+0
+
0+0+
123456789012345678
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS -
-
'1' = CENTER(5);
'2' =
CDESIGN(4)
'A'
= 1, 2
'.'
= NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+' = RIVER(8); '-'
= CHANNEL & RIVER; ' ' = NO REF ATTS
'*'
CHANNEL(12);
COMPARISON OF FIRST CENTER PATTERN
GENERATED BY DECISION RULES (CENTER)
WITH CENTER PATTERN
GENERATED BY DESIGNERS (CDESIGN)
PAGE 168
design2 p
IF DESIGN SYSTEM CRASHES GIVE THE DEBUG COMMAND:
DUMP NAME 12000 3D000
R 0.02/0.02
rd dataguay p
R 0.78/0.80
rd execguay p
printf execguay p
SET_CARRIAGE WIDTH 60
READY OFF
MESSAGE OFF
RD LBESIDE P
RD LGRAVITY P
RETURN
comment execguay p sets the carriage width to 60 characters,
comment turns the ready message off, turns other messages
comment off, and reads in two files which are needed in the
comment Guayana assignment rules.
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 3,17
DECISION PATH WAS U1 U2 U3 U4 U6
comment this last command assigned urbanization to the land
comment according to the decision rules developed in
comment Chapter II.
comment to complete the algorithm of Chapter II it is
comment necessary to go through the whole series of
comment assignments: 3u, 3i, Ic, 3u and 3i - 4 times.
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 8,11
DECISION PATH WAS Ul U2 U3 U4 U7
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 9,12
DECISION PATH WAS Ul U2 U3 U4 U6
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 9,4
DECISION PATH WAS 11 12 13 16 110
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 8,5
DECISION PATH WAS 11 12 13 16 110
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 7,6
12 13 16 110
DECISION PATH WAS 11
PAGE 169
rd assignc p
"CENTER" HAS BEEN PLACED AT 7,9
DECISION PATH WAS Cl C2 C6
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 4,17
DECISION PATH WAS U1 U2 U3 U4 U7
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 5,16
DECISION PATH WAS U1 U2 U3 U4 U7
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 6,17
DECISION PATH WAS Ul U2 U3 U4 U7
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 6,7
DECISION PATH WAS 11 12 13 16 110
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 7,7
DECISION PATH WAS 11 12 13 16 110
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 10,2
DECISION PATH WAS 11 12 13 16 110
comment now starts the second time through the series.
rd assignu p
HAS BEEN PLACED AT 7,17
"URBANIZATION"
DECISION PATH WAS U1 U2 U3 U4 U7
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 8,17
DECISION PATH WAS Ul U2 U3 U4 U7
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 5,17
DECISION PATH WAS U1 U2 U3 U4 U7
rd assigni p
HAS BEEN PLACED AT 11,2
"INDUSTRY"
12 13 16 110
DECISION PATH WAS 11
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 11,1
12 13 16 110
DECISION PATH WAS 11
PAGE 170
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 4,15
DECISION PATH WAS 11 12 13 16 Ill
rd assignc p
HAS BEEN PLACED AT 8,9
"CENTER"
DECISION PATH WAS Cl C2 C3 CS
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 9,17
DECISION PATH WAS Ul U2 U3 U4 U7
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 10,16
DECISION PATH WAS Ul U2 U3 U4 U7
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 10,17
DECISION PATH WAS Ul U2 U3 U4 U7
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 2,17
DECISION PATH WAS 11 12 13 16 Ill
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 2,18
DECISION PATH WAS 11 12 13 16 110
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 7,8
12 14 17 112
DECISION PATH WAS 11
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 7,18
DECISION PATH WAS Ul U2 U3 U4 U7
comment that was the first of the tiird series.
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 6,18
DECISION PATH WAS Ul U2 U3 U4 U7
"URBANIZATION" HAS BEEN PLACED AT 8,18
DECISION PATH WAS Ul U2 U3 U4 U7
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 8,8
12 14 17 112
DECISION PATH WAS 11
PAGE 171
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 9,9
DECISION PATH WAS 11 12 14 17 112
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 9,8
DECISION PATH WAS 11 12 14 17 112
rd assignc p
HAS BEEN PLACED AT 9,9
"CENTER"
DECISION PATH WAS Cl C2 C3 C5
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 5,18
DECISION PATH WAS Ul U2 U3 U4 U7
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 9,18
DECISION PATH WAS Ul U2 U3 U4 U7
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 10,18
DECISION PATH WAS Ul U2 U3 U4 U7
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 10,8
12 14 17 112
DECISION PATH WAS 11
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 8,7
12 14 17 112
DECISION PATH WAS 11
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 9,7
12 14 17 112
DECISION PATH WAS 11
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 5,13
DECISION PATH WAS U1 U2 U3 U5 U8
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 6,14
DECISION PATH WAS Ul U2 U3 U5 U8
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 9,11
DECISION PATH WAS Ul U2 U3 U5 U8
PAGE 172
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 10,7
12 14 17 112
DECISION PATH WAS 11
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 11,7
12 14 17 112
DECISION PATH WAS 11
rd assigni p
HAS BEEN PLACED AT 10,6
"INDUSTRY"
DECISION PATH WAS 11 12 14 17 112
rd assignc p
"CENTER" HAS BEEN PLACED AT 7,8
DECISION PATH WAS Cl C2 C3 C5
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 6,9
DECISION PATH WAS Ul U2 U3 U5 U8
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 10,11
DECISION PATH WAS Ul U2 U3 U5 US
rd assignu p
"URBANIZATION" HAS BEEN PLACED AT 6,8
DECISION PATH WAS Ul U2 U3 U5 U8
rd assignu#i p
"INDUSTRY" HAS BEEN PLACED AT 11,6
DECISION PATH WAS 11 12 14 17 112
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 10,5
12 14 17 112
DECISION PATH WAS 11
rd assigni p
"INDUSTRY" HAS BEEN PLACED AT 12,6
12 14 17 112
DECISION PATH WAS 11
comment that completes the series.
store d4guay p
PAGE
comment to exhibit some of the bookkeeping, and to find
comment out where to look for the information which has
comment been kept, "list" and "show" commands follow.
list center
CHARVAR
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
CHARCON
1 OF CENTER IS CALLED ORDER
1 OF CENTER IS CALLED C1
2 OF CENTER IS CALLED C2
3 OF CENTER IS CALLED C3
4 OF CENTER IS CALLED C4
5 OF CENTER IS CALLED C5
6 OF CENTER IS CALLED C6
7 OF CENTER IS CALLED C7
8 OF CENTER IS CALLED D8
9 OF CENTER IS CALLED D9
10 OF CENTER IS CALLED D10
11 OF CENTER IS CALLED D11
12 OF CENTER IS CALLED C1234
13 OF CENTER IS CALLED C1235
14 OF CENTER IS CALLED C126
15 OF CENTER IS CALLED C17
for i= 12. step 1.
until 15. show center(i)
CENTER(C1234) = 0.
CENTER(C1235) = 3.
CENTER(C126) = 1.
CENTER(C17) = 0.
for i= 23. step 1.
until 30. show industry(i)
INDUSTRY(112358) = 0.
INDUSTRY(112359) = 0.
INDUSTRY(1123610) = 9.
INDUSTRY(1123611) = 2.
INDUSTRY(I124712) = 13.
INDUSTRY(1124713) = 0.
INDUSTRY(112414) = 0.
INDUSTRY(1114) = 0.
for i=17.
step 1. until 22. show urbanization (i)
URBANION(U12346) = 2.
URBANION(U12347) = 16.
URBANION(U12358) = 6.
URBANION(U12359) = 0.
URBANION(U1210) = 0.
URBANION(U110) = 0.
173
PAGE 174
APPENDIX
1
0 1 23 4 5678
123456789
1
S*o
+.*.*.
1 1
+0.**.
+.*.+. 2 1 2 .
*.+.+.+.+. 2+. 1 1 . 1
2
3
4
5
.
2 2
1.
1 .
11
1
2
3
4
5
6
*.*.*.
2 1 1 1 .+.+. 1 . . 1 1
6
7
*. . 2 2 C 3 . I 1 .+.+. . 1 1
7
8
9
10
2 . 2 2 3 . 1 .+.+. 1 . 1 1
2 2 . . 2 2 C . 1 1+.+. . . 1 1
*.*.*.
*.*.
*. 2 2 2 2 2 2 2 .. 1+.+.
11
12
.+.
2 2 . . . 2 2 ...
2 .+.+.+.+.
.....
13
. . . . .
+.+.+.
14
+.
0
. . . . .
15
.
.
.
1
0.
.
1
8
9
10
11
12
13
+.
123456789012345678
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS - INDUSTRY(7); '3' = CENTER(5)
'C'
= 2,
3
'1'
= URBANION(6);
=
'2'
= NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+'
'*' = CHANNEL & RIVER; '
CHANNEL(12);
'
= RIVER(8); '-'
= NO REF ATTS
SECOND PATTERN FOR CIUDAD GUAYANA
GENERATED BY DECISION RULES
SHOWING URBANIZATION, INDUSTRY AND CENTER
=
PAGE 175
APPEND I X
1
123456789012345678
1
2
. *.
+.*.*. A
+.*.+. . A 2
.+. A A 2
*.+.+.+.+.
. 1 1 A .+.+.
A .
•.*.*.
A A .+.++.
. 2 2
*. .
3
4
5
6
7
8
9
10
11
1 12
. . **.
. .. 2 2 . 1 .+.+.
.
A A
2 1 .
1 1 1
1 1
. 1 1
A . 1 1
. A A+.+. . . 1 1
• .* . . 0 . . . . .
. 222 . A+.+. .. 1 111
* ..
. . .
22 2
. 2 2+. .
0
. . . . 2 2 2+.+.+.+.
. . . . . .+.
.* . . . . +.
14
15
123456789012345678
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS UDESIGN(3)
'A'
= 1, 2
-
'1' = URBANION(6);
'2'
= NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+'
= CHANNEL & RIVER; '
'*'
CHANNEL(12);
= RIVER(8); '-'
' = NO REF ATTS
COMPARISON OF SECOND URBANIZATION PATTERN
GENERATED BY DECISION RULES (URBANIZATION)
WITH URBANIZATION PATTERN
GENERATED BY DESIGNERS (UDESIGN)
=
APPEND I X
PAGE 176
1
123
4567890123
45678
.*.
.Al
+.1**
*
0o
. A
+.*0+
.
*A
1
0
1
.
.2
..
*.+.+.+.+. A+
. . . 2
.
.
.
*.*.*. A 2 2 . .+.+.
.
.
.
*.
A A 1 . . . . +.+.
.+.+
. .. .. ..
*.*.*. A 2 A 1 .0 .
*.*. A A . 2 A 1 1 . . .+.+..
.+.+.
A 1 1 .
*. A A A
1A 2 2 2 1 1 . . . .. .
2 2 2 2 . 1 .+.+.+.*+.
. .
13
.. +.+.+.
14
. . . . . .+.
15
. . . . . +.
2
3
4
5
6
7
8
9
10
11
12
.
123456789012345678
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS - IDESIGN(2)
'A'
= 1, 2
'.'
'1'
= INDUSTRY(7); '2'
= NON EMPTY LAND
= RIVER(8); '-' =
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+'
CHANNEL(12);
'*'
= CHANNEL & RIVER; ' ' = NO REF ATTS
COMPARISON OF SECOND INDUSTRY PATTERN
GENERATED BY DECISION RULES (INDUSTRY)
WITH INDUSTRY PATTERN
GENERATED BY DESIGNERS (IDESIGN)
=
PAGE 177
APPENDIX
1
123456789012345678
1
.*.
2
3
+.*.*. . .
. . . .
+.*.*
4
+.*.+. .
.+ .
*.+.+.+.+.
5
6
*.*.*
. .
* .
7
. +.+
. 1
..
. .
.
. .
. .
.
. . .....
+. .
*.*.*.* . . .
1 2 2 .+.+ . . . .
.a.*.
. . . . 2 A . . .+.+. . . . .
8
9
10
.
*
..
.
0
.
0
.0
0 0.
.0 . 0
. .
. +.+.+.+.
13
. . . . .
+.+.+.
14
.
15
. . . .
.
.
.
.
+.+
.+.
12
.
.
0
.
.
.
.
0
11
.
1
2
3
4
5
6
7
8
9
10
11
.
12
.
13
14
0+.
.+.
15
123456789012345678
1
LEGEND:
PRIMARY ATTRIBUTE ASSIGNMENTS - CDESIGN(4)
'A'
= 1,
2
'.'
'1' = CENTER(5);
'2' =
= NON EMPTY LAND
REFERENCE ATTRIBUTE ASSIGNMENTS - - '+' = RIVER(8); '-'
CHANNEL(12);
'*' = CHANNEL & RIVER; ' ' = NO REF ATTS
COMPARISON OF SECOND CENTER PATTERN
GENERATED BY DECISION RULES (CENTER)
WITH CENTER PATTERN
GENERATED BY DESIGNERS (ICENTER)
=
APPEND IX
PAGE 178
DECISION RULE FOR URBANIZATION
SUBSTITUTION ON
FOR I= 1. STEP 1. UNTIL 10. URBANIZATION(1) = 0.
COMMENT THIS SETS THE CHARCONS WHICH KEEP TRACK OF
COMMENT THE DECISION PATH TO 0. INITIALLY
TLISTO *=*
CENTER .OR.
URBANIZATION .OR.
INDUSTRY
EX GRAVITY (TLISTO X Y)
COMMENT LOC X,Y IS THE CENTER OF GRAVITY OF ALL ACTIVITIES
URBANIZATION(U1) = 1.
COMMENT URBANIZATION(U1) IS ONE OF THE CHARCONS KEEPING
COMMENT TRACK OF THE DECISION PATH
TLIST1 *=*
SLOPE05
.OR.
SLOPE615
.ANDN.
TLISTO
COMMENT TLIST1 IS THE DEVELOPMENT ZONE FOR URBANIZATION
COMMENT FORMED AT THE FIRST NODE ON THE DECISION TREE
COMMENT DECISION TREE
0. GOTO UR3
IF TLIST1 .EQL.
COMMENT IF IT DOESN'T EXIST THEN THE PROGRAM BRANCHES TO UR3$
EX BESIDE (URBANIZATION)
COMMENT THIS PRODUCES A LIST NAMED "BESIDE" WHICH CONTAINS
COMMENT ALL THE LOCATIONS NEXT TO "URBANIZATION"
TLIST2 *=* TLIST1 .AND.
URBANIZATION(U2) = 1.
IF TLIST2 .EQL.
0.
GOTO UR3
TLIST1 *=* TLIST2
TLIST2 *=* TLIST1 .AND.
URBANIZATION(U3) = 1.
IF TLIST2 .EQL. 0.
EX BESIDE(RIVER)
BESIDE
SLOPE05
GOTO UR5
TLIST1 *=* TLIST2
TLIST2 *=* TLIST1 .AND.
URBANIZATION(U4) = 1.
BESIDE
URBANI ZATION(U7)=1.
IF TLIST2 .EQL. 0. GOTO UR8
URBANIZATION(U7) = 0.
URBANIZATION(U6) = 1.
TLIST1 *=* TLIST2
GOTO UR8
UR5$ CONTINUE
EX BESIDE(RIVER)
TLIST2 *=* TLIST1 .AND. BESIDE
URBANIZATION(US) = 1.
URBANIZATION(U9) = 1.
IF TLIST2 .EQL.
0.
GOTO UR8
URBANIZATION(U9) = 0.
TLIST1 *=* TLIST2
URBANIZATION(U8) = 1.
UR3$ CONTINUE
APPENDIX
PAGE 179
NEAREST TLIST1 X,Y BESTI, BESTJ MINDIST
PUT URBANIZATION BESTI, BESTJ
EXPAND +**"URBANIZATION" HAS BEEN PLACED AT $3BESTI,$3BESTJ/+
ORDERU = ORDERU + 1.
URBANIZATION(ORDER BESTI,BESTJ) = ORDERU
UR12$ CONTINUE
EXPAND +DECISION PATH WAS U1 +
T = 1.
THROUGH UR10 FOR 1=2. STEP 1. UNTIL 10.
IF URBANIZATION(I) .EQL. 0. GOTO UR10
EXPAND + U$31 +
IF I .GEQ. 10. A = 100.
IF I .LES. 10. A = 10.
T = T*10. + 1
UR10$ CONTINUE
EXPAND +//+
THROUGH UR11 FOR I=11. STEP 1. UNTIL 17.
J = 1 + 6.
IF T-URBANIZATION(I) .EQL. 0. URBANIZATION(J) = URBANIZATION(J) + 1.
UR11$ SUBSTITUTION OFF
READ CONSOLE
UR3$ URBANIZATION(U10) = 1.
EXPAND +NO LAND SUITABLE FOR "URBANIZATION" +
EXPAND +WAS FOUND/+
GOTO UR12
DECISION RULE FOR URBANIZATION
APPEND I X
PAGE 180
DECISION RULE FOR INDUSTRY
SUBSTITUTION ON
FOR I= 1. STEP 1. UNTIL 14. INDUSTRY( I)=0.
TLISTO*=*CENTER.OR. INDUSTRY.OR.URBANIZATION
EX GRAVITY (TLISTO X Y)
I NDUSTRY(1) =1.
TLIST1*=*SLOPE05 .OR. SLOPE615 .ANDN. TLISTO
IF TLIST1 .EQL.
0. GOTO IQ3
EX BESIDE (CHANNEL)
TLIST2 *=*
TLIST1
.AND. BESIDE
INDUSTRY(2)=1.
IF TLIST2.EQL. 0. GOTO IQ4
TLIST1 *=*TLIST2
TLIST2 *=* TLIST1
INDUSTRY(3) = 1.
.AND. SLOPE05
0. GOTO IQ6
IF TLIST2 .EQL.
EX BESIDE(INDUSTRY)
TLIST1 *=*
TLIST2
TLIST2 *=*TLIST1 .AND. BESIDE
INDUSTRY(5)=1.
INDUSTRY(9)=1.
0. GOTO 1Q8
IF TLIST2 .EQL.
INDUSTRY(9)=0.
TLIST1*=*TLIST2
INDUSTRY(8)=1.
GOTO IQ8
IQ6$ INDUSTRY(6)=1.
EX BESIDE(INDUSTRY)
TLIST2*=*TLIST1 .AND. BESIDE
I NDUSTRY(11)=1.
IF TLIST2 .EQL.
0.
GOTO 1Q8
INDUSTRY(11) = 0.
TLIST1 *=* TLIST2
INDUSTRY(10) = 1.
GOTO IQ8
IQ4$ INDUSTRY(4)=1.
EX BESIDE( INDUSTRY)
TLIST2*=*T LIST1 .AND. BESIDE
IF TLIST2 .EQL. 0. GOTO IQ3
TLIST1 *=* TLIST2
TLIST2 *=* TLIST1 .AND. SLOPE05
I NDUS.TRY(7 )=1.
INDUSTRY(1 3)=1.
IF TLIST2 .EQL. 0. GOTO IQ8
INDUSTRY(1 3)=0.
INDUSTRY(12)=1.
TLIST1 *=* TLIST2
APPENDIX
PAGE 181
IQs$ CONTINUE
NEAREST TLIST1 TO X,Y BESTI,BESTJ MINDIST
PUT INDUSTRY BESTI, BESTJ
EXPAND +**"INDUSTRY" HAS BEEN PLACED AT $
ORDERI = ORDERI
S3BESTI,$3BESTJ/+
+ 1.
INDUSTRY(ORDER BESTI,BESTJ) = ORDERI
IQ12$ CONTINUE
EXPAND +DECISION PATH WAS 11 +
T=1.
THROUGH 1Q10 FOR 1=2. 1. 14.
IF INDUSTRY(I) .EQL. 0. GOTO 1Q10
EXPAND + 1$31 +
IF I .GEQ. 10. A = 100.
IF I.LES. 10. A = 10.
T = T*A + I
IQ10$ CONTINUE
EXPAND +//+
THROUGH IQ11 FOR 1=15. 1 • 22.
J = I + 38
INDUSTRY(J) = INDUSTRY(J)
IQ11$ SUBSTITUTION OFF
READ_CONSOLE
IQ3$ INDUSTRY(14)=1.
EXPAND +NO LAND SUITABLE FOR "INDUSTRY"
EXPAND +WAS FOUND/+
GOTO 1Q12
1
F
I
T-
I
)
.
EQL.
NDUSTRY(
DECISION RULE FOR INDUSTRY
+
1.
APPENDIX
PAGE 182
SUBSTITUTION ON
FOR 1= 1. STEP 1. UNTIL 7. CENTER(1)
TLISTO *=*
CENTER .OR.
= 0.
URBANIZATION .OR. INDUSTRY
EX GRAVITY (TLISTO X Y)
TLIST1 *=* SLOPE05 .AND..NOT.
RIVER .AND..NOT. CENTER
CENTER(C1) = 1.
IF TLIST1 .EQL.
0. GOTO CE3
EX BESIDE (CENTER)
UNACCEPTABLE .AND..NOT.
TLIST2 *=* TLIST1 .AND. BESIDE
CENTER(C2) = 1.
IF TLIST2 .EQL. 0. GOTO CE7
EX BESIDE (URBANIZATION)
TLIST1 *=* TLIST2
TLIST2 *=* TLIST1
CENTER(C3) = 1.
IF TLIST2 .EQL.
0.
.AND.
BESIDE
GOTO CE6
TLIST1 *=* TLIST2
CENTER(C4) = 1.
CE7$ CENTER(C6) = 1.
GOTO CE4
CE6$
CENTER(CS) = 1.
CE4$ CONTINUE
NEAREST TLIST1 X,Y BESTI,BESTJ MINDIST
PUT CENTER BESTI, BESTJ
EXPAND +**"CENTER" HAS BEEN PLACED AT $3BESTI,$3BESTJ/+
ORDERC = ORDERC + 1.
CENTER(ORDER BESTI, BESTJ) = ORDERC
EXPAND +DECISION PATH WAS Cl +
T=1.
THROUGH CE5 FOR 1=2. STEP 1. UNTIL 6.
IF CENTER(I) .EQL. 0. GOTO CE5
EXPAND + C$31
T=T*10. + I
+
CE5$ CONTINUE
EXPAND +//+
THROUGH CE8 FOR 1= 8. STEP 1. UNTIL 11.
J = I + 4.
IF T- CENTER(1) .EQL. 0. CENTER(J) = CENTER(J) + 1.
CE8$ SUBSTITUTION OFF
READ_CONSOLE
CE3$ CONTINUE
EXPAND +NO LAND SUITABLE FOR "CENTER" WAS FOUND/+
EXPAND +DECISION PATH WAS C1 C7/+
CENTER(C17) = CENTER(C17) + 1.
SUBSTITUTION OFF
READ_CONSOLE
DECISION RULE FOR CENTER
APPENDIXX
PAGE 183
CIUDAD GUAYANA DATA
SETUP 1 15 1 13 60 60
DEFINE IDES IGN WITH 2 CHARVARS: ORDERLO ORDER H I,
DEFINE UDES IGN WITH 2 CHARVARS: ORDERLO ORDER HI*
DEFINE CDES IGN WITH 2 CHARVARS: ORDERLO ORDER HI*
PUT IDESIGN 2,17(2. 7.) 4,13(1. 1.) 5,11(1. 1
5,16(2. 7.) 5,17(2. 7.) 6,7(8. 14.) 6,8(15. 25.) 6,9(8 . 14.)'
7,6(8. 14.) 7, 7(15. 25.) 8,5(8. 14.) 8,6(15. 25.) 8,7( 2. 7.)'
9,3(1. 1.) 9,4 (8. 14.) 9,6(2. 7.) 9,7(15. 25. ) 10,2(15 . 25,)'
10,3(1. 1.) 10 ,4(1.
1.) 10,5(2. 7.) 10,6(15. 25.) '
11,2(15. 25 .) 11,3(8. 14.) 11,4(8. 14.) 11,5( 15. 25.)
12,1(15. 25
12,2(15. 25.) 12,3(15. 25.) 12, 4(15. 25.
PUT UDESIGN 3, 15(1. 1.) 3,16(1. 1.) 3,17(6. 1 1.) '
4,14(1. 1.) 4, 15(2. 5.) 4,16(2. 5.) '
5,13(12. 14 .) 5,14(1. 1.) 5,15(6. 11.) '
6,10(1. 1.) 6, 14(6. 11.) 7,8(15. 25.) 7,9(12. 14. ) 2
7,11(1. 1.) 7,12(1. 1.) 8,8(15. 25.) 8,9(12. 14.) 8,15(1. 1.)'
9,11(2. 5.) 9, 12(2. 5.) 10,7(15. 25.) 10,8(6. 11.
10,9(6. 11. )10,11(5, 10) 11,6(15. 25.) 11,7( 15. 25.)'
11,8(15. 25 .) 11,10(15. 25.) 11,11(15. 25.)
12,5(15. 25 .) 12,6(15. 25.) 12,7(15. 25.)
PUT CDESIGN 8, 11(2. 2.) 8,10(4. 4.)
7,8(5. 5.) 7,9(3. 3.)
DEFINE CENTER WITH 1 CHARVAR: ORDER * AND WITH 15'
CHARCONS: Cl = 0. C2 = 0. C3=0. C4=0. C5=0. C6=0. C7=0.
D8=1234. D9=1235. D10=126. D11=17.
C1234=0. C1235 = 0. C126=0. C17=0.
DEFINE URBANIZATION WITH 1 CHARVAR:ORDER * AND
WITH 22 CHARCONS: Ul=0. U2=0. U3=0. U4=0. U5=0. U6=0. '
U7=0. U8=0. U9=0. U10=0. U11=12346. U12=12347.'
U13=12358. U14=12359. U15=1210. U16=110. '
U12346=0. U12347=0. U12358=0. U12359=0. U1210=0.
U110=0.
DEFINE INDUSTRY WITH 1 CHARVAR: ORDER * AND WITH '
30 CHARCONS:11=0. 12=0. 13=0. 14=0. 15=0. 16=0. 17=0. 18=0.
19=0. 110=0. 111=0. 112=0. 113=0. 114=0. '
D15=12358. D16=12359. D17=123610. D18=123611. D19=124712.
D20=124713. D21=12414. D22=114. 112358=0. 112359=0.
1123610=0. 1123611=0. 1124712=0. 1124713=0.
112414=0. 1114=0.
RIVER *=* NULL
SLOPE05 *=* NULL
SLOPE615 *=* NULL
UNACCEPTABLE *=* NULL
CHANNEL *=* NULL
ORDERC = 1.
ORDERU=I.
ORDERI=1.
APPENDIX
PAGE 184
PUT INDUSTRY 4,13 5,11 9,3 10,3 10,4
PUT URBANIZATION 3,15 3,16 4,14 5,14 6,10 7,11 7,12 8,15
PUT RIVER 1,17 2,15 2,16 2,14 3,12 3,13 3,14 4,10 4,11 4,12
5,12 6,4 6,5 6,6 6,12 6,13 7,4 7,14 7,15 8,2 8,3 8,4 8,13 '
8,14 9,1 9,2 9,13 9,14 10,1 10,12 10,13 11,12 12, 8 12,9 '
12,10 12,11 13,7 13,8 13,9 14,7 15,7 5,6 5,7 5,8 5,9 5,10
PUT SLOPE05 3,17 4,14 4,17 5,16 5,17 5,18 6,1 7 6, 18 7,8 7,9
7,17 7,18 8,7 8,8 8,9 8,11 8,17 8,18 9,7 9,8 9,9 9,12 9,17 '
9,18 10,3 10,5 10,6 10,7 10,8 10,16 10,17 10, 18 '
11,5 11,6 11,7 12,1 12,2 12,3 12,4 12,5 12,6 13,1 13,2 13,3 '
13,4 13,5 14,1 14,2 14,3 14,4 14,5 15,1 15,2 15,3 15,4
PUT SLOPE615 2,17 2,18 3,16 3,18 4,15 4,16 4, 18 5,11 5,13 5,14
5,15 6,7 6,8 6,9 6,10 6,14 6,15 7,6 7,7 7,11 7,12 8,5 8,6'
8,15 8,16 9,3 9,4 9,6 9,11 9,15 9,16 10,2 10, 4 10,9 10,11 '
11,1 11,2 11,3 11,4 11,8 11,10 11,11 12,7 13, 6 14,6 15,5 15,6
PUT UNACCEPTABLE 1,18 4,13 6,11 6,16 7,5 7,10 7,13 7,16 '
8,10 8,12 9,5 9,10 10,10 10,14 10,15 11,9 11, 13 12,12 13,10
PUT CHANNEL 1,17 2,15 2,16 3,13 3,14 4,11 6,4 6,6 6,5 7,4 8,2'
8,3 8,4 9,1 9,2 10,1 5,6
GRIDSIZE = 1.
N = 1.
P = 15.
Q = 1.
R = 18.
BESTI = 0.
BESTJ = 0.
MINDIST = 0.
A = 0.
READCONSOLE
DATA FILE FOR CIUDAD GUAYANA, 1961
INCLUDING THE DESIGNERS' DESIGN
'
PAGE 185
BIOGRAPHICAL NOTE
William Lyman Porter was born in Poughkeepsie, N.Y. on February 19,
1934. Prior to entering the doctoral program he attended Yale University
where he majored in Architecture. He received a B.A. degree in 1955 and
a B. Arch. in 1957.
In 1958 he was drafted into the U.S. Army where he served as a
draftsman-designer. Upon his release from active duty he was commissioned
2nd Lieutenant in the U.S. Army reserve. He was honorably discharged in
1964.
From 1960 to 1962 he worked for Louis I. Kahn, Architect in Philadelphia.
It was there that his interest in the analysis of design and of the creative
process was nurtured. He became registered to practice architecture in
Pennsylvania in 1961.
From 1962 to 1964 he worked as an urban designer on the fast-growing
industrial city of Ciudad Guayana in the southeastern part of Venezuela.
In 1964 he entered the doctoral program in M.I.T.'s Department of
City and Regional Planning. He received a Mellon Fellowship from M.I.T.
in 1964 and a Stouffer Fellowship from the Joint Center for Urban Studies
of Haraard and M.I.T. in 1966.
He has been an Instructor in Civic Design at the University of Pennsylvania in 1961, and in City Planning at M.I.T. in 1966. He is presently
Assistant Professor of Urban Design in the Departments of Architecture and
Urban Studies and Planning.
Mr. Porter has written for publication "Changing Perspectives on
Residential Area Design," in Regional Planning for Developemnnt by L. Rodwin
and Associates (M.I.T. Press, 1969). and, with K.J. Lloyd and Aaron Fleisher,
"DISCOURSE: The Development of a Language and System for Computer Assisted
City Design" in the Proceedings of the Design Methods Group Conference,
Spring 1968, (forthcoming).
Mr. Porter is a member of the American Institute of Architects, and
the Boston Society of Architects. He is married and has three children.
PAGE 186
FOOTNOTES
CHAPTER I: FOOTNOTES
(1.01) The team which has developed DISCOURSE has included: Katherine
J. Lloyd and Wren McMains as overall systems programmers; Corrie Menger,
Stanley Hoderowski, and Chris Reedy programming various aspects of the
system; Michael O'Hare, David Sandahl as users of the system contributing
to its development by placing specific demands upon it; and Aaron Fleisher
for overall direction and coordination.
The project has enjoyed the support of IBM, The Urban Systems
Laboratory and the Department of Urban Studies and Planning.
(1.02)
"It is a profoundly erroneous truism, repeated
by all copy-books and by eminent people when they are
making speeches, that we should cultivate the habit
of thinking of what we are doing. The precise opposite
is the case. Civilization advances by extending the
number of important operations which we can perform
without thinking about them. Operations of thought
are like cavalry charges in a battle--they are strictly
limited in number, they require fresh horses, and must
only be made at decisive moments."
Alfred North Whitehead, An Introduction to Mathematics
(New York: Oxford University Press, 1958), pp. 41-42.
(1.03) There is a wide range in the literature of what is meant by
"city, "civic" or "urban" design and "physical" city planning. Locating
the point of view represented here in that range is not crucial for the
developemnt of the DISCOURSE language. The discussion of city design is
intended to help the reader to imagine some of the implied or understated
examples throughout the rest of the thesis.
(1.04) Kevin A. Lynch, "Quality in City Design," Who Designs America,
Laurence B. Holland, editor (Princeton Studies in American Civilization,
No. 6, New York: Anchor Books, 1966).
(1.05)
Ibid., p. 142.
(1.06) For a discussion fo the not-so-easily measured see, for example,
Susanne K. Langer, "The Social Influence of Design," Who Designs America,
op. cit. and
Kevin A. Lynch, "City Design and City Appearance," Principles
International City
and Practices of Urban Planning (Washington, D.C.:
Managers' Association, 1968), pp. 249-271.
(1.07) A great deal has been created at the time of problem formulation or recognition. A restricted number of variables has been identified,
major points of conflict noted, and a hint of the nature of the solution
because of the kinds of conflict to which the designer's attention has been
drawn. See also Mackworth in his discussion of "problem finding" versus
"problem solving." He regards the former as the activity demanding much
greater originality.
N.H. Mackworth, "Originality," American Psychologist, Vol. 20,
No. 1, 1964, pages 51-66.
FOOTNOTES
PAGE 187
CHAPTER I: FOOTNOTES
(1.08) Let me stress here and elsewhere that the aspects of the
design process which I have addressed are those concerned with the designer's
communication with himself and his thinking through of the design issues
independently during the time he is generating design ideas; not all the
activities of the designer during that process.
(1.09) Lynch speaks of several of these in his "Quality in City
Design," op. cit., pp. 141-152.
(1.10) Although the sequence of technical studies associated with
the physical planning and design of cities may be continuous, that is,
goals constantly being formulated and revised, new information gathering
activities being initiated and so forth, with respect to any particular
design product there will probably have been the kinds of changes over
the duration of generating it which are spoken of in the text.
For a description of the sequence of technical studies see:
F. Stuart Chapin, Land Use Planning (Urbana: University of
Illinois Press, 1965), pp. 457 ff.
(1.11) This is probably especially true for problems which are
complex and ill-defined. For surveys of such sequences of activities
see for example:
Graham Wallas, The Art of Thought (New York: Harcourt Brace,
1926).
J. Hadamard, The Psychology of Invention in the Mathematical
Field,(Dover publications Inc., 1954).
And for suggested strategies to improve problem solving see:
George Polya, How to Solve It (New York: Doubleday, 1957).
Newell, Shaw, and Simon,"The Processes of Creative Thinking,"
Contemporary Approaches to Creative Thinking, Gruber et al, editors (New
York: Atherton Press, 1964), pp. 63-119.
(1.12) A fascinating discussion of the designer, the design and the
representation of the design as well as ways in which the representation
is helpful to the designer occurs in:
Gordon Pask, "The Conception of a Shape and the Evolution of
a Design," Conference on Design Methods, J.C. Jones and D. Thornley,
editors (Oxford: Pergamon Press, 1963), pp. 153-168.
(1.13) Manuals are generally available which describe general purpose
languages for scientific and commercial application. (eg. FORTRAN, COBOL)
(1.14) The DISCOURSE interpreter permits a wide variety of new
instructions to be written by the user. These are called "RCOM files" in
this thesis and there are many examples in Chapters IV and V and in the
Appendix. Where shorter execution times are demanded, any RCOM file can
be written in AED and called as a basic command of the DISCOURSE language.
PAGE 188
FOOTNOTES
CHAPTER I: FOOTNOTES
(1.15) ICES System: General Description, Daniel Roos, editor (Cambridge:
Department of Civil Engineering, Massachusetts Institure of Technology, 1967)
(1.16) D.T. Ross, AED-O Programming Manual--Preliminary Releases
Nos. 1-4 (Cambridge: Electronic Systems Laboratory, 1964).
1 spent over two years as part of a team working on the plan(1.17)
ning and design of Ciudad Guayana. For further information about the project see:
Lloyd Rodwin and Associates, Planning Urban Growth and Regional
Development: The Experience of the Guayana Program of Venezuela (Cambridge:
The Joint Center for Urban Studies, The MIT Press, 1969).
(1.18) Minsky's discussion deals with both the notions of algorithms
and functional equivalence which are the intuitive basis of the analysis in
Chapter II.
Marvin Minsky, Computation Finite and Infinite Machines
(Englewood-Cliffs: Prentice-Hall Series in Automatic Computation, George
Forsythe, editor, Prentice-Hall Inc., 1967), pp. 132-137.
(1.19) For the idea to represent DISCOURSE as a range rather than a
point on this continuum I owe my thanks to my colleague in the Department
of Architecture, Professor Leon Groisser.
For an explanation of TRANSET, see:
Earl R. Ruiter, ICES TRANSET I Transportation Network Analysis
Engineering Users' Manual (Cambridge: Department of Civil Engineering,
Massachusetts Institute of Technology, 1968).
CHAPTER II: FOOTNOTES
(2.01)
This work began as a seminar in city design led by Professor
Aaron Fleisher in which were John Boorn, Christine Boyer, and Hans Bleiker.
There was agreement neither on the analysis nor the direction of research;
the interpretation of the work may, therefore, not be shared.
(2.02) The GPS (General Problem Solver) did reproduce the subject's
protocol with memarkable fidelity.
Newell and Simon, "GPS, A Program that Simulates Human Thought:
Computers and Thought, E.Feigenbaum and J. Feldman, editors (New York:
McGraw-Hill Book Company, 1963), pp. 279-293.
(2.03)
See FOOTNOTE (1.18).
(2.04) Please refer to the APPENDIX for some of the documentation of
this experiment. See also FOOTNOTE (1.17).
(2.05)
L. Rodwin, Op. cit., pp. 1-39.
PAGE 189
FOOTNOTES
CHAPTER II:
FOOTNOTES
(2.06) A great deal has been decided here: "land suitable for
industry" is only a small leap forward from "SLOPE615" toward making
the description of the environment equal the specification of the design.
For experiments of wider applicability it would probably be better to
pick environmental descriptions with less commitment.
(2.07) The problem of determining how well the generated design
"fits" the designers' pattern is difficult to solve with a rule. A
square by square comparison of the generated pattern gives a value no
different for values computed from very unsuccessful patterns. Examining
clusters of squares may yield values which are different from those
compiled from outlandish patterns. These issues seem to be similar to
many discussed under the name "Pattern Recognition" where the success of
the human pattern recognizer seems to have much to do with the context he
is able to provide which helps him to recognize the particular object he
is looking at.
O.G. Selfridge and U. Neisser, "Pattern Recognition by Machine"
Computers and Thought op. cit., pp. 237-250.
(2.08) Putnam introduces the notion of "law cluster" in which the
status of a law is provisional and depends for its importance upon its
degree of connectedness to other laws and upon the meaning of that to
which it refers. A new discovery, therefore, affects both laws and the
objects of meaning.
H.P. Putnam, "The Analytic and the Synthetic," Minnesota
Studies in the Philosophy of Science, Feigl and Maxwell, editors (Minneapolis,
University of Minnesota Press, 1962), Vol. I I.
(2.09) A particularly creative solution could be defined as one which
has a high degree of both novelty and and usefulness because of the types
of elements and relationships among them both in the solution and the
implied problem formulation. No solution can be tested on either criterion
without reference to considerations external to the created solution itself.
The novelty of a solution may be a product of the surprise
elicited by previously remote elements being brought into a new combination.
The combination may be achieved because the elements play complementary
roles in performing some task; because they are similar in some respect to
each other; or because they are joined by some third element which is common
to both in either one of the two previous respects. The sense of surprise
is often justified by the observers' knowledge that effort in the area of
the new solution had consistently met with failure or that the solution
represents the filling of a gap in previous knowledge.
Mednick notes that although the creative person may have a
relatively higher number of associations in all, the non-creative person
may be able to produce more associations at first; this might be ascribed
to the to the high strength of the bonds between elements which the non
creative person has and which prevent him from moving beyond to less strong
or more remote associations. Mednick also found a good correlation between
high scores on a remote association test and independent ratings of creativity
of the same students by their professors.
PAGE 190
FOOTNOTES
CHAPTER II:
FOOTNOTES
S.A. Mednick, "The Associative Basis of the Creative Process"
Psychological Review. 1962, Vol. 69, No. 3.
The usefulness of a solution may be judged in part by the
acceptance which it is rendered by people to whose advantage it ought to
be to accept it. The criteria used by officials granting patents run very
much along these lines. The same sorts of problems face administrators
when trying to judge relative merits of productive scientists.
J.H. Mc Pherson, "A Proposal for Establishing Ultimate Criteria
Its Recognition
for Measuring Creative Output," Scientific Creativity:
and Development, Taylor and Barron, Eds. (Wiley and Sons Inc. 1963).
(2.10) Some rules, at least for land planning, which would seem to
be at about the right level of specificity can be found in:
F. Stuart Chapin, "Location Requirements" and "Space
Requirements", Urban Land Use Planning op. cit., pp. 370-456.
(2.11) Design of the highway from the observer's point of view yields
horizontal profiles for which the reasons may not be obvious in plan view.
See:
C. Tunnard and B. Pushkarev, "The Paved Ribbon: The Esthetic
of Freeway Design," Man-Made America: Chaos or Control? (New Haven: Yale
University Press, 1963), pp. 157-206.
D. Appleyard, K. Lynch and J. Myer, The View from the Road,
(Cambridge: MIT Press, 1964).
(2.12)
G. Pask, op. cit.
(2.13) Polya suggests"working backwards" as one of the strategies for
improving the chances of problem solving.
G. Polya, op. cit. pp. 225-232.
(2.14) Bartlett's discussion of "artistic thinking" is illuminating
in this regard. See also his discussion of "interpolation" (with reference
to "working backwards") where the subject has both beginning and end points
to work with.
F.C. Bartlett, Thinking: An Experimental and Social Study
(New York: Basic Books, 1958).
CHAPTER III:
FOOTNOTES
(3.01)
See FOOTNOTE (1.18).
(3.02)
A.N. Whitehead, An Introduction to Mathematics, op. cit., p. 186.
"The notion of Law, that is to say, of some measure of
regularity or of persistence or of recurrence, is an
essential element in the urge towards technology, methodology, scholarship, and speculation."
PAGE 191
FOOTNOTES
CHAPTER III:
FOOTNOTES
A.N. Whitehead, Adventures of Ideas (New York: The MacMillan
1952), p. 139.
Also, see:
A.N. Whitehead, An Introduction to Mathematics,op. cit., pp.]07-1 2 0
for a discussion of functions.
Company,
(3.03) G. Hodge, "Use and Mis-Use of Measurement Scales in City
Planning," Journal of the American Institute of Planners, Vol. XXVIII,
No. 2, pp. 112-121.
ment.
(3.04)
See:
Many communities have their own rules concerning land developF.S. Chapin, op. cit., p. 311.
Principles and Practice of Urban Planning, op. cit., p. 126.
(3.05)
F.S.
Chapin, op.
cit., pp.
383-456.
(3.06) J. Herbert and B. Stevens, "A Model for the Distribution of
Residential Activity in Urban Areas," Journal of Regional Science, Vol.2,
No. 2, 1960.
(3.07) J. Ruesch and W. Kees, Nonverbal Communication: Notes on the
Visual Perception of Human Relations (Berkeley: University of California
Press, 1956), pp. 89-148.
Books,
(3.08) Susanne K. Langer, Philosophy in a New Key (New York: Mentor
The New American Library of World Literature, Inc., 1952), p. 228.
(3.09) These categories reflect a taxonomy developed by Donald
Appleyard which are partly evident in:
Donald Appleyard, "City Designers and the Pluralistic City,"
L. Rodwin and Associates, op. cit., pp. 422-452.
, "An Environmental Improvement Program for
Metropolitan Boston," (Cambridge: Joint Center for Urban Studies, 1966).
and R. Okamoto, "Environmental Criteria for
Ideal Transportation Systems," (Chicago: Barton-Aschman Associates, Inc.,
1968) (Mimeographed).
(3.10)
cussed in:
Filling the gaps and creating the gaps to be filled is disF.C. Bartlett, op. cit.
(3.11) Harvard's Sixth Urban Design Conference, "Designing Inter-City
Growth," Progressive Architecture (August, 1962).
(3.12) F. Maki and J. Goldberg, "Some Thoughts on Collective Form,"
(St. Louis: Washington University, 1962) (mimeographed).
(3.13)
H. Blumenfeld, "The Urban Pattern,"
The Annals (March, 1964).
FOOTNOTES
CHAPTER I II:
PAGE 192
FOOTNOTES
K. Lynch, "The Pattern of the Metropolis," The Future
Metropolis, L. Rodwin, editor (New York: George Braziller, 1961),
pp. 103-128.
(3.14) Rusch makes an interesting analysis of Picasso's series of
"Bulls" in which are identified two kinds of changes from drawing to drawing:
movement toward the final sketch, the larger changes; and movement toward
good form.
Charles W. Rusch,"The Psychological Basis for an Incremental
Approach to Architecture"(unpublished Master's thesis, The University of
California, Berkeley, 1966).
(3.15) See for example, L. Alloway, "City Notes," Architectural
Design, (December, 1961).
A. Rockman, "The City as an Information Processing Machine,"
Habitat, Vol. X, No. I (January/February, 1967), pp. 44-47.
For a condensed review of techniques for empirical research see:
K.H. Craik, "The Comprehension of the Everyday Environment,"
J.A.I.P., Volume XXXIV, No. 1, January 1968; for adjectival scales see:
C.E. Osgood, G.J. Suci, and P.H. Tannebaum, The Measurement of
Meaning (Urbana: University of Illinois Press, 1957).
(3.16)
See the discussion of "frames of reference" in:
J.P. Guilford," Frames of Reference for Creative Behavior in
Its Educational Implications (New York: John
the Arts," Creativity:
Wiley and Sons, Inc., 1967), pp. 189-203.
Some designers seem to work from simple geometric ideas to
distortions of these ideas to accommodate the particularities of the site
and the program. I carried out several such studies under the direction of
Louis I. Kahn, Architect of Philadelphia.
(3.17)
(3.18) The opportunity area as perceived is constrained by a number
of factors such as time pressure, and information available; these and more
are discussed in:
J.G. March and H.A. Simon, Organization (New York: John Wiley
and Sons, Inc., 1963) pp. 150-161.
(3.19) J. Boorn, "A Choice System for Environmental Design and
Development" (unpublished PhD. thesis, Department of Urban Studies, Massachusetts Institute of Technology, 1969).
(3.20) March and Simon discuss "aspiration levels" as being closely
related to "criteria of satisfaction"; they propose that " the aspiration
level tends to adjust to the level of achievement."
March and Simon, op. cit., p. 182.
If so, this is a strike against the normatively minded for it
suggests that the designer will tend to adjust goals or the fit between
goals and performance before he adjusts his performance.
PAGE
FOOTNOTES
193
FOOTNOTES
CHAPTER I I:
(3.21) Joachim F. Wohlwill, "The Physical Environment: A Problem
for a Psychology of Stimulation," The Journal of Social Issues, Vol. XXII,
No. 4 (1966), pp. 29-38.
CHAPTER IV:
FOOTNOTES
There is an excellent introductory survey of data organization
(4.01)
in:
Surveys, Vol.
G.G. Dodd, "Elements of Data Management Systems," Computing
I, No. 2, June 1969, pp. 117-133.
(4.02) In the present version of DISCOURSE there is a core-resident
array in which one bit position is assigned to each ATTRIBUTE at a LOCATION
(to each ATTXLOC).
If the ATTXLOC is there or not there it can be found
out very rapidly by searching that array. Thus all logical operations can
be performed very rapidly without having to read in the data which are
stored on the disk. The pointer structure is reserved for the CHARVARS, the
numerical information which is LOCATION specific and which may be changing
in its demands upon computer storage. Pointers also serve to gain access
to the various parts of the data from the two basic dictionaries of ATTRIBUTES
and LOCATIONS.
(4.03) For a precise definition of the use of the features of the
language and for further illustration of the use of the language see:
DISCOURSE Notation Manual I (Cambridge: Department of Urban
Studies and Planning, M.I.T., March, 1969) (Mimeographed).
DISCOURSE Reference Manual I (Cambridge, Department of Urban
Studies and Planning, M.I.T., March, 1969) (Mimeographed).
(4.04)
M.I.T.,
An engaging and clear explanation of EDIT can be found in:
O,G, Selfridge, The CP-67/CMS Primer, OSM-19 (Cambridge:
Information Processing Center, June 1966)(Mimeographed).
CHAPTER V:
(5.01)
FOOTNOTES
Please refer to the discussion in Chapter IV page 88, "step 6."
For further discussion of the uses of simulations see:
A. Fleisher, "The Uses of Simulations," Computer Methods in
the Analysis of Large Scale Social Systems, James M. Beshers, editor (Cambridge: Joint Center for Urban Studies, 1965). pp. 144-146.
(5.02)
PAGE 194
FOOTNOTES
CHAPTER VI:
FOOTNOTES
(6.01) Examples have been developed by students in subject "11.06"
dedicated to the analysis of design at M.I.T. in the spring of 1969.
They ranged from making clear the criteria of growth of residential areas
in a rapidly expanding city to making obvious but complicated computations
of floor-area-ratios parking ratios and the like of a center city district.
Michael O'Hare has been working on the development of a residential
location simulation which began with Harvard's New Communities Project.
(6.02)
John P. Boorn, op.
cit.
(6.03)
Kevin A. Lynch, "Environmental Adaptability,"
Vol. XXIV, No. 1, 1958/1.
J.A.I.P.,
(6.04) R. Stotz, "Inexpensive Graphics," Computer Graphics in
Architecture and Design, M.Milne, editor (New Haven: Yale School of Art
and Architecture, 1969)
(6.05)
Steven Coons, "Computer Aided Design,"
Ibid.
(6.06) H. Reichenbach, The Philosophy of Space and Time (New York:
Dover Publications, Inc., 1957).
The description of the discovery of non-euclidean geometries;
the destruction of the point of view that the axioms of geometry are true
a priori. Contains discussions of euclidean, non-euclidean and topological
spaces; and introduces the idea of "coordinative definitions:"
"They are arbitrary, like all definitions; on
their choice depends the conceptual system which
develops with the progress of science...Wherever
metrical relations are to be established, the use of
coordinative definitions is conspicuous. If a distance is to be measured, the unit of length has to
be determined beforehand by definition. This definition
is a coordinative definition.... If the definition is
used for measurements, as in the case of the unit of
length, it is a "metrical" coordinative definition."
H. Reichenbach, op. cit., pp. 14-15.
If euclidean or any other geometry is not inevitably a part
of the world which we see, but, instead, part of our conceptual apparatus,
it behooves us to choose the geometry which is most convenient for us at
the time. The same point of view is expressed by Carnap with respect to
with respect to the variables of the description (in DISCOURSE, the
ATTRIBUTES, CHARCONS and CHARVARS):
"...the decision to use certain types of variables
is a practical decision like the choice of an instrument;
it depends chiefly upon the purposes for which the
instrument--here the language--is intended to be used
and upon the properties of the instrument."
PAGE 195
FOOTNOTES
CHAPTER VI:
FOOTNOTES
R. Carnap, Meaning and Necessity (Chicago: The Univers i ty
of Chicago Press, 1956), p. 43.
(6.07) F. Harary et al, Structural Models: An Introduction to the
Theory of Directed Graphs (New York: John Wiley and Sons, Inc., 1965).
BIBLIOGRAPHY
PAGE 196
Bartlett, F.C. Thinking: An Experimental and Social Study.
Basic Books, Inc., 1958.
Bruner, Jerome S. A Study of Thinking. New York:
Inc., John Wiley and Sons, Inc., 1956.
Processes:
310.
Chapin, F.S.
1965.
London:
Sciences Editions,
"On Going Beyond the Information Given," The Cognitive
Readings. Englewood Cliffs: Prentis Hall, 1964. pp. 2 93-
Land Use Planning.
Urbana:
University of Illinois Press,
Lynch, K.A. "Quality in City Design," Who Designs America. Laurence B.
Holland, editor. Princeton Studies in American Civilization, No. 6.
New York: Anchor Books, 1966.
Minsky, M. Computation: Finite and Infinite Machines. Englewood Cliffs:
Prentice-Hall Series in Automatic Computation. George Forsythe,
editor, Prentice Hall Inc., 1967.
Newell, A. Shaw, J.C., and Simon, H.A. "The Processes of Creative
Thinking," Contemporary Approaches to Creative Thinking. H.E. Gruber
et al, editors, New York: Atherton Press, 1964.
Pask, Gordon. "The Conception of a Shape and the Evolution of a Design,"
Conference on Design Methods. J.C. Jones and D. Thornley, editors.
Oxford: Pergamon Press, 1963 . pp. 153-168
Polya, George.
How to Solve It.
Simon, Herbert.
Models of Man.
New York:
New York:
Doubleday, 1957.
John Wiley and Sons, 1957.
Whitehead, A.N. An Introduction to Mathematics.
University Press, 1958.
New York:
Oxford