Organizational Architecture
Markus Schatten
Here some introduction to modern organization theory since I think this is a
vital part of our disscusion. There is a huge paradigm shift in organization
theory going on in the last 20 years. The accent is being moved from
hierarchy to heterarchy, from birocracy to adhocracy, from structures to
processes, from real to virtual, and from closed to open. Since I assume
that you are not familiar with this terminology (correct me if I'm wrong),
I'll give you a quick introduction.
The network organization or heterarchy is an organization consisting of
nodes (which in this sense can be other organizations, organizational units,
teams or particular people/employees) which are connected without a
permanent structure. It is an organization without hierarchy. The
connections between nodes are information based in their nature, and the
most important issue in such organizations is the sharing of information. In
the last 15-20 years a lot of such organizations were described in the
literature. Network organization, fishnet organization, the spiders web, the
spaghetti organization, the networked organization, the heterarchy to name
just a few. I won't bother you with all those concepts, but let me introduce
you to the concept I like most and which, I think, is most valuable to us;
the fishnet organization. I think I allready said a few words about this
concept, but for those who don't remember or havent read it here's the idea.
When you look at a fishers net on the beach it is fully heterarchical. There
are a lot of nodes which are interconnected, but there is almost no
structure. But if you take one node and lift it up you get a hierarhical
structure, if you lift another you get two hierarchies. In this manner you
can create and destroy hierarchies as you wish. Now taking this into an
organizational context, imagine an organitazion with no fixed hierarchy, no
exact order of command. The "invisible hand" which lifts the particular
nodes (people, teams etc.) up and down is their knowledge and reputation in
certain situations which occure in the environment of the organization. This
situations reseamble project teams for special purposed (e.g. a new product
line to be developt, or a new chance occures on the market) which create
this dynamic hierarchies. One thing to note here is that there's a whole
theory about hierarchical structures, and these are the most common
structures in organizations today. But, they have a lot of dissadvantages -
they are to slow, to birocratized and to rigid. On the contrary heterarchies
are fast, ad-hoc and easily to change and move.
Here we come to the terms birocracy and adhocracy. Birocratic organizations
have a strongly formalized behavior, e.g. everything is prescribed -
procedures, policies, constitutions etc. Thus birocratic organizations do
not easily change their behavior and course of action. Such behavior is good
in steady conditions where the environment doesn't change often and is
relatively stable. But in todays dynamic conditions where things change form
second to second (and the Internet is one of the main causes for such an
environment) such organizations easy get into trouble, need to restructure
or die. Thus the idea of adhocracy came into play. Ad-hoc organizations have
a rather unformalized behavior, can easily change and adapt to new features
form the environment. Their policies, procedures and constitutions are
created ad-hoc ("on the fly") and always changable.
Another issue to consider is the shift from structure to process.
Organizations were traditionaly organized according to organigrams
(organization charts) - structures consisting of different departments e.g.
production, sales, management, CRM, HR etc. But this meant that one and the
same proces (for example order processing) had to be fullfiled in different
departments - the proces was not a continous set of actions but broke into
parts (de-parted). The quality of these processes depended on all
departments which were involved (e.g. for the processing order example the
sales department, the ware housing department, the customer relationship
department etc. depending on the particular organization would be involved).
This means that the proces had multiple points of potential failure. The
bussines process rengineering paradigm and the ISO standards (e.g. 9000
series, 14000 series etc) shifted the accent on the proceses. Thus an
organization is not any more a system of departments, but a system of
processes. Every process has an owner which is responsible for the whole
proces, rather than just a part of it. The process owner could easily be a
team or a whole organizational unit if it's a bigger core process. In this
manner the core processes are easily determined and other non-core processes
could more easily be outsorced.
Virtual organizations are a term which came into play since the Internet
gone crazy. The idea behind a virtual organization is to create a virtual
entity which only seems to be an organization, but in fact are parts of or
whole other organizations. Virtual organizations emerge in situations where
one organization cannot answer the needs of the market by it self and thus
in collaboration with other organizations creates a virtual entity to get
the job done. This is a very strong concept and if undersood right extremly
powerfull. Virtual organizations exist in cyberspace and thus are something
we realy have to consider.
In a book I wrote in collaboration with prof. dr. sc. Miroslav Zugaj from my
faculty and in a paper which was published on the international conference
on production and engineering in Bihac we describe our term of the open
organization. A similar concept is the open organization described in
open-organizations.org or the permanently beta organization. The idea is to
use the concepts of OpenSource in organizations. A lot of interesting
concepts come into play in this manner like constant feed-back loops,
virtuality, autopoiesis, heterarchy etc. but I do not want to bother you
with those. The only thing I need to state here that a lot of organizations
are "opening up" nowdays since they need the constant feed-back of their
customers. On-line communities, forums and discussion rooms on organizations
portals are just some examples of such behavior.
Now, why do I tell all this. An information system is allways a subsystem of
a organization or a social system. Since the system we try to build should
be oriented towards the future all these concepts should be considered and
undestood. So we need to build a system which will be able to support a
heterarchical, ad-hoc, process oriented, virtual, and open organization. I'm
in the phase of publishing a paper on the autopoietical information system
which should be such a system.
The hererachy or more speciffic the fishnet organization could be
accomplished through a combination of SD2 (or a similar system), AD (or
similar) and OpenSource project management. Since SD2 defines a global
hierarchy it needs to be used locally in order to have a dynamic structure
(with multiple hierarchies). So the idea is to create a system similar to
SourceForge only for political projects and political organizations which
are managed locally through SD2. This opens the possibility to define
responsibility and fast reactions in emergence situations. Decisions in
normal operation should be made through AD. Also in this manner the problem
of small issues is resolved since every issue (no matter how small) could
have it's own project because every user can start a new project at will.
Adhocracy should be accomplished through a wiki system for every particular
project or organization. In this manner, because of the nature of wiki
systems to change fast, policies, constitutions, procedures etc. could be
changed quickly and thus organizations and projects on the system would be
adhocracies.
Process orientation can be accomplished through the Balanced ScoreCard
(BSC). The BSC method defines different perspectives of an organization's
vision, missions, strategies and strategic goals. These strategic goals are
accomplished through the accomplishment (and masuremant) of processes. Thus
in this way projects and/or organizations would be process oriented. Another
good thing with the BSC method is that it defines measurements and thus
enables the measurement of an organizations/projects efficiency. This detail
is important for the integrated decision making process I allready spoke
about, since it needs a way to implement and control (measure) decisions.
Virtuality is easily accomplished in an Internet environment. Every
project/organization can possibly have subproject/suborganizations. The
merging of OpenSource projects is a common thing to happen in the OpenSource
community. In this way we could create the posibility of creating virtual
projects/organizations through the system by merging/combining different
(sub)projects/(sub)organizations. Since every project/organization should
have it's own forum and since all projects/organizations will be
cathegorized there will allways be a chance that members of different
projects/organizations communicate and thus collaborate.
Openess is accompished through the TOP principles of political action which
I allready described (the document General Definition). In this manner all
organizations and projects stay open to the public, and all their processes
are fully transparent to the public.
This is the main structure of the system how I see it. Any questions,
suggestions, ideas and critics are trully wellcome!
Best regards
Markus Schatten, dipl. inf.
e-mail: markus....@foi.hr
http://www.tiaktiv.hr
illegale
Even I know what you are talking about, you are pretty hard to follow.
There are several concepts I suppose people are not fammiliar with and
that need some good explanation. So, I supose it would be usefull to
cut this stuff to basic concepts and to explain them in a way everybody
will get fammiliar with.
What I liked the most out of your texts is (guess hat :-)) fishnet
organisation, as long as it gives pretty good visualisation of info
flow in such optimized info flow heterarchy.
Though, as long as I am a little bit more familiar to your ideas, let
me ask you why do you find ad/sd-2 combo better to multileadership
pageranking system that could be in my opinion possibly good
aproximation for fishnet organisation?
To say a little bit more about AD, what makes that concept be better
than others? I am sorry to say that I only asume it has time/importance
category that is not involved in pageranking system such as sd-2 is.
Did I get it right about the essence of wlennes of AD or are there some
other parametrs I missed due to not so good lurking :-)
ATB,
Gale
Markus Schatten
On Thursday 20 April 2006 18:04, illegale wrote:
> Huhuhu. :-)
>
> Even I know what you are talking about, you are pretty hard to follow.
>
;-) I hoped people would ask questions....
> There are several concepts I suppose people are not fammiliar with and
> that need some good explanation. So, I supose it would be usefull to
> cut this stuff to basic concepts and to explain them in a way everybody
> will get fammiliar with.
>
> What I liked the most out of your texts is (guess hat :-)) fishnet
> organisation, as long as it gives pretty good visualisation of info
> flow in such optimized info flow heterarchy.
Well, there does not seem to be interest in explanation... That's sad since I
think this is crucial. Modern organizations are a great example of how the
Internet changes the way people think, manage knowledge and achieve better
results. By "only" creating a democratic forum, all this great possibilities
wouldn't be used.
>
> Though, as long as I am a little bit more familiar to your ideas, let
> me ask you why do you find ad/sd-2 combo better to multileadership
> pageranking system that could be in my opinion possibly good
> aproximation for fishnet organisation?
Well, I actually don't find AD/SD2 combo to be better ;-) My idea was to use
this good concepts in a multileadership system (since these concepts are only
a small part of the story) ;-)
>
> To say a little bit more about AD, what makes that concept be better
> than others? I am sorry to say that I only asume it has time/importance
> category that is not involved in pageranking system such as sd-2 is.
> Did I get it right about the essence of wlennes of AD or are there some
> other parametrs I missed due to not so good lurking :-)
>
I do not say AD, SD-2, CLD2 or any other concept is better; I wanted to leave
them as possible options in a multi-concept system. In this way users can
choose, test and optimize them. Eventually the "best" concept and/or
combination of concepts will filter out (autopoiesis).
Best regards
--
illegale
> Dear Gale
>
> On Thursday 20 April 2006 18:04, illegale wrote:
> > Huhuhu. :-)
> >
> > Even I know what you are talking about, you are pretty hard to follow.
> >
>
> ;-) I hoped people would ask questions....
Hoped wrong :-)
> > There are several concepts I suppose people are not fammiliar with and
> > that need some good explanation. So, I supose it would be usefull to
> > cut this stuff to basic concepts and to explain them in a way everybody
> > will get fammiliar with.
> >
> > What I liked the most out of your texts is (guess hat :-)) fishnet
> > organisation, as long as it gives pretty good visualisation of info
> > flow in such optimized info flow heterarchy.
>
> Well, there does not seem to be interest in explanation... That's sad since I
> think this is crucial. Modern organizations are a great example of how the
> Internet changes the way people think, manage knowledge and achieve better
> results. By "only" creating a democratic forum, all this great possibilities
> wouldn't be used.
I suppose it is up to presentation. You are habdling pretty cool
concpet. If you want t introduce them to other members, it is better to
do it easaly, one by one.
> >
> > Though, as long as I am a little bit more familiar to your ideas, let
> > me ask you why do you find ad/sd-2 combo better to multileadership
> > pageranking system that could be in my opinion possibly good
> > aproximation for fishnet organisation?
>
> Well, I actually don't find AD/SD2 combo to be better ;-) My idea was to use
> this good concepts in a multileadership system (since these concepts are only
> a small part of the story) ;-)
So, we are pretty simmilar thinking.
> > To say a little bit more about AD, what makes that concept be better
> > than others? I am sorry to say that I only asume it has time/importance
> > category that is not involved in pageranking system such as sd-2 is.
> > Did I get it right about the essence of wlennes of AD or are there some
> > other parametrs I missed due to not so good lurking :-)
> >
>
> I do not say AD, SD-2, CLD2 or any other concept is better; I wanted to leave
> them as possible options in a multi-concept system. In this way users can
> choose, test and optimize them. Eventually the "best" concept and/or
> combination of concepts will filter out (autopoiesis).
Absolutely agreed.
ATB,
Gale
echarp
them difficult to recognise.
I'm a cluetrain and agile programming guy.
Simple, ad-hoc and transparent is better than hierarchic, rigid and
opaque.
Hierarchies are useful, if they can mutate and change as their parts do.
MG
I think all of us have there own agenda and a certain vision until now
on how to reach the goals of the agenda.
In this way we could all get what we need and some good darwinism will
most certainly occur on the way!
And we will also get to know each other and form a very powerful
network in the same time.
illegale
Markus Schatten
Markus Schatten
On Friday 21 April 2006 13:47, echarp wrote:
> I share most of the concepts presented, although their presentation make
> them difficult to recognise.
If there's anything of the concepts what you would like to understand better
feel free to ask! ;-)
>
> I'm a cluetrain and agile programming guy.
:-D
>
> Simple, ad-hoc and transparent is better than hierarchic, rigid and
> opaque.
>
> Hierarchies are useful, if they can mutate and change as their parts do.
Exactly! Very nice and easy explained.