The psychology
of description
This page is published under the terms of the licence
summarized in the footnote.
All
free-to-read materials on the http://avancier.website
are paid for out of income from Avancier’s training courses and methods
licences.
If you find
them helpful, please spread the word and link to the site in whichever social
media you use.
The psychologist W Ross Ashby was concerned with autonomic functions that maintain the state variables of the body.
He eschewed discussion of higher intelligence, but we turn to that below.
This paper discusses how we think about things, model them and communicate our mental models to others.
Much of the content has been edited from an email correspondence with Marc Lankhorst (which is not to say Marc agrees with everything said in all the papers).
Contents
Descriptions
that are mental models
Descriptions
that are communicable
Concepts in words and documents
Comparing
mental and documented models
Concepts
in system/architecture description
Architecture
description as a collective memory
Descriptions in memory
How
can an organism know what is out there in the world? react
to it? predict it? anticipate
it? direct it?
Psychologists assume a major function of our brains is to typify or model entities and events in our environment.
Having reasonably accurate mental models helps actors to recognise things in reality and predict their behaviour.
Experiments suggest that a wide range of animals can use their minds to accurately predict or direct future events.
The evidence suggests that brains hold mental models of the world – of things and how they behave
E.g. a honey bee brain creates and uses mental models of pollen sources.
Some philosophers (idealists) share the view of psychologists that reality is constructed in the mind.
That doesn’t mean there is no world out there; only that brains hold mental models of the external world.
And it doesn’t mean mental models are wholly accurate; only accurate enough to help organisms survive.
Descriptions in communication
How can an organism describe something (a food source, or a threat) to another member of the same species?
It must translate mental models into communicable models.
E.g. a honey bee can translate its mental model of a pollen source into the form of a dance.
A human being usually translates a mental model into the form of words, spoken or documented.
So, actors can create mental models, dances, speeches and documents to describe realities.
We humans describe things in our minds, in verbal and visual communications, and in documentation of various kinds.
Descriptions are models (be they mental, documented or other) that idealise realities.
We not only remember past realities and observe present realities, but also envisage future and fictional realities.
Generally, a description is anything (created and used by
an actor) that encodes concepts or properties abstracted from other things.
Our psycho-biologist position is that descriptions evolved from the ability of actors to form mental models of things observed and envisaged.
Concepts encoded in sense memories
Animals
can create (in some coded form) a sense memory of observing the structure and
behaviour of a thing.
For example, remember the look and feel of the concept we humans call “thorniness”.
We humans naturally remember words as sense memories; e.g. remember the sound of the word “thorny” and the sensation of saying it.
And after some education, we can associate that sense memory with the look of the written word “thorny”.
So, we can associate the look and feel of thorniness with the sound and look of the word “thorny” in a mental model of the concept.
Code to code translations
A mental model is one form in which a concept is coded – a first kind of description.
You can translate that by voicing the sound of the spoken word “thorny”– a second kind of description.
Or translate that word be writing into the visual image of written word – a third kind of description.
Or paint a picture of a thorn – a fourth kind of description.
Or build a clay model of a thorn - a fifth kind of description.
All five descriptions can be recorded; and (ruling out telepathy) all but the first can be used to communicate.
As
a human, you can encode what you observe and envisage in many forms of description.
And translate descriptions between coded forms. Here are some examples:
· Sensations of sight, sound, and touch can be translated into the biological code of mental models.
· A mental model can be translated into the codes of sound (speech) or sight (writing, mime).
· A sounded word can be translated into the codes of sight (sign language) or touch (braille).
Educated people continually swap back and forth between mental and documented models
We decode the one and use it inform or change the other.
The importance of words to humans
Self-aware beings can bring mental models to mind at will, inspect them and manipulate them.
Words are an especially important way of encoding descriptions to humans, not least because we can manipulate them so easily.
I believe we can not only speak using words, but also think using words (as I do in writing this sentence).
Encoding of concepts in verbal descriptions is especially important to humans.
We have developed verbal descriptions to a fine art; we
remember numbers and names as words, and describe things using words.
We not only learned lots of individual words, we associated one with another,
We associate named concepts (like “planet”) with other
named concepts (“massive”) and named things (Mars and Venus).
We can manipulate words and invent new ones.
Surely, this ability is what enables human actors to think so much more widely and deeply than other animals?
And for obvious reasons, we deal here mostly with concepts and properties encoded in written words.
How
can an organism know what is out there? predict what
will happen in the world? react to it? direct it?
Psychologists assume a major function of our brains is to typify or model entities and events in our environment.
Having reasonably accurate mental models helps actors to recognise things in reality and predict their behaviour.
Experiments suggest that a wide range of animals can use their minds to accurately predict or direct future events.
The evidence suggests that brains hold mental models of the world – of things and how they behave
E.g. a honey bee brain creates and uses mental models of pollen sources.
|
Pollen
source location |
|
Honey
bee mental models |
|
<create and use> <abstract concepts from> |
|
Honey bees <observe and envisage > Pollen sources |
The idealist psycho-biologists may insist that reality is constructed in the mind.
That doesn’t mean there is no world out there; only that brains hold mental models of the external world.
And it doesn’t mean mental models are wholly accurate; only accurate enough to help organisms survive.
Some can encode concepts not only internally in mental models, but also externally in symbols observable by others.
How can an organism describe something (a food source, or a threat) to another member of the same species?
It must translate mental models into communicable models.
Honeybees
are clever little creatures.
They
can form abstract concepts, such as symmetry versus asymmetry.
And
they use symbolic language — the celebrated waggle dance — to direct their hivemates to flower patches.
New
reports suggest that they can also communicate across species, and can count —
up to a point. http://www.livescience.com/2909-bees-count.html
.
|
Pollen
source communication |
|
Honey
bee dances |
|
<create and use> <abstract concepts from> |
|
Honey bees <observe and envisage > Pollen sources |
The honey bee’s dance is a kind of map that models the direction and distance of a subject that is a pollen source.
The survival of the bee hive depends on the accuracy of this dance-map.
Other bees read the dance-map, test it, and, if they find the pollen, then the map is accurate enough.
Our everyday experience is of translating concepts/properties (already coded in our minds) into a different code for communication to others.
Human actors create mental models to record descriptions and create mimes, speeches, documents (etc.) to communicate and sometimes record descriptions
Obviously, words are an especially important form of encoded description to humans.
|
Human
verbal communication |
|
Words |
|
<create and use> <abstract concepts from> |
|
Humans <observe and envisage > Realities |
You not only speak using words, you can also remember and think using words (as I do in writing this).
Surely, this ability is what enables humans to think so much more widely and deeply than other animals.
We retrieve concepts encoded in mental and documented models – we decode the one and use it inform or change the other.
Cognitive psychologists assume our brains hold mental models that typify real-world things and how they behave.
These mental models of observed and envisaged things form a
network of inter-related types in the brain.
Inevitably, given our imperfect and continuous learning process,
our network of types is flexible, fuzzy and contradictory in places.
Is meaning in the model?
No. Meaning is only understood at model coding time, and model decoding time.
An
encoded concept has no meaning until an actor decodes and uses it.
A model is a model, regardless of its form, be it in the mind, on paper, a dance, a stone carving, whatever.
A model becomes meaningful when it is used – if only to create a new memory.
How are models tested?
The job of system (enterprise or software) architect is to envisage and describe a target system - a future reality.
The target system is a fictional reality when first described. How do we judge it be a good and true description? Again:
We can ask judge or jury (peer group, or other stakeholder group) to examine a thing or a description, and give us a verdict.
We can devise test cases with predicted results, and compare those predicted results against the actual results of running the tests.
For either, architects usually have translate their mental models into more communicable documentation.
Both humans and computers are able to monitor and direct entities and events in the world about them.
Computers can detect events in the world, store a model of external entities, and send messages to direct the behaviour of those entities..
But computers lack the flexibility needed to deal with novel situations, and the self-awareness to inspect and change their own models and rules.
Our brains enable
us to make sense of, interpret, represent or model the world, and to make
predictions about that world.
Self-awareness enables us to inspect and change our own mental models.
You do somehow retain a persistent memory of sensations you have experienced and concepts you have conceived in the past.
You can probably remember your primary school buildings (though there is surely some memory decay and shift over the years).
You can bring that memory back into the spotlight of conscious attention, more or less accurately.
Yet you knowingly think about those school buildings only occasionally and transiently.
Your stream of consciousness is directed hither and thither by external perceptions and internal motivations.
By will or by accident, past memories/thoughts are brought into the spotlight of attention.
Once in the spotlight, you can communicate or document a memory, and create new memories in doing so.
Following the real-time stream of your consciousness, you can express concepts in spoken words.
Instead or as well, you can express concepts in writing, in persistent documented models.
Humans have two
ways to boost the advantages that thinking gives us.
We have language: we can think using words and
use words to share thoughts with others.
Talking and listening helps to firm up and extend the mental
models of speakers and listeners.
We have documentation: we can store and read
thoughts not only in mind but also in documents.
Writing and reading a documented model helps to firm up and extend the mental models of its writers and readers.
The spotlight of our attention can shine on a variety of sensations, experiences and memories.
We may bring to mind sights, smells, sounds, music and words.
We express concepts using body language, words, and graphical substitutes for words.
Our thoughts and words
are bound to each other.
We don’t form
sentences and draw diagrams just to record completed mental models, or show
them to others.
We do it to help us
form, test, stimulate, remember and extend our mental models.
Where is the concept of a thing found?
A concept of a thing may be found in a mental model and/or a documented model.
We read and write mental models and documented models in very different ways.
We all know how we create and maintain documented models.
Nobody knows how we create or maintain a mental model.
Yet documented models are inextricably linked with mental models.
You may assume mental models precede documented models.
But often, we correct or clarify a mental model by documenting it and reading it back.
And reading a documented model can lead us to form new mental models.
A mental model is relatively fuzzy and fragile.
It may be changed by the process of thinking about it, or entirely manufactured as in "recovered memory syndrome".
"When it comes to memories, you'll do well to remember that they are very unreliable." http://www.psychologytoday.com/blog/look-it-way/201001/recovered-memory-syndrome
Moreover, one person can have several separate and contradictory mental models of one named thing in the world.
A documented model is not a second class citizen - merely a "denotation" of a mental model.
There is a two way process; each kind of model informs the other.
A documented model is a cognitive product that may capture one or more mental models.
A musical score captures the concept of a symphony, arguably more definitively than in any mental model of its composer.
That same musical score may stimulate new mental models in the minds of those who read it, or listen to a performance.
There can be any number of perfect copies of a documented model.
There can be any number of imperfect copies of a mental model, in different minds or in one mind.
Provided at least one model (mental or documented) exists, then it might be used to create the modelled thing
If all copies are destroyed, then it is impossible to build the modelled thing, except by a fluke.
All descriptions of operational systems are partial and flawed abstractions.
Which one(s) you call an "architecture description" depends on the rules of your chosen standard or method.
An architecture description is composed of stakeholder-agreed documented models of the system.
The documented models can be seen as collating various fuzzy and partial mental models of the system.
At the start of a building project, you (an architect) might form a mental model before drawing a documented model of it.
But perhaps you will instead select a documented model from a library of a pre-documented design patterns?
Thereafter, you proceed by round trip engineering - cycling between mental and documented models.
Your first-published documented model leads stakeholders to form their own mental models, which you must respond to.
At the end of the day, what is in your mind about the complete architecture?
· You are not aware of some elements – designed by somebody else
· You have forgotten some other elements that you worked on
· You remember variations and versions of the same element - not always sure which have been superseded and which is the latest
· Some of your mental models may be inaccurate, or perhaps entirely false memories.
The concept(s) of the final "architecture" cannot be found in one mind, and the sum of what is in many minds is a mess.
Various mental models have been collated and coordinated in documented models, which have grown incrementally to form a coherent whole.
To stabilise mental models and make them shareable, they must be externalised in documented models.
And if we want the whole architecture to be locatable in one place – we need to agree a set of documented models.
In short, once our building has one agreed architecture description, then the architecture concept(s) can be found in two forms.
· fuzzily and inconsistently in many fragile mental models.
· more clearly and coherently in the architecture description’s documented models.
No stakeholder can hold all the concepts and properties of a complex system description in mind.
Architects create views to address stakeholders’ different mental models, in so far as they are knowable and reconcilable with others,
Architects incrementally respond to concerns, requirements and feedback on what architects documented earlier.
Architects add items to the description one by one, creating a large and complex structure of concepts that no one of them can conceive.
Architects forget or misremember what was so carefully thought through and worked out earlier - by documenting models.
Architects have to re-read the description to bring to mind what it says - before they extend or change it.
Whatever partial and flawed recollection one architect has of the whole architecture, nobody else shares it.
An architecture description is a collective memory of what matters about a system of interest.
If there is such a thing as a collective consciousness, then the architecture description is that thing.
Many mental models of a system may exist, but none encompass all the system’s properties or are definitive of the system.
The purpose of documenting an architecture description is to hold an “intensional definition” of a system, agreed as meeting the concerns of stakeholders.
The purpose is not so much to denote or represent a system concept, as to be the system concept that has been agreed.
Footnote: Creative Commons Attribution-No Derivative Works Licence
2.0 06/01/2016 19:27
Attribution: You may copy, distribute and display this copyrighted work
only if you clearly credit “Avancier Limited: http://avancier.co.uk” before the start and include this footnote at the end.
No Derivative Works: You may copy, distribute, display
only complete and verbatim copies of this page, not derivative works based upon it.
For more information about the licence, see http://creativecommons.org