A cybernetic view of description and reality
Copyright 2014 Graham Berrisford. Now a chapter in “the book” at https://bit.ly/2yXGImr. Last updated 22/04/2021 11:20
THIS IS A FIRST DRAFT
All social animals hold information in memory, and communicate it in messages. This chapter outlines some information and communication theory. It features a WKID hierarchy, and the principle that in symbolic communication, coding is ubiquitous. It goes on to describe how message senders communicate, and so share knowledge, with message receivers.
“To understand is to know what to do” Wittgenstein
Don’t think of information processing as a side issue, peripheral to something else. Our lives and businesses depend on creating, processing, using and sharing information. This is evident from the hundred words below that we have for doing those things.
Abstracting, Accounting, Agreeing, Analyzing, Answering, Architecting, Asserting, Attributing, Averaging, Believing, Billing, Briefing, Calculating, Characterizing, Chronicling, Coding, Communicating, Composing, Conceiving, Counting, Deciding, Declaring, Decoding, Defining, Demonstrating, Describing, Designing, Detecting, Directing, Drawing, Elucidating, Encoding, Envisaging, Equating, Explaining, Expressing, Factualizing, Fictionalizing, Generalizing, Hearing, Hypothesizing, Idealizing, Illustrating, Informing, Instructing, Judging, Justifying, Knowing, Lying, Mapping, Measuring, Messaging, Modelling, Narrating, Observing, Orating, Ordering, Perceiving, Picturing, Postulating, Presenting, Proposing, Qualifying, Quantifying, Questioning, Reading, Recording, Recounting, Reifying, Remembering, Rendering, Reporting, Representing, Reproducing, Sensing, Sorting, Speaking, Specifying, Stating, Story-telling, Summarizing, Testing, Theorizing, Thinking, Totaling, Translating, Transmitting, Truth-telling, Typifying, Understanding, Valuing, Verifying, Viewing, Voting Visualizing, Worrying, Writing, X-raying, Yelling, Zeroing.
The importance of information in enterprise architecture
What makes a society "social" are the communication events by which actors exchange information. The information includes descriptions of the world, declarations, decisions and directions about what is to be done.
A human institution, organization or enterprise is a social entity in which some or much information is standardized. For thousands of years, government and commercial activities have been conducted in response to regular, repeated messages such as orders, payments, applications for positions or assistance, notifications of decisions, appointments and other events, along with declarations of goals and plans. Information appears not only in messages but also in records, role and rule definitions. The messages, records, roles and rules contain information that actors rely on to understand should be done, and to do it.
"The Information Age" is the name given to the shift, starting in the 1960s, from industries established by the Industrial Revolution to an economy enabled and supported by information technology. The term "Enterprise Architecture" (EA) was coined in the 1980s in relation to the impact of this shift on business activity systems that create and use business data.
"In 1982, futurist and inventor R. Buckminster Fuller estimated that up until 1900, human knowledge doubled approximately every century, but by 1945 it was doubling every 25 years. And by 1982, it was doubling every 12-13 months. In retrospect, this may sound a little quaint since experts now estimate that by 2020, human knowledge will double every 12 hours. But the real question is, “How is it making us smarter?”"
The source seems to be a paper by IBM https://ia601003.us.archive.org/2/items/TheToxicTerabyte/The%20Toxic%20Terabyte.pdf.
"It is projected that just four years from now, the world’s information base will be doubling in size every 11 hours.”
If you send an email to 1,000 people rather than one, have you created 1,000 times more knowledge? In natural language, the terms data, information and knowledge are often used interchangeably. How to distinguish them?
In natural language, the terms data, information and knowledge are often used interchangeably.
Data? To remember and communicate about phenomena they observe and envisage, animals construct data structures in memories and messages. All data structures are physical; they are organized forms of matter and energy.
Information? The meanings of data structures in messages and memories appear only in their use by their creators and users. (In human and digital data storage, many different data structures encode the same information, e.g. the value of pi).
Knowledge? To know something, to understand a memory or message structure, is to know when and how to use it to do something. The information that an onrushing train will kill you is useful – it guides you to step off a railway track. Knowledge (along with emotions like love and fear) helps you to survive, thrive and pass your genes on. So, knowledge may be defined as information that is accurate enough to be useful, and science might then be defined as the sum total of human knowledge that it is useful to all.
I have tried to relate data, information and knowledge in the table below. And added wisdom, which implies something of the “insight learning” mentioned earlier. To be honest, one purpose for this table is to save us from having to discuss any of several other WKID variations that have been proposed and criticized elsewhere, for example here WKID hierarchies. This table might also be criticized, but it does help us (in this chapter) to discuss the exchange of knowledge in a social system of communicating actors.
the ability to apply knowledge in new situations.
information that is accurate enough to be useful.
meaning created/encoded or found/decoded in data by an actor.
a structure of matter/energy in which information has been created/encoded or found/decoded
You might say a frog’s eyes hold a data structure, a memory or pattern of the insect type. When the frog’s eyes detect an instance of the insect type – that meaningful information is encoded into a nerve impulse message. The frog’s brain decodes that message, processes it, then encodes another message sent to the frog’s tongue. If the tongue captures the insect, then the information was true, and may be called knowledge.
Can these things be quantified? Aside from counting digits in digital storage, much discussion of these things is unscientific. The amount data stored in some way is surely increasing exponentially. The amount of information or meaning extracted from that data must be increasing in proportion to the number of actors who use it, and the speed at which they can decode it. The amount of knowledge (here, useful rather than useless or misleading information) might be increasing somewhat more slowly. And the amount of deduplicated knowledge that is useful to all of us, and so increases the body of knowledge we call science must be less again. I guess it is increasing somewhat faster each year, but there may be a peak science growth year, after which its growth begins to slow?
Several information/data distinctions have drawn in computing science, including these.
Some contrast information in
with data in
a message or data flow
a memory or data store
a conceptual or logical data model
a physical data model
E.g. relating a) the notes on a page of piano music to b) the sounds made by the piano. The score for a piano concerto is a data structure, yet meaningless to anybody not musically trained. Information appears in the process of composing the score and playing the piano. The composer who creates the score must know the sound structure to be created. A piano player who uses the score should understand and produce the sounds the composer had in mind.
Data is any phenomenon that is variable (has a variety of values) and created or used to store or convey information. A phenomenon is used as data structure when it is a) encoded to convey information/meaning or b) decoded as conveying information/meaning. It can be any physical structure or motion, of matter or energy, that is created or used as a data structure or signal. E.g.
· The shadow on a sundial – used to by people as representing the time of day.
· The state of your office door (open or closed) – used by people tell when you are open to visitors.
· Dance movements – used to express emotions, or by a honey bee to describe the location of some pollen.
· The score of a piano concerto – used by a piano player to make music.
· Words – used as in this book to convey complex information.
· The biochemical structures in your brain – used to remember entities and events.
A description is meaningful to an actor only in the process of creating or using it
Information is not an intrinsic property of data; it does not exist in a data structure on its own. Information is an extrinsic property of data; it exists in only in process of encoding or coding some data, in relation to something else. That something might be a state of affairs observable in the real world, or some action to be taken.
In other words, there is no information or meaning in a data structure on its own. Data creators must perform processes to encode/create meanings in structures. And data users must perform processes to decode/find meanings in structures. The information of interest to sociology only exists in those processes - in the intentions or purposes of data creators and the interpretations of data users.
This chapter goes on to describe how message senders communicate, and so share knowledge, with message receivers.
This chapter is not about Shannon’s information theory. (Which relates to what can be coded in and retrieved from signals such as radio waves, and about the integrity of data structures.)
And we don’t begin here by thinking of information like a linguist or software engineer might do. Millennia before verbal languages and computing, animals encoded input signals into memories. Then decoded memories into the stream of consciousness when determining responses to events. And social animals encoded and decoded messages (say, alarm calls) without needing to learn a language.
Try to imagine how you’d mimic these incredibly effective processes in software. Inevitably, you’d find yourself drawn down paths you understand. You’d think of data structures – stored in databases, and transmitted in data flows. You’d presume the need for formally defined languages, symbols, syntax and semantics.
But surely, a truly general information theory cannot start from or depend on how software works? It ought to start from the evolution of information created and used by animals. Many animals can not only remember information internally in memories, but also share information messages.
Read this triangle left to right thus: Animals <create and use> Information <represents> Phenomena.
<create and use> <represents>
Animals <observe and envisage> Phenomena
The information at the apex is encoded in memories and messages.
Memories and messages: holders of information.
In biology, internal memories and external messages are of different kinds. Memories are neural patterns; messages take the form of sounds, smells and gestures. In software, the distinction between memories and messages is blurred.
Communication: the exchange of information between senders and receivers.
Actors may exchange information directly by sending/receiving messages. Or else, indirectly by writing/reading information stored in some memory both can access. They respond to information in messages, often in a way determined by information in memory.
A description must in correlate in some way to the described phenomenon. There are three ways a description can do this.
· An iconic description, like a statue or photograph, mimics some features of the described phenomenon,
· A signifying description, like the smoke of a fire, points to effects produced by a described phenomenon.
· A symbolic description encodes some features of the described phenomenon using some kind of code, verbal or graphical language.
Our main interest is in the last, in symbolic description.
“considering the extremely complex codings used by the brain. 8/2. Ubiquity of coding… when a “Gale warning” is broadcast. It starts as some patterned process in the nerve cells of the meteorologist…”
A brain can encode some information in the structure of its memory, which becomes useful when the brain decodes it, by reversing the encoding process. Similarly, a sender can encode some information in a message, which becomes useful when a receiver decodes the message using the same code. The information/meaning in a memory/message exists in the processes of encoding and decoding it.
To create information is to write or encode a memory or message that represents something. To use information is to read or decode a memory or message, and use it for some purpose. Nobody understands how our biochemistry encodes thoughts, but those processes exist.
Communication stacks in a communication network
Ashby’s cybernetics is about how information is stored and communicated. In the processes of creating and using information, rather than its physical form. These processes encode and decode information into and out of physical forms.
Suppose you want me to look at the moon, and then ask me to.
· Your conscious thought is encoded in
o neural impulses from your brain to your mouth, then
§ vocal chord movements, then
· sound waves, which are translated into
§ ear drum movements, then
o neural impulses from my ears to my brain, and
· conscious thought in my mind.
During the communication, your request is expressed or encoded in several forms, public and private. Note that the information/data distinction is recursive in a communication stack. The actors at level N+1 abstract the logical information of interest to them from the physical data structure at level N.
Ashby presented the example below. To send and receive the gale warning message involves a succession of coding and decoding steps. Each time a message passes between humans, it passes down and up a communication stack. It passes down from the sender’s brain to some physical medium for transport. And back up from that physical medium to the receiver’s brain, where the message is stored in memory, at least for a while.
“Let us consider, in some detail, the comparatively simple sequence of events that occurs when a “Gale warning” is broad-cast. It starts as
· a patterned process in the nerve cells of the meteorologist, and then becomes
· a pattern of muscle-movements as she writes or types it, thereby making it
· a pattern of ink marks on paper. From here it becomes
· a pattern of light and dark on the announcer’s retina, then
· a pattern of retinal excitation, then
· a pattern of nerve impulses in the optic nerve, and so on through her nervous system. It emerges, while she is reading the warning, as
· a pattern of lip and tongue movements, and then travels as
· a pattern of waves in the air. Reaching the microphone it becomes
· a pattern of variations of electrical potential, and then goes through further changes as it is amplified, modulated, and broadcast. Now it is
· a pattern of waves in the ether, and next
· a pattern in the receiving set. Back again to
· a pattern of waves in the air, it then becomes
· a pattern of vibrations traversing the listener’s ear-drums, ossicles, cochlea, and then becomes
· a pattern of nerve-impulses moving up the auditory nerve.
… this very brief account mentions no less than sixteen major transformations
through all of which something [announcer’s the intention] has been preserved,
though the superficial appearances have changed almost out of recognition.” (1956, 8/2)
Note that in any communication stack, the information/data distinction is recursive. The actors at level N+1 abstract the logical information of interest to them from the physical data structure at level N.
After receiving a message, a listener can verify the accuracy of the warning by watching the weather.
If the correspondence between a structure in human memory and the phenomenon it represents is to be useful, the brain must decode the memory by reversing the encoding process. Similarly, if the correspondence between a structure in a message and the phenomenon it represents is to be used as intended, a receiver must decode the message using the same code the sender used to encode it. The meaning of a message only exists in the process of encoding it or decoding it.
Representing a structure
Suppose A = a map and B = a territory. And
there are correspondences between the structures of the two entities. Then the
map carries some information about the territory. And the territory carries
some information about the map.
Representing a behavior
Suppose A = a musical score and B = a musical performance. And there are correspondences between the structure of A and the behavior of B. Then the musical score carries information about the process of the performance. And the performance carries information about the structure of the musical score.
Remembering a thing
· A = the state of something in the environment. M = the state of a message conveyed by eyesight to your brain#
· B = the state of a memory in your brain.
To register and remember the existence of the thing
· A is encoded into M
· M is decoded and encoded into B.
Ultimately, the meaning of a memory is not found in the memory alone. It is found in the process by which B is decoded by retrieval and used. This last is the information of most interest to psychology.
Communicating an idea
Suppose A = the state of a message sender’s brain. M = the state of the message. B = the state of a message receiver’s brain.
To communicate an idea
· A is encoded in M
· M is decoded and encoded into B
The aim of human-to-human communication is not to draw a biological correspondence between A and B. How far the structures of two brains can be correlated at the biological level is unclear. By contrast, correspondence at the sociological level can be observed and verified.
Recognize the looseness of natural languages
Synonyms and ambiguities abound in natural language. Consider the terms we use to describe things. We may speak of a thing as having properties, qualities, characteristics, attributes or features. Some use those terms interchangeably; some draw distinctions between them. There is no general agreement as to whether the terms should be distinguished or how.
Recognize that one type can be instantiated in many times and places
Our type theory begins with four assertions:
· to describe a thing is to typify it
· to typify a thing is to describe it
· one type can be instantiated in many things
· one thing can instantiate (embody, realize, manifest, conform to) many types.
We use the same terms (property, quality, characteristic, attribute and feature) with reference to both:
· conceptual types (say, length: a measurable distance)
· instantiations of conceptual types (say, the 100 metre length of a race).
Speakers assume that listeners can tell from the context which meaning applies; but it is not always clear.
Recognize that descriptions can be complex types
A type is conceptual or intensional definitions of a thing that instantiates the type. There are simple types like length, and complex types such as:
· the conceptual type that is a symphony score
· an instantiation of that conceptual type in a symphony performance.
Recognize that one phenomenon can be described in several ways
We use a map (mental or documented) to help us understand a territory. The map relates selected features of a territory to each other. The triangle below relates mappers to maps and territories. Read it left to right thus: Mappers <create and use> Maps <represent> Territory.
<create and use> <represent>
Mappers <observe and envisage> Territories
Different maps, showing different features, may be drawn of one territory. You may use different maps for country walks, planning journeys, driving, studying geographic features, studying demographic features. And you may need to find or buy a new map now and then. And more generally, any phenomenon can be described in several ways. Say, light can be described in terms of waves or particles
Recognize that holism is not wholeism
A map is never a complete or perfect representation of a territory. It only needs to tell us enough that we can find or learn what we want.
Taking a holistic view of a business does not mean you study the whole of the business, every conceivable element of it. It can mean you observe or envisage elements relevant to your motivation for understanding or describing the business. Then, study or describe how those elements interact to produce outcomes they cannot produce on their own. Or else, you identify some unexplained or desired outcome of the business in operation. Then, you find or describe which elements of the business do or must interact to produce that outcome.
Recognize that description can take many physical forms
You may observe or envisage some properties instantiated some phenomenon, now or in the future. And encode those properties in a description or model future use. The description may be held in biochemical, mental, documented or other physical form.
This one three chapters that outline compatible theories of information, description and types. In short, the ten principles of description introduced are:
· A good regulator has a description of what it regulates
· Consciousness is a process that enables us to compare possible futures with past experience.
· Every type is a description
· Every description is a type
· To describe a complex thing using symbols is to typify it in terms of types already understood
· A description is meaningful to an actor only in the process of creating or using it
· Descriptions can be iconic and/or symbolic
· Coding is ubiquitous in the creation, use and sharing of symbolic descriptions
· We share knowledge by verifying descriptions we share
The final chapters pursue some implications of these principles.