James Miller's Living Systems Model
An Interpretation and Application of the Model
to Weak Signal® Research by Collaborative Design and Group Genius
Bryan S. Coffman
January 15, 1997
Part I: Information and Its Relation to Weak Signal
In 1978 James Miller published the most thorough cross-discipline
analysis and synthesis of the functions and behavior of livings systems
ever set down in one book (nearly 1100 pages of extremely dense and small
text). In a thoughtful and scientific way it spans systems from the individual
cell to the supranational system. The book is accessible to laymen but
also popular and useful for scientists and researchers.
As Miller moves up the chain of complexity from cell to
group to supranational system, he uses 19 generic, critical subsystems
to provide structure to the material. Nine subsystems process matter-energy;
nine process information, and one processes both matter-energy and information.
In Part I of this paper I'll address only the subsystems
that process information. It is clear that these particular subsystems
do not operate in isolation from the other ten, however in order to examine
the model in light of weak signal research, it will not be necessary to
consider them until Part II. The reader may find it of value to first
scan the series on Weak Signal Research that begins here.
First I'll list Miller's definitions so the reader will
have some understanding of the terms. Then I'll show how the subsystems
work together in a typical communication and information processing event.
Finally, I'll add comments that illustrate how the model relates to weak
signal research, and in particular, the implications for organizational
structure and development.
Miller uses the word "information"
in the same sense that the word "message" is used in weak signal
research. In Miller's model, the value of the message is implied before
receipt. In weak signal research, it's important to understand that many
messages that reach the organization may not be capable of being transduced,
decoded or associated, and therefore will have not information associated
with them. A message only becomes information after the receiving system
or subsystem assigns value to it by calculating an adjustment of its behavior
in state space. In weak signal research, information is used in two senses:
Claude Shannon's sense, as a measure of the uncertainty or surprise in
the communication event as a whole; and Norbert Wiener's sense, as a measure
of the degree of order (or complexity) in the system.
||Examples in the Enterprise
|Input Transducer (IT)
||"The sensory subsystem which
brings markers bearing information into the system, changing them
to other matter-energy forms suitable for transmission within
||Mail rooms, servers that receive
E-mail and attachments from outside of the organization, telephone
systems that may include receptionists, fax machines, individuals
who work in and support these functions, also policies governing
the use of these elements.
|Internal Transducer (NT)
||"The sensory subsystem which
receives, from subsystems or components within the system, markers
bearing information about significant alterations in those subsystems
or components, changing them to other matter-energy forms of a
sort which can be transmitted within it."
||In-boxes on desks, personal E-mail
systems, telephones on desks of individuals in the organization,
individuals assigned to perform these tasks, also policies governing
the use of these elements.
|Channel and Net (CH)
||"The subsystem composed of
a single route in physical space, or multiple interconnected routes,
by which markers bearing information are transmitted to all parts
of the system."
||Hallways, fiber optic cable, twisted
pair, line of sight for visual reception and transmission of information,
the arrangement of space within a building to support the receipt,
analysis and transmission of messages, individuals who act as
||"The subsystem which alters
the code of information input to it through the input transducer
or internal transducer into a 'private' code that can be used
internally by the system."
||Display screens, audio speakers
in telephones, the function of human minds to understand written
language or diagrams, individuals who translate or interpret messages
||"The subsystem which carries
out the first stage of the learning process, forming enduring
associations among items of information in the system."
||Filing systems, techniques for deciding
which messages go where, skills in analysis, synthesis, and assigning
value and meaning to messages, software used in the analysis process,
some databases and expert systems, regulated fields of analysis
such as accounting, individuals who perform these skills.
||"The subsystem which carries
out the second stage of the learning process, storing various
sorts of information in the system for different periods of time."
||Filing cabinets, disk drives and
removable cartridges, paper that has words or images printed on
it, voice mail systems, libraries, individuals responsible for
maintaining these systems.
||"The executive subsystem which
receives information inputs from all other subsystems and transmits
to them information outputs that control the entire system."
||Any individual or collection of
individuals or devices who receive messages, associate them with
past experience based on contents in memory, and then choose a
course of action that may alter the behavior or state of the system
or its components. This system may be distributed, instead of
centralized. Decider functions also use outputs for the purpose
of requesting specific inputs from other subsystems or from other
systems. Also systems that determine what type of messages to
scan for, which to admit,and which to turn away from the system.
The Associator, Memory and Decider together play 'Spoze.
||"The subsystem which alters
the code of information input to it from other information processing
subsystems, from a 'private' code used internally by the system
into a 'public' code which can be interpreted by other systems
in its environment."
||Keyboards, computer mice, microphones,
various components of computers, the use of writing instruments
to put diagrams or words on various surfaces, individuals assigned
to document processes on computer or paper in the form of minutes
or journals or notes.
|Output Transducer (OT)
||"The subsystem which puts out
markers bearing information from the system, changing markers
within the system into other matter-energy forms which can be
transmitted over channels in the system's environment."
||Mail rooms, network servers that
send E-mail, telephones, advertising media including television
and radio, individuals working in, with or in support of these
functions, also policies governing the use of these elements.
For reference, here is a diagram of the interaction of the
nine information processing sub-systems.
IT=input transducer; DC=decoder; NT=internal transducer;
AS=associator; ME=memory; DE=decider; EC=encoder; OT=output transducer;
CH=channel and net
Message Processing Between and Within
The diagram above shows the subsystems that have just been defined in
relation to one another. It also shows the subsystems of three systems
interacting in a data corrected, error free communication event. The transmitting
system composes a message, encodes it for general use and transduces it
across its boundary into an external channel and net. The message is received
by both a comparator system and the receiving system where it is transduced
inside these systems and decoded. The receiver system sends a copy of
the message in the form it was received to the comparator system. The
comparator system compares the message it received from the transmitting
system to the message it received from the receiving system. It then sends
a message to the receiving system that indicates any discrepancies in
the two messages in the form of an error correction signal. Now, the receiving
system is confident enough of the integrity of the message to have it
transduced internally and prepared for the Associator.
The Associator fills two roles. First it assembles
messages into collections, or "ecosystems". It builds composite
models that show how these sets of messages relate to one another to form
a whole. Then it compares these messages and message models to those stored
in Memory to look for similarities or differences. It looks in particular
for discrepancies that may indicate the system is moving away from homeostasis
(balance) in the form of internal or external threats. It may also identify
opportunities in the same way.
As an example, an accounting department may
collect data from many parts of the corporation and assemble these into
collections of data based on models. Then it looks at the results in light
of past experience to find clues that might help management determine
future plans of action. The message collections and the results of the
comparison to past collections are forwarded to management for decision-making.
Memory stores previous messages and models
or patterns of how these messages have interacted in the past. It also
keeps track of what decisions were made based on these models, and the
results of the implementation of these decisions--their relative success
or failure. This constitutes a library of strategies for success and failure
that the system will refer to and modify over and over again. The depth
and quality of this library, together with the breadth and quality of
the models of message "ecosystems" constitute a measure of the
information inherent in the structure of the system--its order and complexity.
The Decider examines the analysis assembled
by the Associator and chooses a course of action to take based probably
on a fuzzy logic. There are no clear cut decisions to be made by living
systems. There are always unexplored options, incomplete perception, and
the element of surprise. The Decider determines which state (or behavior,
roughly) the system will move into during the next time period. The set
of all possible combinations of states that the system could move into
is called its fitness landscape. Some states are more fit than others
(they usually yield more success), and these are represented by peaks
on the landscape. As internal and external conditions change, the landscape
may deform: yesterday's peak may be tomorrow's valley. If an organization
finds itself cycling among a finite set of states, then it is said to
be cycling in an attractor--it's attracted for some reason to habitually
pursue this set of behaviors. To continuously improve, organizations must
discover and apply algorithms to help them explore their fitness landscape,
find efficient attractors that include high fitness peaks and continually
learn and evolve as the landscape shifts beneath their feet.
The Decider may include senior management,
but it is not and should not be limited to that small group. Network organizations
rely upon a distributed Decider function to bring effective action to
bear on a local situation in a timely fashion. This becomes more crucial
as the rate of change accelerates, or the rate at which the fitness landscape
deforms as a result of internal and external adaptation increases.
The Role of the Comparator
The purpose of the comparator is to test and provide messages on the integrity
of the transmitted signal. Here is John Pierce's diagram of how this process
works in communication and information theory, in particular as it relates
to the electronic transmission of data. It is similar to the diagram above,
but uses different terminology.
How is this done in practice? Consider a
conversation, discussion or dispute. Sometimes there is actually a third
party who acts as a "fair witness" in a discussion between two
other individuals. The fair witness listens to a statement from one individual
and then both he and the receiving individual can indicate what they heard.
Of course, this is not really comparing messages--it's comparing the values
and meaning assigned by these two systems to a message. Other times, there
may be only two individuals and the discussion involves statements followed
by replies that begin with the words, "by that, do you mean..."
In this way we act as comparators for ourselves. The work of the comparator
is a part of the function of the Associator.
Implications for Weak Signal Research
and Organizational Structure and Development
There's one interesting twist with respect to information theory
and weak signal research: the receiver is almost never the intended target
of the weak signal. Instead, the receiver is usually an eavesdropper on
some other conversation in a different industry or discipline. Like a
sonar operator sorting out friendly and enemy signals. There's nothing
covert or clandestine about this, however, it has important implications
for the role of the comparator and the internal transducer.
Enterprises that receive weak signals, will usually have
difficulty testing their integrity because they are difficult to decode
and associate to existing models. A real estate finance company may think
that some recent genetic research may have either a direct, or more likely,
indirect or possibly metaphorical effect on its work. But chances are
that nobody in the company can translate the terms of art in genetic research
into something they can understand and apply. Once a layman's book on
the subject is published, this becomes less of a problem, but then the
signal is also not so weak anymore, and exploiting it is up for grabs
among many competitors. The enterprise, therefore must enlist or cultivate
special assistance in finding primary research, translating it, and synthesizing
it into forms that will be understandable and therefore potentially useful
for the organization's Decider function. This means it must seek out alert
input transducers so that the messages from other disciplines are not
blocked or discarded, knowledgeable decoders to translate the messages
into something the associator can handle, competent internal transducers
to get these strange messages to the right destinations with the right
descriptions, a flexible associator that can build models using outside
data, and competent comparators to confirm their interpretation.
This process yields the great side benefit of building a
broader and more resilient Associator. Imagine the power of a company
that can actually associate real estate knowledge with genetic engineering,
and use the resulting synthesis to leverage market supremacy or even dominance
in their industry. Or maybe even reshape the industry.
On another, related note, most information processing
subsystems in organizations are tuned to receive and process standard
signals--ones that are expected (even though their content and value may
be surprising). Research departments scan the environment for these signals
and often have standard procedures for procuring them. Such messages may
relate to the economy, politics, competitors. Large systems are usually
put into place to gather messages from internal and external system components
to aid in the manufacture of tangible and intangible products and services.
Because so much of this is routine, enterprises must engage
in special efforts to redesign each subsystem to process weak signals--surprising,
disturbing, or messages that seem unrelated or irrelevant to the work
at hand. The typical response to this need for redesign is to fashion
a new department and charge it with the task of handling weak signals.
These efforts will fail. They will fail because, even though the new system
has remodeled all of its information processing subsystems to handle the
new type of messages, the subsystems of the parent organization are still
unable to process the information or understand its value. The only true
solution is to distribute the capacity and demand for weak signal processing
broadly in the organization and allow individual subsystems to adapt to
local conditions. Management provides a portion of the stimulus for redesign,
but the processing of weak signals must develop an intrinsic worth or
value for the organization to continue to lend genuine focus to the activity.
Part II: Matter-Energy Subsystems
Part III: Reproduction Subsystem and Evolution
The components of the model and its
definitions are copyrighted by James Grier Miller in his book Living Systems,
copyright © 1978, McGraw-Hill. All rights reserved.
interpretative work copyright © 1997, MG Taylor Corporation. All rights
terms and conditions