[ox-en] Patterns are not simply abstractions
- From: "Franz Nahrada" <f.nahrada reflex.at>
- Date: Sat, 27 Feb 2010 09:18:35 +0100
Hi Stefan,
thanks for publishing my still torso-like presentation and your comments.
Yes, the design patterns in software are an offspring of what Christopher
Alexander formulated for architecture. Pattern Theory is gaining ground in
learning and communication thoory as well. I think its a very fruitful
paradigm that enables us to see beyond traditional methodological
constraints.
An interesting observation is that Alexander gave a speech in 1996 talking
to the software developers community at the 1996 ACM Conference on
Object-Oriented Programs, Systems, Languages and Applications (OOPSLA) in
San Jose, a place I personally remember well because I was at the Apple
WWDC there six years earlier. Alexander dared, as a non-specialist and
non-programmer, to adress the community with some inherent criticism.
http://www.patternlanguage.com/archive/ieee/ieeetext.htm
Christopher Alexander: "When I look at the object-oriented work on
patterns that I've seen, I see the format of a pattern (context, problem,
solution, and so forth). It is a nice and useful format. It allows you to
write down good ideas about software design in a way that can be
discussed, shared, modified, and so forth. So, it is a really useful
vehicle of communication. And, I think that insofar as patterns have
become useful tools in the design of software, it helps the task of
programming in that way. It is a nice, neat format and that is fine.
However, that is not all that pattern languages are supposed to do. The
pattern language that we began creating in the 1970s had other essential
features. First, it has a moral component. Second, it has the aim of
creating coherence, morphological coherence in the things which are made
with it. And third, it is generative: it allows people to create
coherence, morally sound objects, and encourages and enables this process
because of its emphasis on the coherence of the created whole."
you write:
However, from this definition you can't define patterns in advance but
can only observe them and define them as some sort of best practices.
More theoretically speaking I think (design) patterns are just another
name for a certain type of abstractions. In fact some people may seem
to recognize the concept of abstractions by learning about (design)
patterns. I'd see this less as a strength of (design) patterns and
more as a lack of scientific education, though...
In fact it's difficult to find abstractions / patterns and you really
need to be an expert on the subject with lots of experience. In
particular you need to distinguish carefully the *content* of a
pattern with the pattern itself.
I think you come from the concept of design patterns that Alexander
adressed as a half-way approach.
In my view it is very necessary to state to state that patterns are not
just theoretical abstractions, but in fact they are at the very core of
"what holds reality together". Patterns are indicatable, recognizeable,
recurring objective structures that hold and reenforce activities.
Patterns are at the bottom of the phenomenon of life or vitality. They are
the outcome of evolution, like the species of a well evolved ecosystem.
I dont think that you can truly separate the patterns from their content,
the content being processes and dynamics that usually flow and use many
patterns. A pattern is a structure that acts as a problem-solving or
energy-enhancing dynamic "shell", mostly embedded in a particular sequence
of other patterns that play out well together. I tried to show in addition
that despite their abstarct character most patterns have historical
nature, that not everything exists equally at any time but rather in
pattern clusters, - as we see in evolution that interlinked species thrive
with each other, but at the same time through time many things are
retrieved again and again when their time has come. This is the McLuhanite
turn i wish to support. Like I try to show with "my" "village" pattern
that takes different content in medieval time and -maybe - in
postindustrial time, but still has a defineable array of properties over
time:
* a limited number of human beings that can perceive each other as a
community rather than a society
* a form of life and living space / settlement that interplays with nature
and landscape as the "significant other"
* a need and culture to find informal social solutions rather than formal
* a local economy
so it is a subdued or underused, neglected, deprived pattern in the
industrial age.
What brings us to the perception and predicteablity of patterns is not
just speculation, but the timeless experience of ups and downs of
patterns. Thats where McLuhan in my view complements Alexander perfectly.
It allows us to marry the merits of Alexandrine pattern thinking with the
structural depth of historical analysis of Marxists like Eric Hobsbawm.
They cannot stand beside each other as isolated worldviews, they need to
be integrated.
when I wrote
Take for
example the permaculture movement. Permaculture is revolutionizing our
relation with the biological environment, because nature is not any
more considered an object of human labor and control, but rather as a
self-organizing system in a way very similar to design patterns: doing
almost all of the work and maintaining itself, if arranged properly.
and you answered
But "arranging things properly" *is* perfect control - isn't it? So
I'd not say that this is a new pattern in this respect. It's a new way
of doing agriculture, yes, but the agriculture pattern is kept.
I agree with you that this is in a way a continuation of agriculture, but
patternwhise "agriculture" seems to me too much expanded and to trivial to
really have pattern quality.
(by the way as I said permaculture is not a romantic way of looking at
nature, but a deep understanding of checks and balances in a complex
system).
Also "love" is for me not a pattern. You cannot describe in a non-trivial
or tautological way what is love. "Marriage" would be a pattern, or
"Romantic Love" - very clearly describeable and recurring patterns through
history. Patterns always give way to alternatives, choices. They refer to
a plurality of phenomena and the way we can analyze them, choose between
them.
I hope I made myself clear.
Franz
-----------
Hi Franz and all!
Some comments.
7 minutes ago Stefan Merten wrote:
Alexanders approach, though ontologically similar, is radically
different in its perspective. Rather than focussing on the dynamics of
constant change, he rather asks what is recurring and what is stable.
He reserves the term pattern to structures that prove as timeless,
because they have an inherent quality of enhancing and supporting a
wide range of activities.
In fact that is a feature of (design) patterns which I know also from
software architecture where design patterns today play a major role.
However, from this definition you can't define patterns in advance but
can only observe them and define them as some sort of best practices.
More theoretically speaking I think (design) patterns are just another
name for a certain type of abstractions. In fact some people may seem
to recognize the concept of abstractions by learning about (design)
patterns. I'd see this less as a strength of (design) patterns and
more as a lack of scientific education, though...
In fact it's difficult to find abstractions / patterns and you really
need to be an expert on the subject with lots of experience. In
particular you need to distinguish carefully the *content* of a
pattern with the pattern itself.
The idea
that you love a physical person in a highly exclusive way - like you
love God - was brought up by the medieval renaissance.
I'd agree that this is a major shift. But isn't this better described
as a paradigm shift than as a new pattern? From the pattern "love" I'd
say this is not a new pattern because the pattern is the same and only
the content is changed.
Take for
example the permaculture movement. Permaculture is revolutionizing our
relation with the biological environment, because nature is not any
more considered an object of human labor and control, but rather as a
self-organizing system in a way very similar to design patterns: doing
almost all of the work and maintaining itself, if arranged properly.
But "arranging things properly" *is* perfect control - isn't it? So
I'd not say that this is a new pattern in this respect. It's a new way
of doing agriculture, yes, but the agriculture pattern is kept.
It is probably no incident that this special view of nature has
emerged simultaneously with the use of cybernetics and modern
information theory.
Sure this is not by chance. The romantic view of nature is the
counter-movement to more and more technology. However, it is in many
ways a backward oriented movement dreaming of a view of nature which
never was reality - not even before the romantic era which all
invented this paradigm of seeing nature.
In a way and even with a certain degree of justification, the market
system could be considered as a first attempt to enact such autonomous
behaviour of subsystems
Yes. Markets are perfect examples of autonomous agents interacting
which each other without central control.
The information that is
conveyed by money is almost zero and necessarily overshadowed by
inherent bias.
True. One reason is the abstraction money represents. Money abstracts
From nearly everything - but this makes things comparable and this is
a precondition for complex markets.
So some might still work on a pattern of informed and informing money,
but in fact the Internet by its very existence indicates the ability
of much more precise information transfer between autonomous
subsystems.
If you communicate measures of things then it is really misleading to
call this money.
Grüße
Stefan
-----
Hi list!
Here comes Franz' presentation.
Grüße
Stefan
=== 8< === 8< === 8< === 8< === 8< === 8< === 8< === 8< === 8< === 8< ===
================================================
A Pattern Language Of The Postindustrial Society
================================================
--------------------------------------
Roadmap for Synergy in Times of Change
--------------------------------------
[http://fourth.oekonux-conference.org/program/index.html Speech given
at the 4th Oekonux Conference in Manchester]
Abstract
========
The term pattern language was coined by Christopher Alexander and it
is quite suitable for the attempt to describe a society in which
peer-to-peer principles are dominating pillars of the social fabric.
Rather than to be derived from a "principle" the p2p-society is built
on the mutual reenforcement of patterns, structures that only in their
combination and interaction allow to solve the complex challenges of
life.
The speech builds on the contribution to the "Open Source Yearbook
2008" but will increase the number of patterns, describe them closer,
outline their mutual relations.
A Pattern Language Of The Postindustrial Society
================================================
Introduction: Reality is real because it is mediating
-----------------------------------------------------
Societal change is and has never been the result of one single factor.
In the case of capitalism, many convincing points have been made to
explain different sources of the emergence of a different mode of
production, of social dominance, of human relations.
Take for example the notion of Sombart about the appearance of the
Liber abbaci by Leonardo Pisano around 1200. The second section of
Liber abbaci contains a large collection of problems aimed at
merchants. They relate to the price of goods, how to calculate profit
on transactions, how to convert between the various currencies in use
in Mediterranean countries. On the other side the same author points
in another book to the development of modern warfare, a thesis which
is greatly supported by the central role the costs of maintaining
armies for the incentive for feudal lords to switch to capitalist
methods of production. Again there are theories that point to cultural
and religious structures that are greatly enhancing the dissolution of
the hierarchical society, for example Weber's inquiries on the
protestant ethics. Others point to the emergence of colonialism as the
main source of wealth accumulation. Others point to the host of
inventions and the emergence of modern science that made it possible
to revolutionize production and create the one necessary step "out of
nature" that was the necessary precondition for a society in which
abstract value is dominating and prevailing.
All this is to show that there are *structures that are enhancing
certain activities while blocking others*. A societal change comes
about when several of these structures develop an intricate synergy.
They resonate with each other, find each other, give life and energy
to each other, they even grow together and merge into a megastructure.
And, most interestingly, they come from different sources.
This will enable us to overcome dogmatism (like "Free software is THE
Germ Form of a new society") and become more aware of our constructive
possibilities.
For me there are two thinkers that, from a totally different angle,
have tried to formulate this fundamental view that enables us to not
only understand how historical changes come about, but why certain
structures do manage to stay alive and others don't. They help us
understand that social change does not simply come from one point, but
is a result of positive, multipolar, self-feeding interactions.
The two thinkers are Marshall Mc Luhan with his theory of media and
Christopher Alexander with his theory of patterns. They both oppose a
worldview that separates dead matter from living ideas, like
fundamentally laid out in the work of Descartes. They rather give us a
key to understanding the world as an interplay of related processes,
which are not isolated from each other, but constantly interacting,
enhancing, supressing, expanding, reversing. McLuhan has tried to
capture this quality in the term of the medium: he has rightfully
pointed out that each and every relation that we have with the world
is through structures that mediate these relations, and that the
structures deeply shape the outcome of our perception and position in
this world. Thus it is justified to look at history of society as a
history of media, which not only allowed for certain perceptions and
impressions, roles and values to prevail, but which also interact in a
dialectical way with each other.
McLuhans dialectic is a tetrade rather than a triade; he has added one
important element to Hegelian thought that makes all the difference in
the world. There are four laws of media, and three of them are well
known from traditional dialectic worldviews. There is a powerful
structure emerging which on one side enhances certain qualities. On
the other side this emergence supresses other qualities. The growth
and universalisation of a structure leads into a point where there is
a paradox reversal, when pushed to its extreme limits it reverses into
something new by its own internal dynamics. This is where the Keimform
Theory has a place, because it states nothing can grow than by feeding
processes that carry the potential of entirely new structures to a
certain point undermining the nurturing structure. The fourth law of
media, however, is the most interesting, it is called the law of
retrieval, and McLuhan expressed it bluntly when he said: The content
of any new medium is an older medium.
Alexanders approach, though ontologically similar, is radically
different in its perspective. Rather than focussing on the dynamics of
constant change, he rather asks what is recurring and what is stable.
He reserves the term pattern to structures that prove as timeless,
because they have an inherent quality of enhancing and supporting a
wide range of activities. Alexanders interest as an architect is
simply to find out what makes the quality of an environment that
nurtures and channels human activites best. So he is also keen to look
at phenomena of permanent retrieval, timeless proven solutions that we
embrace again and again. His vigor is to understand the quality of
such solutions, analyze what differs a good solution from a bad
solution and create an understanding for the circumstances in which
such solutions work and support each other.
Maybe at this point its interesting to try to synthesize the two
views. We have seen that the emergence of capitalism brought about an
interplay of patterns, *intellectual patterns* like modern math or
experiment-based science or the university, *spiritual patterns* like
a worldview that puts individual responsibility at the center,
*relational patterns* like the territorial state replacing the
personal dependency state, *conflict patterns* like the standing army,
*economic patterns* like colonialism, urban trade centers, eventually
factories, *technical patterns* like gunpowder warfare, steam engine,
machine, and so on.
An Analogy to Life
------------------
You get the picture. The reality and especially the emergence of a
totally new reality is not adeaquatly understood if I say: the central
concept of this new society is, for example, the motivation by money
and the growth of money. Let's stick still for one last minute with
the example of the capitalist society. On one side this idea that the
monetary pattern is the central pattern or the central paradigm is
very true - however, one could say with equal right that there is
rather a pattern cloud emerging that facilitates certain behaviours.
There are even essential patterns to facilitate non-monetary behaviour
that are needed. So, for example we can see a hidden parallel
development of the core family and the emotional emphasis on love
together with the impersonalisation of economic relations. The idea
that you love a physical person in a highly exclusive way - like you
love God - was brought up by the medieval renaissance.
A few theorists were able to draw the lines that connect the logic of
the economic value sphere to the seemingly totally different personal
value sphere. Part of this is the polarisation of gender roles and
gender stereotypes. The male and the female appear as a social
creation, although they are not. There are several patterns that
facilitate some core functions of society like reproduction, care, and
that also serve as specific containers for problems just created in a
particular society like dealing with the astounding range of
inablities of highly specialised individuals. One such example of a
modern pattern would be "self-declared religion" where people rally
around affirmational institutions for subjective moral values in clubs,
associations or sects of any kind to simply escape the contingent and
arbitrary nature of their general socal relations.
So here we go. To understand the concept further, we can draw on the
analogy to living nature. In fact life is the paradigm for patterns.
Life is patterns of matter that are organized for self-reproduction.
If you play the game of life you might see certain patterns prevail
and out of chaos emerges order. if you play it further you see that
living organisms not only take very different shapes, but their life
would not take place without other living organisms. So in any given
place of life, or biotope, you have an active dependency of organisms
on each other. And this organisms tend to be very, very different.
Besides water and salt, every meat is transformed grass. Grass is
transformed light, carbon and earth. In more highly developed stages,
this goes beyond simply devouring and leads to inhabiting. One
structure feeds the other, facilitates the other. Patterns interact
and form pattern clusters. Each pattern is self-feeding, but it is
also feeding others that might feed back or not. Plants might profit
From predator animals because the limit the number of plant-eaters.
So the idea of patterns is to take this analogy and base our woldview
not just on the objectivity, but also very basically on the
interactivity of things. Whereas Alexander has reserved the word
"pattern" to structures that facilitate sustainable and recurring
interactivity, we could rightfully add that there is a relative
historical coherence of patterns and not just the absolute and
timeless coherence that Alexander thought of. In times of change,
however, we might wish to replace obsolete patterns by "more timeless
ones" by inventing what is missing and retrieving what was lost.
Patterns and the Central Social Paradigm
----------------------------------------
So the question and the question ahead of us is to conceive the
universe of patterns generated by and generating the emergence of a
new social formation. There is a major shift in the DNA, in the very
core structure of society. Another analogy to life could be very
helpful: a biotope in the desert is different from a biotope in the
rain forest. The rain forest is built on massive amounts of precursing
biomass that hold essential external conditions, even although the
climate zone might be the same as the desert. This is the analogy
which is maybe a little bit helpful to understand the Marxist concept
of productive forces. The productive forces are the humus and the
shade on which patterns can grow. The biotope itself, under certain
conditions, might take the form of a protective meta-organism,
providing shade and shelter and even more. Some biologist - like
Thomas Boller from Basel - argue if the strains of Mycorhiza fungae
found in a forest soil are not only the sole nourishers of a wide
range of plants (something nowadays accepted in biology), but also
serve as a kind of communication and protection system by which older
trees nourish younger trees or warn against bugs and .
So out of this biological example we could also draw the analogy that
at a certain point of resourcefulness it makes much more sense for a
society to switch to cooperative behaviour rather than competitive
one.
It seems we are at a turning point or tipping point in human history
where this structural change happens, although it was long-anticipated
and long-awaited. But we need to look at the biology of this new
biotope, at the species emerging that form this new resourcefulness,
that maintain it, replenish it, enhance it.
Maybe we could start with the Internet as the mycorhiza of our new
biotope, allowing us to grow on intentional communication. The
internet has certain essential qualities which are
* bidirectionality of communication,
* the ability to "narrowcast" and "broadcast" on demand,
* the ability to transform all kinds of media into and off a digital
format and therefore drive all kinds of perceptive, responsive and
productive activities by man and machine, and, maybe most important,
* ubiquity and
* synchronicity.
Each one of these qualities requires and produces a set of innovations
and retrievals that only in their interplay could form a P2P society.
Currently, these innovations and retrievals are dealt with in
subcultures, who usually wrongfully take the part as the whole. One
could say that the Free Software Culture in itself, although it is
already complex in itself, is a set of organisms that are ambiguously
able to survive in the old biotope as well as participate within and
help generate other elements of a new one. But there are other
organisms emerging on the way that play a similar role. Take for
example the permaculture movement. Permaculture is revolutionizing our
relation with the biological environment, because nature is not any
more considered an object of human labor and control, but rather as a
self-organizing system in a way very similar to design patterns: doing
almost all of the work and maintaining itself, if arranged properly.
It is probably no incident that this special view of nature has
emerged simultaneously with the use of cybernetics and modern
information theory. In a way the permaculture view on nature is
similar to a programmers view who creates interacting objects that are
capable of fulfilling autonomous tasks in interaction with each other.
The simple notion is that nature is by far to complex for humans to
manage the system themselves, so we look for a system of checks and
balances that works, and we find thousands of natural systems and
organisms that already have a highly interactive and synergetic
potential, they are fine-tuned to permanently react on time and
produce what's needed to keep the system alive. Cybernetics by the way
was born because in air defence human reaction is too slow, we need
information based agent systems that react autonomously. Cynically
enough, one could argue that both Cybernetics and the Internet are a
product of warfare innovation, but indeed as we have seen at the
beginning, this is true for most social and technical innovations that
nowadays shape our culture.
Maybe the central social paradigm that unifies Permaculture and Free
Software Movement, but as we might see dozens of other emerging
organisms or patterns could be characterized as: if you give up the
central control over the process you might be able to achieve better
results than if you try to coordinate. But there is one essential
requirement for this to work: each subsystem must be able to retrieve,
process and communicate information from and to other subsystems. In
the case of nature, most of this information is stored in the DNA, and
it is most likely the product of blind and massive trial and error
processes in the billion year history of evolution - speeded up by
selection and mutation mechanisms that we are still hardly beginning
to understand. But as you obviously know nature has autonomously
developed an incredible wealth of signalling and information
transferring systems.
In a way and even with a certain degree of justification, the market
system could be considered as a first attempt to enact such autonomous
behaviour of subsystems (as we have heard from Michel Bauwens, giving
up direct control and switching to indirect control was the driving
factor in raising efficiency even from Slave System to Feudal serfdom,
and of cause from Feudalist Serfdom to Capitalism) and many consider
the monetary system simply as a channel of information. If we for a
moment consider this position instead of simply rejecting it, than we
discover that the market system is probably a very primitive and
error-prone system of information transfer. The information that is
conveyed by money is almost zero and necessarily overshadowed by
inherent bias. The irony is that even the most zealous apologists of
the market system have to admit that the current crisis is based in
the fact that some derivative forms of money purposefully deleted all
kind of information about their validity. But in fact this exformation
property of money is built in the most elementary function of money.
Marx showed that money does not and cannot inform us about the
proportion of human labor needed for production, because the
subsystems accumulating money can only do this by constantly acting
out of proportion. So a cheap pricetag on a commodity can mean many
things, from the externalisation of costs to the use of child labor,
From the lack of quality and durability of the product to a temporary
overproduction.
So some might still work on a pattern of informed and informing money,
but in fact the Internet by its very existence indicates the ability
of much more precise information transfer between autonomous
subsystems.
The central social paradigm of the emerging P2P society could be
described as the information based recognition and useage of positive
and productive output of subsystems. The built in assumption is, that
with the help of information based relations in the end the other
subsystems will be able to spontaneously produce enough output to
sustain and complement each other better than by market relations.
Out of these few assumptions I propose we build a pattern language of
a post-industrial society.
Towards A Pattern language
--------------------------
I had the intention of developing a rough sketchy version right before
this conference, but out of external reasons there was a lack of time
to do it. So I can just suggest some patterns and pattern clusters to
work with, and ask for contributions and common development.
I should also make clear that there is already an astounding effort
going on to build what is called a "pattern language for the
communication revolution". That community or subculture around Doug
Schuler has made a tremendous wealth of pattern proposals available
online and recently published a 136 of these patterns in a book called
"Liberating Voices". The reason I am not completely subscribing to
this pattern language is that many of these patterns are intermediate
and not self-standing patterns. What I mean by that is that a number
of patterns are rooted in the protest and civil rights movement and
they are focussed on the current emergence of the Civil Society.
We need to take this pattern into account, but extend the productivity
and self-reliance beyond the proven patterns to reach a true
coherence. Therefore I suggest to do a pattern language of a highly
speculative nature and include not only what's out there already and
proven, but also what needs to be created.
We might follow Alexander's example and organize the patterns in an
order where the most general pattens come first and the smaller and
fractal patterns follow.
1. Patterngroup: Global patterns
1. Planetary management
http://www.pege.org
2. Commons
maybe needs to be specified
3. Fractality and Subsidiarity
(crosslink to Alexanders Independent Regions, Roberto Verzolas
"avoid global variables")
4. Civil Society
5. The Internet
2. Patterngroup: Cultural Patterns
1. The voluntary, self-driven culture
Frithjof Bergmann: Its not an easy thing to find out what we
really, really want.
Andrius Kulikauskas: helping each other to grow by holding each
other accountable to our own personal values.
2. Self-directed cultural communities
Example: Linux, Wikipedia. But that just the tip of an
incredible Iceberg.
focussing on diversity and even contradictinng schemes
3. P2P Education
Examples VideoBridges, but also immersive training like OSE
4. Knowledge culture
Monasteries of the 21st cebtury, deep science community
5. Hypercycling and social hacking
solidary economy, social entrepreneutsip, cooperatives.
3. Technical patterns
1. Living Systems
study and learn from nature. Permaculture, Regenerative Design
2. Design Language
3. Flexible Fabrication
4. Local Patterns
1. Global Villages
2. Arcology
3. Learning Center
4. Life Maintainance Organisation
_________________________________
Web-Site: http://www.oekonux.org/
Organization: http://www.oekonux.de/projekt/
Contact: projekt oekonux.de