What Is the Systems Approach?

Author(s): Gordon K. C. Chen

Source: Interfaces, Vol. 6, No. 1, Part 1 of Two (Nov., 1975), pp. 32-37
Published by: INFORMS
Stable URL: http://www.jstor.org/stable/25059250
Accessed: 24-07-2015 19:24 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact [email protected].

INFORMS is collaborating with JSTOR to digitize, preserve and extend access to Interfaces.

http://www.jstor.org

This content downloaded from 103.27.8.46 on Fri, 24 Jul 2015 19:24:11 UTC
All use subject to JSTOR Terms and Conditions

INTERFACES
Vol. 6, No. 1
November
1975

APPROACH?
WHAT ISTHE SYSTEMS
K. C. Chen

Gordon

of Massachusetts

University
Gordon

K.

C.

Chen

at the
of management
professor
to Management
He
is a contributor
recent
and Systems
book, Operations

is associate

of Massachusetts,
Amherst.
University
and
is the co-editor
of the
Science
A Simulation
Approach.
Analysis:

term "Systems Approach"

The
in manage
has appeared
frequently
and systems literature
In
of
its
use, its mean
recently.
spite
popular
a
is
seldom
Even
definition
is given,
defined.
when
and
where
ing
clearly
seems
as to exactly what
there is little agreement
confusion
it means. The
to stem from the fact that a "system" often means
different
things to dif

ment

ferent

systems.

an

To

people.

mean

it may

engineer

To

object.

physical

an

econ

of operation
be a mode
such as the free enterprise
system or
a
mean
To
it
either
the hard
scientist
may
system.
computer
or the software
or
of the computer
information
programs

it may
omist
an inventory
ware
system

the

Thus,

term

has

so

become

it

that

generic

is almost

meaningless.

a concept
of this paper
of the
and definition
is to develop
purpose
a
as
for solving problems.
systems approach
methodology
then do we mean by the "Systems Approach?"
What
uses the familiar
Churchman
fable about several blindmen
each touch
a
an
of
to
different
his
of
The
illustrate
part
concept
systems.
elephant
ing
one should not
of the story is that when
with a problem
moral
confronted
just look at a part of it but should, instead, view the whole picture which may
be referred to as the total systems concept.
The

Such
case.

concept
it

However,

For
questions.
whole
problem

also

has

certainly
raises

some

how

are

example,
not
and

just

its

part

of

it,

in

especially

appeals,

important
we
to know

conceptual
we
are
since

and
not

in many

the

elephant

methodological
with
dealing
respects

our

the
per

In fact, the blindmen

is limited to what we can sense and perceive?
ception
more
all
and
evident
these days
is
and elephant
the
syndrome
prevalent
more
are becoming
as the problems
and more
Take
the drug
complex.
how many
different
this
views are there concerning
for example:
problem,
law enforcement
officers can give us one answer
social ill? The
wide-spread
and the sociologists,
another. The
may have a theory for it,
psychologists
can cite a few more
case histories.
Even among
researchers
and the medical
in the same field, their opinions may differ widely.
themselves
the experts
is the type of problem where
the elephant
Unlike
problem,
drug addiction
no one knows what kind of monster
it really is. Maybe
all of the experts'
more
are
of the ele
than
the
blindmen's
interpretation
nothing
opinions
all.
after
phant
to the total systems concept
relevant
second question
is: "If a
The
a
and not in part, what
constitutes
in whole
is to be considered
problem
are as clear cut as that of
all problems
whole
and what does not?" Not
an elephant whose
is at least visible and identifiable.
total picture
describing
and po
of the problems,
Most
those
social, economic,
involving
especially
Copyright

1975, The

Institute

of Management

32

This content downloaded from 103.27.8.46 on Fri, 24 Jul 2015 19:24:11 UTC
All use subject to JSTOR Terms and Conditions

Sciences

for
litical behavior,
elude such identification.
the traffic problem,
Consider
example: how should we view it in terms of a whole
system? Should we take
into consideration
such factors as the pedestrians,
the sizes, shapes, or noise
we
since
of
etc.?
draw
the
line? Furthermore,
should
Where
levels
vehicles,
a
a
of
is
far
how
the
should
every system
system
subsystem
larger system,
it has been said in the marketing
field
be drawn? For example,
boundary
in trans
that had the buggy whip manufacturers
realized
that they were
business
instead of buggy whip business
portation
they probably would have
as it
survived
This
is akin to the total system viewpoint.
Plausible
today.
may

the

sound,

cannot

concept

be

too

stretched

far,

ren

without

however,

itself meaningless.
the buggy whip
business
What
if, for example,
dering
were regarded as in a business, period,
instead of in any particular
business;
would
this broader,
all-inclusive
the industry
solve its
concept have helped
on
to the question
not so much
the answer
problems? Obviously,
depends
the type of business
it was in as to how it viewed and defined
its problems.
as a marketing
If the industry had insisted on viewing
its declining
business
its efforts on promoting
sales in a decaying market,
problem
by directing
to find itself losing out in the end. On
it should not have been surprised
the other hand,
for buggy whips
if it was aware that the falling demand
was a natural
it would
have defined
of changing
consequence
technology,
the problem
in those terms and coped with
it as such.
The preceding
the crucial
illustration
suggests that in problem-solving
we are dealing with
is not just whether
the total system or where
question
we draw the system boundary,
of
and attributes
but what proper
entities
the system should be. In other words,
is whether
the important
question
or not we have identified
illustrate how a problem
the right problem. To
can be solved readily if it is properly
and defined,
identified
the following
examples may give further illumination.
a problem
in a new
In his recent book, Rudwick
cites a case involving
four
with
The
has forty stories and is equipped
office building.
building
a few months,
elevators
serving all forty floors. Opened
complaints
only
that it took
into the building
office alleging
began pouring
superintendent's
too long to wait for the elevator on the first floor during
rush
the morning
hour. This would have ordinarily
been viewed as a classical queuing
problem
in management
science and treated as such. In fact, several attempts were
made

to

reduce

automatic
improve

readjustment

device

to

expedite

the

elevators.

much,

an

operator

was

assigned

the

control
matter

the

time.

waiting

First,

was

to

to

made

that

When

the

regulate

the

not

did

eleva

to be unsatisfactory
also. Finally
tors manually
in the morning.
This proved
After
the problem.
was called
in to investigate
the elevator manufacturer
analyzing
native

the
solutions.

situation,
One

the

manufacturer's

was

to

engineers
the

replace

existing

suggested
control

several
device

alter
with

some
more
such as directing
one having more
flexible
features
complex
was estimated
to cost
cars for expresses
and some for local stops. This
to add another
to be in
was
elevator
alternative
about $7,000. Another
a glass enclosure
the
which
with
of the building
stalled outside
provides
an advertising
an outside
effect.
with
view and the building
riders with
was de
cost about $50,000. While
the building
This would
management
an
came
one
with
cons
some
these
of
the
and
up
pros
proposals,
bating
a
was
cost
at
How
of
idea
the
that
solved
only $300.
problem
ingenious
It was
the installation
the solution was quite
this done? Actually
simple.
so that the riders could
set in the lobby near the elevator
of a television
to the first
to return
for the elevator
watch
the program
while waiting
INTERFACES

November

1975

33

This content downloaded from 103.27.8.46 on Fri, 24 Jul 2015 19:24:11 UTC
All use subject to JSTOR Terms and Conditions

The
with
riders were no longer
and com
impatient
long waiting
subsided.
plaints
Now what
lesson can one learn from such an experience?
An obvious
one is that how and how well a problem
is solved depends
entirely upon
as
case was viewed
how and how well
it is defined. When
the elevator
an elevator delay problem,
to
to
focus
tended
solve
the problem
attempts
on the reduction
as a problem
of the delay time. But when
it was defined
of boredom
from waiting,
the solution was simply to ease the boredom.
to demonstrate
The
that in problem-solving
it is
also helps
example
to look at the total picture but is essential
neither necessary nor sufficient
to locate the key issues underlying
from
the problem which may be obscured
or may
and immediate
the surface of the problem
lie beyond
its obvious
it
after all had
realms. Perhaps
the elephant
story is not too farfetched
blindmen
the
the
in a different
describe
casted
been
say,
light,
having
the correct answers.
skin of the elephant. They would
likely have provided
a way of per
in short then, is essentially
The
"Systems Approach",
on
a
and
and
by identifying
focusing
thinking
through
problem
ceiving
to
In
other
in
thereto.
order
the critical elements
words,
apply
pertaining
to know about
and
of the problem
the nature
systems approach we need
we are dealing with
first. But what
is a problem?
what kind of problem
And how can it be identified?
as the devia
From a systems point of view, a problem may be defined
state of affairs at a given point of time.
tion of the actual from the desired
as follows:
It can be defined
floor.

Pt=\Dt-At\
Where Pt =
=
Dt
=
At
t=

Problem

at time t
state of affairs at time t

Desired

state of affairs at time t

Actual
A point

in time

and given an equilibrium

By this definition
can exist if and only
a problem
i.e.
At
Pf > 0
exists.
ditions
actual

a) The
i.e.

state remains

Dt?:Dt_1

the same, but

position
if any of
the desired
is

givenAt=At_i=Dt_i.This

goal

between
Dt
the following
state has

changed,

changing

problem

as a Type
I problem.
state
The
remains
the
but
desired
the actual state has
same,
b)
is a goal seeking
i.e.Af ?zAt_1gi\enDt
=Z)i_1
=^4f_1.This
as a Type
II problem.
and is to be identified
and

shall be

identified

and
con

here

and the
the desired
c) Both
i.e. Dt =?Dt_ x and
positions,
as
is a goal changing
This
at the same time and is to

changed;
problem

to different
have changed
=
given
At ?=At_1
andZ>f_ t At_x
Dt?*At
a
well as
occuring
goal seeking problem
be called a Type
III problem.
actual

states

How
this classification?
is it going to help identify and solve prob
Why
let us first examine
in
lems? Before answering
these questions
the difference
the Type
characteristics
between
the two basic types of problems,
namely,
II problems.
I and Type
are essentially
I problems
As stated earlier, Type
goal changing
prob
are goal seeking problems.
the ob
II problems
lems and Type
Therefore,
34

INTERFACES

November

This content downloaded from 103.27.8.46 on Fri, 24 Jul 2015 19:24:11 UTC
All use subject to JSTOR Terms and Conditions

1975

For example, when

of the two are entirely different.
jectives and purposes
a
a
new
an organization
it is faced with
is contemplating
market,
entering
a
But
if
it
to
is
I,
attempting
regain
goal changing, Type
declining
problem.
II problem. The
former
market
share, it is dealing with a goal seeking Type
case
as

may
a

as a preventive
regarded
or
action.
remedial

be

corrective

or

process,

anticipative

and

the

latter,

as a prob
the former may be viewed
the systems concept,
of
and
the
latter,
systems
design
analysis and main
synthesis,
on systems
former involves
with
systems prognosis
emphasis
with
while
the latter involves
systems diagnosis
forecasting,
a
and control. Thus,
the former assumes
systems guidance
and
that
is
of
role
initiating
creating,
leading,
learning,
leadership
capable
an action; while
and path of a pre
the direction
the latter merely
follows
action only when
determined
and is to take corrective
goal and destination
a deviation
occurs. In terms of systems hierarchy,
the former is a system of
a dominant
the highest
order in which
role, and the
plays
being human
can often now routinely
be
to a lower order system which
latter belongs
or
means.
handled
other
mechanical
by computers
in terms of the systems concept
and having
defined
Having
problems
to apply
we are now
in a position
the
delineated
their characteristics,
the two major
Since
these problems.
toward
systems
approach
solving
and are
and characteristics
have different
types of problems
objectives
to
of
their
of
the
different
solution,
hierarchies,
necessity,
systems
approaches
also take different
forms.
to Type
starts with a clear definition
I problems
The
systems approach
to the old one from which
state or goal as opposed
of the new desired
dimensions
it has changed.
In defining
the new desired
state, all relevant
to
much
be
need
and
how
such as what, where,
when,
specified.
who,
a new market,
it must
in entering
For example,
if a firm is interested
at
area
or
what
to
in
which
be
introduced
is
what
service
product
specify
In terms
lem of systems
tenance. The
and
planning
on
emphasis

time,

who

of

will

be

its

target

and

customers,

how

many

are

expected,

etc.

and
it should be analyzed,
the new goal is defined,
Once
questioned,
new
state
that
desired
of
dimension
This
is
the
"why". Any
challenged.
a
value
or is weakened
has
the
fails to meet
challenge
questionable
by
or redefined.
In defining
the new desired
state,
and should be reconsidered
state.
to define what
is not the desired
it is important
is, as well as what
a
what
is
so
and
is
will
what
delineate
not, and
problem
help
Doing
It
considerations.
irrelevant
and rule out
draw a clear system boundary
a problem.
a desired
state is not defining
that defining
should be noted
state and the actual
between
the new desired
But defining
the difference
a problem
or existing
is to be solved
How
state is defining
the problem.
not
but
is
how
defined
also
the
upon how
upon
problem
depends
only
state is defined.
the desired
is always surrounded
it can be achieved,
state, before
Any new desired
or
constraints may be of a re
sorts
These
all
obstacles.
of
constraints
by
source,

technological,

environmental,

natural

or

even

artificial

nature.

There

is the assessment
tasks in the systems approach
fore, one of the important
often
takes the form of a feasibility
constraints.
of all conceivable
This
or re
can be avoided
etc. Some constraints
test project,
study or a pilot
or
constraints
trade
offs.
Some
moved
system designs
by
by alternative
to avoid or remove. For example,
are logically
it is logically
impossible
a
one
a
toss
a
of
coin. In such
to
tail
with
head
and
both
get
impossible
state is unfeasible.
cases, the new desired
INTERFACES

November

1975

35

This content downloaded from 103.27.8.46 on Fri, 24 Jul 2015 19:24:11 UTC
All use subject to JSTOR Terms and Conditions

a new desired
state is defined
and deemed
feasible and once a
a
is
it
well
should
be
followed
defined,
process.
problem
by
simplification
that the systems approach
One
of the contributions
in problem
makes
is its emphasis on the various
and methodologies
it has
solving
techniques
for system simplification.
The process may
involve
the simplifica
developed
or removal
tion of the new desired
of constraints,
state, the simplification
or the simplification
of the problem
itself.
defined
the new desired
from the actual
state, its difference
Having
the task of the systems approach
is to find ways
state, and the constraints,
or reducing
of removing
and means
the difference.
is an alternative
This
Once

process.

generating

pattern

scanning,

are

There

them are such methods

recognition

of

generating

inquisition,

logical

ways

many

as creative
and

search

alternatives,

combinational

among

dimensional

extension,

cumula

analysis,

etc.
and learning,
simulation
and modelling,
computer
experiments
details are too numerous
and beyond
the scope of this paper.
once generated
must
to a
The
alternatives
be evaluated
according
set
of
criteria
and
of
A
number
been
orders
of
have
methods
given
priorities.
to evaluate and prioritize
them are such tech
alternatives,
among
developed

tive
The

niques

as

decision

analysis,

and

cost-benefit,

cost-effectiveness

etc.

analyses,

It must be noted, however,

because of the individual
of different
preferences
decision makers,
the final selection
of the alternatives
may or may not be
state originally
consistent
with
the new desired
In such cases,
prescribed.
a
to the problems
the solutions
instead of
be
may
compromise
simply
an optimal or rational one.
to Type
assumes
state is
The
II problems
that a desired
approach
is to diagnose
the problem
the cause
already in existence or well established;
a Type
of the actual state deviating
II
from it and to find remedies. Thus,
must
a
are
certain
conditions
which
with
There
symptom.
problem
begins
can be detected.
be satisfied before a symptom
For example:
and measurable.
a) A symptom must be observable
the aid of an information
requires
b) Its detection
a memory,
put

sensor,

a measuring

system,

comparator,

system;
and

namely,
an

out

system.

absence of a signal can be the presence of a sympton.

c) The
of the sympton may be totally unrelated
to
location
d) The
cation of the problem.

the

lo

are available
Various
and methods
for detecting
certain
techniques
of
The
medical
has the most advanced
types
symptoms.
symptom
profession
detection
and devices. Management
Science has in recent dec
techniques
some powerful
ades developed
tools and techniques
in handling
large and
but still largely relies upon experience,
complex problems,
insights, and in
to locate problems.
Behavorial
tuition
scientists have recently
experimented
or modify
with
to
various
but there
human
behavior,
techniques
predict
are still reservations
are
their
wide
There
limiting
encouraging
applications.
are making
advanced
information
systems
management
signs that more
to receive early
for management
and may now be possible
great headway,
erupts.
warning
signals before a problem
to management,
it is still
Even with
the early warning
signs available
a
cause
in
of
is
the
the
process
ill-equipped
problem. Diagnosis
diagnosing
case
to the causes. For example,
in the elevator
of relating
the symptom
as
were
cited earlier,
the riders' complaints
diagnosed
being
(symptom)
a
boredom-from
of
an elevator
instead
delay problem
(wrong diagnosis)
36

INTERFACES

November

This content downloaded from 103.27.8.46 on Fri, 24 Jul 2015 19:24:11 UTC
All use subject to JSTOR Terms and Conditions

1975

finds a
which
problem
problem
(real cause). For every elevator
of others which do not. Thus,
the
there must be hundreds
solution,
a continuing
to which
future research
process presents
challenge
diagnostic
once
II situation,
In the Type
should be directed.
in the systems approach
the rest of
the problem
is accomplished,
the difficult
task of diagnosing
to the one used for solving Type
I problem,
the approach
is comparable
and follow ups.
alternatives
selecting, and implementing
namely, generating,
of the Systems Approach.
The
pages are a brief overview
preceding
as an aid to problem-solving.
It offers a logical and consistent
definition
or uniqueness.
No
it helps
claim is made
of its completeness
Hopefully,
contribute
human
toward a better understanding
of the complex
problem
to problem
then, the systems approach
solving process. In summary
solving
and
is a way of thinking
the nature
upon
through a problem
depending
to be handled.
the type of problems
It takes a system synthesis and prog
nostic approach
is of the goal changing nature and uses sys
if the problem
tems analysis
if it is a goal seeking problem.
It
and diagnostic
approach
It is interested
specifies as well as classifies,
analyzes as well as synthesizes.
not only in "What is?" but also in "What is not?"
waiting
correct

References
to Cybernetics,
1963.
& Sons, New York,
Introduction
W.R.,
John Wiley
and Management,
1959.
& Sons,
Inc., New
York,
Stafford,
Cybernetics
John? Wiley
of Science" Management
Skeleton
Kenneth
The
E., "General
Boulding,
Systems
Theory
1956.
Science, April,
of Faulty
Vol.
16,
Chu, Wesley
W.,
Research,
Operations
"Adaptive
Systems",
Diagnosis
No. 5, Sept.-Oct.,
1968.
C. West,
The
1968.
Delacorte
Press, New
York,
Churchman,
Systems
Approach,
M. and James G. March,
Prentice
A Behavioral
Hall,
of the Firm,
Cyert, Richard
Theory
Cliffs, N. J., 1963.
Incorporated,
Englewood
"A
Search
of
Vol.
No.
Dobbie,
3,
16,
M.,
Research,
Theory",
Survey
Operations
James
1968.
May-June,
a
is
"What
Vol.
No.
1969.
Eilon,
Samuel,
Decision?",
4, December,
16,
Science,
Management
Albert N., "Management's
1969.
March,
Automation,
Faught,
System", Business
Early Warning
Edward
and Julian
Feldman
McGraw-Hill
and Thought,
Feigenbaum,
Computers
(Ed.),
Book Company,
New York,
1963.
Incorporated,
Foundations
New
1958.
Feinstein,
Amiel,
York,
McGraw-Hill,
of Information
Theory,
Peter C, Decision
and Value
& Sons, Inc., New
1964.
Fishburn,
York,
Theory,
John Wiley
Industrial
MIT
1961.
Forrester,
Press, Cambridge,
MA,
Jay W.,
Dynamics,
I?aul J., "Heuristic
You Can Do
Problem
Vol.
Gordon,
5,
It", Business
Horizon,
Solving,
No.
1, Spring
1962, pp. 43-63.
in a System
for Computer-Aided
A., "Problem-Solving
Gorry,
George
Strategies
Diagnosis",
MIT
Sloan School
of Management,
1967.
Working
Paper #268-67,
June
Carl E., The Management
McGraw-Hill
Book Co., New
1967.
York,
of Intelligence,
Gregory,
D. Van Norstrand
Hall, Arthur
D., A Methodology
Co., Incorporated,
for Systems Engineering,
1962.
Princeton,
N.J.,
C. H. and B. B. Tregoe,
The Rational
McGraw-Hill
1965.
Co., New
York,
Kepner,
Manager,
"Artificial
1961.
M.,
Minsky,
Intelligence",
of IRE,
Proceedings
January,
A. and H.A.
of Human
Simulation
Newell,
Simon,
December,
Science,
"Computer
Thinking",
1961.
Stanford
and Industrial
Problem
Prentice
L., Systems
for Business
Optner,
Analysis
Solving,
1965.
Inc., Englewood
Hall,
Cliffs, N.J.,
William
Process
of Problem
Industrial
Pounds,
F., "The
Review,
Finding",
Management
Fall, 1969.
Walter
and Thought:
An
Reitman,
R., Cognition
Approach,
Information-Processing
John
& Sons, Inc., New York,
1965.
Wiley
Bernard
and Sons,
Rudwick,
H.,
Inc.,
for Effective
Systems
Analysis
Planning,
John Wiley
New York,
1969.
S. Sankar,
Observables
and Probabilities",
"Structure,
Sengupta,
paper,
presented
working
at TIMS
1968. Department
of Economics,
Cleveland,
OH,
Meeting,
September,
University
of Waterloo,
Canada.
Waterloo,
Ontario,
Science
York University.
Simon, Herbert
A., The New
Decision,
of Management
1960, New
on Problem
manual.
Singer, Frank A., "A Program
Solving", mimeographed
self-instrucjtjonal
School
of Business
of Massachusetts,
1968.
Administration,
MA,
Amherst,
University
& Sons, Incorporated,
New
Wiener,
1948.
Norbert,
York,
Cybernetics,
John Wiley
A Systems
Foresman
and Co., Glenview,
111.
Young,
Scott,
Analysis,
Stanley, Management:
1966.
Ashby,
Beer,

INTERFACES

November

1975

37

This content downloaded from 103.27.8.46 on Fri, 24 Jul 2015 19:24:11 UTC
All use subject to JSTOR Terms and Conditions