1) Globality is a higher level of
complexity: complexity is not a novelty
we are used to complexity: we face it at every
level of our actions.
a complex system can be schematized as an
hierarchy of interrelated components, integrated in sub-systems which contribute
to the system objectives and, by their interactions, actually represents the
system itself. Outside of the system is the environment. The components of the
system are made of simple (atomic) bricks.
this simple scheme can be used to build much
more complex systems as a "self-similar" object: using different
"unit ruler" to observe the system, we will perceive always the same
basic structure. By using a magnifying lens the atomic bricks become actually
complex systems, that are made of lower scale "atomic" bricks, and
having as environment the higher level components of which they are parts.
Looking at the earth from a satellite it can be seen as a system where what we
on the earth perceive as our environment is actually part of the system.
what is new of globalization is that it has
raised the scale of complexity, so that at least a part of what we were used to
consider as our environment becomes actually part of the system itself.
2) Can we learn from past
experience in decision- making / problem solving?
if we are used to complexity, then - as active
members of the system that we perceive as our own - we are used to act (react or
pro-act) to respond to the system challenges.
if so, we should hope that to deal with
globality challenges we can learn from our experience
the trick we uses in dealing with complexity is
to limit the range of the complexity of what we consider "our" system:
we "expel" higher levels out of the system to make them part of
exogenous environment (we cannot act to change the environment; we can only try
to forecast the changes to take pro-active decision), and we accept as
"atomic" unbreakable parts some of the inputs we use to (reactively or
proactively) change "our" system.
the case of the automobile manufacturer can
serve to illustrate the point: he knows that the automobile is simply a part in
a complex transport system, and that he could consider alternative solutions to
realize the transport system objectives (including those where there are no
automobiles). However, he considers the automobile as the system he should deal
with. The transportation system for him becomes part of the complex environment
(which includes all the other transport modes). Notwithstanding this
simplification, the system remains a very complex one (with complex sub-systems
and components) and the innovation changes to respond to the transportation
challenges are quite difficult to meet. He knows that the components that enter
into the automobile product are complex systems on their own terms and can be
the subject of innovation changes. However, he has to assume that they are - at
least part of them - given and unchangeable. The oil shock and the concern for the air quality
have compelled automakers to consider the higher dimension of the transport
system (even if up to now they have found possible to respond by adapting the
automotive product without having to plan higher transport system level
changes).
3) The design paradigm for
decision-making / problem- solving
we are used (since Descartes) to consider a
reductionist approach to problem solving: to break down complex issues into
small pieces and from them to build - bottom-up - a complex solution.
however, this implies an intrinsic linear chain
of cause-effect relationship between the elements of the system where the
problem emerges. Since we know that there are non-linear feedbacks in complex
system, the reductionist problem-solving approach implicitly assume that the
feedbacks could be "frozen", taken for given, be part of the past of
the system and not significantly modified by our intervention to change it. This
approximation is in many practical case a satisfactory one (very small actions
on a highly inertial complex system).
if globalization brings to the fore
interdependence among system members of such a magnitude to challenge the
identity itself of each member (changing system structure), than the
reductionist / separation (decoupling) of variables problem- solving method is
not any more satisfactory.
we should therefore look at other procedures,
which accept intrinsically the non-linearities. One such procedure is that of
the design.
even in the very simple case of an architect
that have to provide a solution for a customer wishing a new house, complexity
(and feedbacks) enters in a not eliminable way. The process start with the
customer having a vague idea of what he really want. Before the architect can
accept the job, a better and clear definition (problem specifications) of the
terms of reference of the problems have to be developed. In a linearized model,
this will be the responsibility of the client. Instead, he needs to interact
with the architect and together try to "design" the specifications.
This is done by a complex process that take into considerations the existing
portfolio of ideas on types of housing, preliminary sketching of alternatives,
etc. The process ends when the client take the responsibility to say that his
initial vague desires are now well explained. We can call this initial phase (of
better defining what the real problem to be solved is), a
"meta-design" phase meaning that the interactive complexity of going
through all the solving steps has already intervened even if only in a
"simulated" design.
Now the problem passes to the architect direct
responsibility. However very seldom we can say that the problem can right away
be broken down into small pieces. The designer has accepted the terms of
references on the basis of potentially conceived solutions. However before he
could develop the real solution, a creative phase has to intervene. And this is
a most non-linear process where all the variables are again put together, the
specification is challenged items by items, somewhat rejected in the designer
mind (the phase is called a "divergent" one), up to the point when the
designer feel that the process can stop and a "final" solution
emerges.
The solution is still somewhat theoretical and
global, lacking details. One or more designing steps follow to detail the
solution. Again, "divergent" stages of thinking might be necessary
before "converging" to satisfactory (good enough) solutions at
subsystem or component level.
We have said that this designing phase is the
sole responsibility of the architect. However, the client will be called in, to
have is agreement on details which can change the agreed specifications, the
terms of reference. And the client intervention might require to modify
envisaged solutions.
The process is an iterative one also when we pass
to the implementation of the detailed design. Here the major responsibility pass
to the constructor. However, no matter how detailed are the blue prints, there
will be instances where the design specifications have to be changed (since they
are contradictory or in contrast with the reality of the available materials or
because of the 3-dimensional effects of a building which was only simulated in
the 2-dimensional blue prints). Most of the interactions in the construction
phase are between the constructor and the architect. However, in some instance,
the client himself has to intervene to take final decisions.
in the design case we are faced with a
"micro-world" complexity which however contains all the features of
complexity and non-linearities of a "macro-world" complexity.
the proposal, here, is to refer to the design
problem-solving process as a paradigm for the highly non-linear case of
globalization challenges.
there are two extreme models of human behavior
in approaching complexity (two cultures): the one represented by the human and
social scientist and the other by the physical scientist. We can refer to the
first approach as "holistic" and the second as "reductionist".
These two models have both contributed to the advancement of our cognitive
approach to understand the world. However, neither one nor the other are suited
to design actions to change our world.
The "design" approach is actually a
compromise between the two: it is an interactive chain of "holistic"
cognitive approach and "reductionist" building of solutions.
the design paradigm suggests that an important
preliminary organizational rule is to be followed: to define clearly the roles
of different actors in order to contrast the fuzziness caused by complexity and
interdependence. Only with a clear definition of actors we can overcome the
effects of feedbacks that blur actors and phases of problem-solving. The process
can therefore be considered as a chain of circles centered around each actor.
Each circle is covered a certain number of times involving the interactions of
all the other actors up to a point when good enough convergence is met, so that
it can be decided to pass the responsibility to the next actor down the problem
solving chain.
one of the major difficulties encountered in
dealing with global issues is actually the lack of such clear role subdivision.
There is a perceived challenge to be met, therefore there is a demand to do
something. But, who materialize such demand in term of playing the role of the
"client"? S&T can represent a potential supply of ideas and
initiatives to develop solutions. But how do we organize the related response?
the difficulty to clearly define actors and
roles is however not new. Even in the case of producing goods to satisfy clear
needs, not always the case is as simple as the one of building a house for an
individual client. The "client" usually is represented by the market
which is not a physical entity that can express the needs in terms of product
specifications. The producer has to guess what the "market" (the
future potential costumer) will appreciate. To do this in practice, the producer
has to simulate himself the role of the client (see the firm's complex internal
organization with well separated functions and roles such as Marketing, Design,
Production). The organization procedure in a firm is simple enough for products
that are already well diffused on the market and the problem-solving has to do
with their slow adaptation to the market and S&T changes. The problem is
much more difficult for the case of a radically new product. In such a case,
changes in the firm organization are often needed (see the project and/or matrix
organization). We know by experience how difficult are such cases, and how many
firms fail in setting a proper organizational procedures to deal with the
novelty.
On the other hand the success cases show the
importance of clear definition of roles to simulate client/ designer/ producer
and the assumption of related responsibility by some one.
globality issues are more similar to the case
of radically new products. The existing organization (at all levels, local,
regional, world-wide) is not apt to deal with some of the more radical issues of
globalization. The design paradigm suggests that a central role to be developed
is that of the "client".
we might very well feel the globalization
challenges, but unless we will clearly succeed in transforming them into problem
specifications it will be difficult to allocate the needed resources, to call
the contribution of the S&T community. We know that there is no physical
"client" to represent the society in general. However, like in the
case of the market, someone has to materialize such a role.
not all the globalization issues pertain to the
same class. Some are real world-wide, some are emerging at local level through
global interdependence. The "game" of decision making and problem
solving has to be played at a scale which is proper to that of the single issues
we are concerned with. But in all the cases the same design paradigm can be
applied and actors have to be clearly defined to set to motion the
demand-supplier mechanism of problem solving.
5) Two
basic complexity features:
"competition vs cooperation" and "development through unbalances"
supposing that we have a satisfactory enough
procedure to approach problem solving under complexity, then we have to ask if
our actions can be big enough to change sensibly its trajectory. In the case of
globalization, are not possible actions too small to hope to produce results in
a relative short time? Even if this is the case, we not necessary have to
despair, provided our actions can count on "leverage" effects produced
by the forces internal to the system. A well known physical examples is that of
resonance: we can produce big effects with small perturbation, provided,
however, that we "know" the "frequency" of resonance of the
system.
in general, the more we know about the system
the higher the possibility to intervene on it successfully. It is therefore
important that we understand some basic features of the dynamics of complex
systems. However, how can we expect to understand the features of a system which
is going through radical changes? We have , in fact, to presume that the
globalization process is actually representing such a radical change in our
"global" system.
a "conjecture" (not demonstrated but
for which there are supporting evidences) is that in going through a radical
change the system will continue to be "self-similar". It might have
increased the number of the levels in its hierarchical structure, but the same
basic features will be reproduced at the new higher scale.
we point here to two major self-similar
features of complex systems: "cooperation vs competition" and
"dynamic development through unbalances".
this conjecture might apply only to the system
which is of our specific concern (the human world system) or be common to a
"class" of systems. In the first case, we can hope to learn what are
the "self-similar" evidence of the system by looking only at it. In
the second case we could have hints and knowledge support by observing other
systems of the same class, since some of these might be well studied.
In fact, we can say that a second
"conjecture" is usually implicitly assumed in our cognitive approach
to the complex human system: all the systems belonging to a certain class (open
and self-organizing systems) share similar behaviour. We can therefore use the
knowledge of other systems as a "metaphor" for the one of our own
concern. The most important metaphor that have been used extensively is that of
the biological system.
the biological metaphor points out as basic
system features that explain the system dynamic, those of "generation of
changes" and "selections of the fittest". To this, one should add
that the "elements" that belong to the system do "compete"
between themselves to pass the selection, but implicitly or explicitly they also
"cooperate" (co-evolution) to be better off in passing the selection
mechanism (some time also succeeding in modifying the environment).
these features are self-similar: we found them
at all the levels of the biological scale (from genes, to cells, to individual,
to species).
the second feature is that evolution
(development) is not continuous but it goes through period of stability and
period of changes (punctuated evolution). In other words, the system cannot
change radically (at the level of concern in the system hierarchy) unless it has
cumulated a large enough unbalance.
the use of the biological metaphor as an
heuristic tool to the cognitive approach to the human system, has pointed to
similar features at different levels of observation (from individual, to groups,
to larger social aggregates, in the social behaviour as well as in the
technological realm).
6) The globalization process has
already produced large unbalances: the global issues
the elementary definition of a system is that
of "a set of elements that interacts among themselves". In a spatially
defined system, an important characteristics is the "range" of the
interaction. Elements can directly interact with elements that are very close,
or, in other cases, also to very far distant ones. Elements, because of
interaction, can aggregate in groups (sub-system) which can interact as a group
with other groups or elements.
The evolution of the human system (thanks in
particular to technological progress) has been characterized by a three-fold
phenomenon: the increase of the "spatial" span of interaction among
human beings (e.g. by new communication means); the reduction of "time to
interact" (not only through telecommunication but also through physical
interaction); the increased "reactivity" of the "global"
system to local actions (see the cases of ecological concern). Globalization can
be defined has the undergoing step (the last one?) in this evolution of
increasing the "range" (space, time, reactivity) of human interaction
to the dimension of the earth itself.
the ultimate consequences of this change should
be that no one is small enough to consider the effects of his action negligible
at the global system scale. We know that there are indeed physical systems where
this is the case (deterministic chaos). However we should hope to be far from
the chaotic end, and that the human system will find ways to restructure itself
so that small actions produce again small effects (and local problems could be
met with local actions).
the increase of range of interactions between
system elements is a potentiality which is not evenly distributed. There are
elements (or groups of elements) that can use the increased interaction range to
their advantages or which are in a context that amplify their action effects.
There is a "reinforcing mechanism" so that the diffusion of the
globalization basic system changes follows privileged paths. We could use the
word "percolation" to synthesize such a non uniform diffusion process
(percolation is the phenomenon by which a fluid diffuse in an apparently
homogeneous filter by preferred paths which the fluids cuts itself through the
filter).
"percolation" produces a separation
among the elements of the systems, between the "integrated" and the
"excluded". This will produce unbalances to which the system will
later react. The system reaction will be proportionate to the degree of
unbalances reached (with the possibility of overshooting).
the risk is that the globalization process is
producing too big unbalances, too widening gaps between the
"integrated" and the "excluded". Is already this the case?
what we call the "globalization
issues" are actually a result of too great unbalances which have already
developed. Unbalances in the uses of natural and artificial resources (such as
the S&T endowment), have resulted in "global" negative
externalities produced by the "integrated" that interest both them and
the "excluded". This, however, is not the only paradoxical situation
of an unbalanced globalization process.
another paradox is that the ones that are
excluded from the increased ranges of the potentially positive factors, cannot
consider themselves as "small" with respect to the global system.
Their actions have also negative global externalities (see the deforestation
effects in the LDCs regions) to which the "integrated" react.
7) From challenge to action: how
to help the globalization process to diffuse at all system's levels through
competition/cooperation to avoid too great unbalances
when passing from the globalization challenges
to actions, we have to distinguish between the process of globalization and the
global issues already apparent.
First preoccupation is to help the globalization
process to develop in a uniform way, avoiding producing great unbalances.
globalization is a process that not only we are
forced to accept, but that should not intrinsically be considered harmful.
by observing the unbalances in human systems we
note that they increase when competition (or self-interest) is pushed to the
extreme without at the same time developing some kind of cooperation.
one possible reaction to reduce unbalance or to
stop its growth, is to develop "barriers" against competition. This
response will have only transient effects, and, at the end, increase the
unbalances to be followed by great oscillations to readjust the unbalances to a
more reasonable level.
In fact, to set up barriers is in contrast to the
intrinsic system trend to increase interdependence and interaction.
"General wisdom" suggests that the ones that contrast
"global" system trends are condemned to be "losers"
("great men have always 'understood' and been 'interpreters' of their
time")
a better solution is instead that to favor the
system trends by taking advantage of the system "leverage" effects.
Our suggestion is that cooperation is the other side of competition and it is
intrinsically tied to it. We cannot consider the one without the other, at all
levels of human actions (individual, local, regional, global).
cooperation should not to be confused with
"solidarity", with altruistic behavior. Cooperation instead has to be
seen as of direct interest of all the cooperating partners.
limiting ourselves to the S&T sector we
note how important is the competition/ cooperation behavior at different level
of actors scale:
i) scientists compete among themselves to assure
individual recognition, while they eagerly present and discuss their results or
intuitions with peers;
ii) entrepreneurs will search for appropriable
"localized" innovation change while he will develop indirect and
direct (e.g. joint ventures) methods to get access to others' know how, so
contributing indirectly to the increase of the endowment of "generic"
technological know how;
iii) cooperation to develop common technological
standards is an important features of our S&T progress provided the
standardization does not extend to the limit of freezing all the freedom to
develop new ideas and to adapt products to the variety of users and needs;
iv) S&T planning will not be conceivable
unless a combination of top-down (a sort of "cooperative" effort to
individuate goals, allocate resources, etc.) and bottom-up (competitive)
actions.
the competition/cooperation loop, to be
effective, has to close at all levels of actions. Instead there is a tendency to
separate the levels where actors have only to compete, to the ones where actors
have only to cooperate. An example of how this behavior have increased
unbalances come from the development of urban social environment. In the past
different "classes" of peoples inhabited the same buildings where
there were a clear distinction of the different social position of the tenants
(even in the height of the different floors). However, the vicinity of poor and
rich tenants made possible, on the other hand, a kind of cooperation-solidarity
to develop (the poorer tenants supplying services to the richer ones). Unbalances
were evident and undesirable. However, the result of shifting the responsibility
to close the competitive/cooperation circle to an higher level (through social
state solidarity) have produced urban ghettos (around social housings) shifting
the unbalances to a much higher scale and making the situation even more
undesirable and unmanageable.
one instance of separation of the competition
and cooperation role is evident in the case of the EC R&D policy: in
principle only 'pre-competitive' R&D can be the object of EC initiatives. On
one hand this might limit too much the effectiveness of the EC use of R&D as
a tool for its objectives. On the other hand, in case it be important an R&D
public policy to develop "products", leaving such possibility only to
the national governments might lead (as it actually happens) to develop national
'flag' enterprises.
a general recipe to respond to the
globalization challenges is therefore not only to push for more cooperation but
to be assured that competition/cooperation develops at all system levels. The
problem is particular difficult when proper actors/institutions are not yet
developed at the new "global" sub-system.
8) From challenge to action: how
to respond to the global issues
globalization has already produced too large
unbalances that can be interpreted in term of "global issues". While
on one hand one should develop policies and actions to avoid the further growing
of such unbalances (counting on the leverage effects of the globalization
process to reduce the scale of the global issues), on the other hand we cannot
wait for the intrinsic ('natural') reaction of the global system to reduce the
unbalances (not only it will take too long time, but the ways might be too
harmful).
to respond to specific challenges the recipe we
suggest is that of the 'design paradigm'. The complexity of the global issues
has to be accepted as an intrinsic feature. However the design paradigm tell us
that we have to 'simplify' the complexity by deciding what is the 'internal'
complexity and what can be considered 'exogenous' to the decision-making scope
of action. In other words, we have to enter into a process of issue
classification and of problem definition, defining the roles of the 'client' and
of the 'designer'.
'global' issues can in fact be classified as
issues whose 'complexity' can be reduced to local dimension, or to regional, or
to really global one. According to the different classes, different actors will
have to take the responsibility to play the role of client and designer.
the design paradigm can be seen as a process of
cooperation/ competition between the client and the designer where the emphasis
is on the one or the other of the two according to the phase of the problem
solving process. The requirement that the role of the client and of the designer
be realized at a system level coherent with the problem 'dimension' is therefore
another aspects of the general rules of avoiding too great unbalances (in this
case unbalances between the ones that represent the demand and those that
represent the ability to respond).
the problem is particularly difficult with
global issues because any approach to respond with solutions will produce
negative externalities as well as benefits. The difficulty comes from the fact
that often the one that suffers of the externalities are not the same that
benefit from the solution (see the case of transport infrastructure). The
definition of what the problem really is and of its 'dimension' depends from the
possibility to close the balance between the losers and gainers from the
solution.
9)
S& T is a powerful
determinant of change: can we aim at S&
T based innovation changes to help
the
globalization process to diffuse reducing the integrated-excluded divide?
S&T is itself part and determinant of the
globalization process. Therefore S&T should be looked at to avoid too great
unbalances to develop.
as a matter of fact, one can notice a
"percolation" phenomenon in the diffusion both of Science and of
Technology which contribute to the globalization unbalances. While Science in
principle is a "generic" endowment of human kind, the ability to do
scientific research depends from the level of local education, from the
availability of financial resources and of infrastructures. Moreover, even if
one has the possibility to perform research activity, the choice of the research
agenda is conditioned by the existing directions of the research activity, by
the 'internal rules' that characterize the ways by which the scientific
community behave. All this will tend to reinforce the role and choices of the
countries that lead the scientific development.
there is an important, even if very difficult,
plan of action to be undertaken directed on one side to increase the
participation of the 'excluded' to the scientific research activity, and on the
other side to shift directions of research.
one should therefore recognize the importance
of investments on 'intangibles' with priority on education and scientific
research. The more such investments will reach the excluded part of the world,
the more there is a chance that also the research agenda will become broader and
cover areas of future interest (in term of opening potential application
concerned with dealing with globality issues).
along with the availability of resources for
such intangible investments, organization intervention will necessary be aimed
at increasing the efficiency of the investment, but also at helping the process
of change of scientific exploration. Ideas such as developing a 'commonwealth of
Science', twinning of research institutions, networks of research between North
and South countries have to be encouraged.
concerning the Technology side of S&T the
'percolation' phenomenon is even more evident. Application of science for
practical purposes tends to be pushed by 'innovators' to get entrepreneurial
advantages (localized technological changes). Competition is the basic
motivation. Other aspects of the globalization process, such as the opening of
the world markets, the increased scale of productions and of size of firms tend
to keep the technological knowledge internal to the innovating enterprise.
it is therefore important that one develop
policies to push the cooperation side also for technological development.
Fortunately, the increasing resources needed to develop important innovation
changes and the necessity to follow many alternatives research routes to assure
final success, push entrepreneurs towards some kind of cooperation
(pre-competitive research, joint ventures, etc.). There is therefore matter for
intervention to increase cooperation, including networking of companies from
advanced and developing countries.
10) S&T and global issues:
can we plan for innovation changes in complex systems?
to deal with the global issues by applying
S&T potentialities means to 'plan' innovation changes. The 'design' paradigm
help in avoiding the trap of considering planning as a 'rational, linear'
approach in problem solving, and in accepting complexity as an intrinsic
feature. However the 'scale' of the global issues and their 'higher
hierarchical' level of complexity make the problem difficult even if we are
determined to use the design paradigm (including creating new institutions if
needed to play the client role).
first of all we have to accept that an
innovation plan rarely could be a one shot operation. Actions might be aimed at
developing the 'building blocks' of future solutions. When this step is
performed, the imagined solution might not be possible, because the resulting
'building blocks' are different than originally conceived, or because the new
developed information might show that more radical changes might be necessary.
in planning innovation changes for complex
systems, a hierarchy of innovation objectives has to be posed according to the
system hierarchy: innovation objectives can consider only components innovation
(tactical innovation policy), or consider subsystem innovation (strategical
policy), or even the entire system change (structural policy).
the case of transport issue will clarify the
idea. The challenges includes energy conservation, environment protection,
avoiding saturation of transport infrastructure and improving quality of life in
urban environment. Contribution to the challenges can be obtained aiming at
innovating the today public or private vehicles (the 'components' of the
transport system). There are however limits in the result that can be obtained
with such constraints in innovation objectives. One can think to innovate at
'subsystem level', e.g. developing new and more efficient public transport that
will produce a shift in demand from private to public transport. 'Components'
(building blocks) for such innovation plan might not, however, be there. So,
efforts should be dedicated to prove that new solutions can be developed for
public transport (e.g. public modes of transport that can have a flexible,
demand responsive, routing). Finally, the results might not be satisfactory,
unless one innovate the entire system, which might require to include, in our
planning, the change of the 'environment' of the transport system (e.g. urban
planning to reduce saturation effects of congested transport demand).
the more we plan for innovation changes that
includes higher levels of the system hierarchy, the longer will be the time for
solution, the higher the uncertainties. It will be difficult to have the needed
society consensus on innovation actions if we can point only to long term
uncertain successes. So, an acceptable plan should include a high proportion of
short term tangible results, to make acceptable the devoting of resources to
more radical and longer term actions. It is therefore important to show, also
for the case of globalization issues, that not all conceivable actions are long
term, difficult to realize, requiring radical organizational and institutional
changes.
A classification of globality issues shows that
indeed they can be classified according to their 'range' (as local, regional, or
global), and, even when having a real global range, that result can be obtained
acting on components, or sub-systems.
an innovation plan to respond to global issues
should make the following assumptions: first assumption: the today 'system' (the global
system, including the environment, the social and the S&T systems) certainly
has 'slacks' available to adapt to the globalization challenges at least to a
certain extent, without having to change its structure or prevailing values.
Priority should therefore be given to take advantages of such slacks (component
innovation changes); second assumption: there are global issues which
could not be dealt with unless innovation changes are aimed at sub-systems; third assumptions: the long term challenges of
globalization cannot be met unless basic changes in value and behavior are
realized (system innovation);
an innovation plan to deal with global issues
has therefore, to be balanced, to include actions that have objectives at the
three levels.
11) Applying the design paradigm
approach to all innovation levels: the institutional, organizational and
policy
consequences
the design paradigm emphasize the
client-designer interaction. The difficulty with many of the global issues is
that not only the issues are far from clear (which is a characteristics of any
design problem in the first phase of the problem-solving approach), but also
that the client is not 'defined' or the role is played by an 'improper' client
(at a level different form the issue level). To pass from challenge perception
to problem definition one will have, therefore, to make institutional
development as an intrinsic part of problem solving.
the specification of the problem to be solved
will depend from the 'client' values, which on their turn will become real
values (action inducing) if the client together with the designer can perceive
possible solutions. Out of metaphor, if the real client is the society at large,
the first difficult task is that of resisting the idea that roles can be
separated between society, politics and S&T. The intermingling of S&T
with the other society systems means accepting the intermingling of the
respective uncertainties (in values, priorities, ideas for S&T
potentialities).
the design paradigm can be applied here at a
'meta' level: 'designing the design' in order at the end to be able to start the
real design process having established actors and roles (client/designer) and an
agenda of issues classified in terms of dimension (local, regional, global) and
of innovation objectives (component, subsystem or system changes).
The examination of few different issues will show
better the importance of the meta- design phase.
the deforestation issue. The challenge is
perceived as a real one. However there is no consensus of what are the possible
solutions (stop deforestation in the LDC's, start reforestation in the rich
northern country, develop 'artificial forest' as a sink of CO2, etc.). Examining
the portfolio of ideas of potential solutions (applying the design paradigm)
sorts the effects to point out how vague is the issue that we want to deal with.
Is it really deforestation because of the potential effects on climate change or
on loss of biological diversity, or on both? And what are the priority values:
the longer term survival of human kind or the short term improvement of the
poorer part of the world? Can S&T help with short term actions that can at
the same time reduce deforestation and increase rentability of forest
conservation for the LDC's? The looking for ideas now has to be shifted in new
directions. The real issues for S&T might be that of finding uses (and
therefore values) for the natural 'waste materials' produced by the forest, or
of how to intervene on forest to increase production of useful materials in a
synergetic approach with nature. At the end of the 'meta-design' exercise a
panoply of issues with related potentiality of responses might emerge, which
could be classified from very long term (needing new direction of basic
research), to more practical short term ones (such as developing more effective
mechanical ways to get materials out of the forest without destroying it).
desertification. The case is quite different
from that of deforestation where (apart from innovative approaches based on
S&T) a simple solution might - naively - be considered possible if the
richer part of he world is willing to pay for the value represented by keeping
the forest. Instead, in the case of desertification the dynamic of the process
is not well known and it will be in any case difficult to point to the
'culprits'. Desertification has certainly a very negative impact on local
populations. However, are we sure that not only to stop the further
desertification, but also to reduce the existing desert area is an objective to
be pursued? What might be the induced effects on climate change? We know already
that projects to recuperate desert lands to agricultural activity, distort
important water resources shifting problems to other areas.
Could instead human kind take advantage of the
existing desert area with its clear skies looking for S&T cheap solar
energy? A project along this line will certainly require huge resources that
will produce return only in the long term. The problem is therefore that of
proper allocation of resources. However the analysis of potential ideas might
point to effects of 'virtuous economic circle' of wealth generation with short
term effects also on local population. The meta-design in such a case will
result in changing a negative global issue in that of the exploitation of a
resource.
marginal agriculture. An important global issue
is that of inhurbation and abandonment of the agricultural land. On one side
this is the result of the increased productivity in agriculture. On the other,
it is simply the abandoning of agriculture activity because of non redditivity.
The phenomenon produces great unbalances:
abandoning of marginal land (such as hills and mountains) in rich countries
(with the consequence of decaying biological process and soil erosion);
abandoning of potentially rich soil in underdeveloped countries which lack the
investment resources needed for an high productivity agriculture (with the
consequence of reduced income and of population migration); increasing
environmental damage due to the use of pests and fertilizer; reduction of the
variety of species because of selecting only the most productive ones;
protection of agriculture in rich countries through custom barriers (a cost to
the rich nations with further effects on poorer countries that see reduction of
export markets for their agricultural products, which are often the only
products that could be exported).
What are here the real issues? What tasks can be
put forward for S&T?
In rich northern countries one issue is that of
keeping people on the land for the importance that the human presence has on
geo-biological equilibrium. In such a case should the objectives for S&T be
that to develop technology that make profitable agricultural activity on
marginal lands?
Or the objective is different, and one should
look for an occupation of the territory for other economic activities with
surplus income destined to a good housekeeping of the territory (not so much to
produce agricultural products, but to assure a good biological and geological
soil maintenance: peasants as gardeners?)? Since the existing system of
protecting an high productivity agriculture has a cost to society (see the case
of the EC Common Agricultural Policy) will the issue (to be translated in terms
of S&T) in such a case be a better occupation of the lands at a lower global
cost to society?
Or, could new agricultural process be developed
that at the same time reduce the productivity, assure occupation and
good-housekeeping of the territory and also a profitable activity?
Can we separate the issues of rich countries and LDC's, or the solutions should be developed in an unitarian scheme?
the agricultural case shows how intermingled
are economic, social, technological, international aspects; how issues
definition depends from values definition and perception of potential solution,
from a balanced mixture of competition and cooperation between different society
sectors (agriculture, fisheries, industrial, service) and between poor and rich
countries, between the ones with food surpluses and the one dying of famine.
The 'meta-design' exercise, will point out to
radical new ideas as well as some more short term ones, with related tasks for
S&T. The clear definition of issues and potentiality of solutions will be a
preconditions to get the determination needed to allocate resources in so
radically new ways with respect to consolidated today approaches.
S&T will help the phase of transforming
challenges into specific issues by contributing ideas of possible solutions. The
completion of such a phase and the definition of actors, is a precondition for
the real contribution of S&T. However, accepting the basic interactive
nature of problem-solving, one should consider that each actor in the
"loop" can react to the inputs he receives from "above", so
'pro-actively' producing feedbacks that will change, later on, such inputs. In
other words, in an interconnected decision-making / problem-solving situation,
there is no excuse for any one actor to stay still, waiting for clear and better
inputs.
S&T actors should therefore start their own
planning for actions. But how to plan S&T? A plan is always a mixture of a
top-down and a bottom-up approach.
In practice the methodology suggested is the
following:
i) assume that the information on the issues at
stakes - as vague and uncommitted as they are - could represent a first sketch
of a top-down settings of general objectives and finalities;
ii) evaluate the existing portfolio of ideas
classifying them in term that can more easily and directly refer to the
globalization issues;
iii) with the help of such classifications, spot
underway directions of S&T development that are susceptible of covering the
areas of concerns for the given issues and plan for priority actions to continue
R&D along such directions;
iv) identify S&T areas of potential interest
on which no activity is underway and make an effort to define objectives and
ideas to initiate R&D on such areas and domain.
such analysis will at least provide a way to
measure how consistent are today directions of S&T development and the need
to change them or the related priorities.
12) Conclusions: applying the
complexity "wisdom" to globalization
the word "rational" for a
problem-solving approach bear an intrinsic analytical / reductionist flavor. To
deal with the globalization issues, we need a better and less compromised word.
We propose that the approach is better represented by the use of the word
"wisdom".
our basic hypothesis is that there
"wisdom" is available that allow us to deal with complexity. The
suggestion is that such a wisdom should be applied to the the very complex case
of globalization.
from the experience of problem-solving in
complex situation we point out the "design paradigm" as the one that
capture the intrinsic features of complexity. In fact the design paradigm
accept: vagueness of problem statement, strong interactions and blurring of
roles of the different actors involved. However, it also provide a
"recipe" to find ways out from an endless looping of interactions.
referring to the design paradigm permits to
point to very simple general "wisdom" recipes (such as that of
recognizing the "dimension" of the problem in order to choose proper
actors) for the behavior of each actors, even before starting the real
problem-solving activity. It also provides more detailed "wisdom"
recipes for problem solving.
the design approach can be applied to the
S&T realm, to develop its role on globalization.
in the problem-solving loop, there is no
privileged points to start the process. Responsibility is diffused and each
actors can start the process.
In particular, S&T actors have a
responsibility of their own, reacting to the request from other actors to
intervene, but also pro-acting anticipating future requests and even
predisposing for such requests to become possible.
Appendix: Proposal for an experimental approach to an
action program on Globalization
Globalization represents a change in human system
complexity. To deal with it, it is necessary to assure variety of responses and
flexibility. We need a creative approach on all the components of the action
process: from organization, to problem definition, to solution design. Because
of the uncertainties even in the definition of what the real problems at stake
are, to assure a variety of approaches is more important than the attempt to
select priority issues or to better focus the actions. We need to learn how to
deal with the globality issues: so the approaches should assure, through variety
and flexibility, that even errors will contribute to such learning.
All that makes difficult to converge the
necessary will power and resources to develop practical actions: in fact we are
too used to consider that action programs should be well spelled-out and
assure selection of priority. We should therefore have the courage to admit that
a clear and well focused program will be misleading. What we need is an
experimental approach,, vague enough to assure the creative contribution from
different sources, and the possibility to make change of directions and priority
as we learn from the progress of the actions.
To assure variety and flexibility one should
avoid a centralized approach and look for multipolar interventions.
Nevertheless, it is important that a proper climate be developed to alert on the
needs for multipolar interventions, to provide leverage effects on actions (no
matter where they come from), to assure a frame of reference for debate,
co-operation, stimulation.
With that in mind, we can underline the important
role that the EC will have in providing such climate to induce actions.
We should therefore propose that the EC will
launch an experimental program on globalization that will foster multipolar
initiatives, call for creative contribution from a multiplicity of actors (both
public and private), provide a starting frame of reference and the seeds for new
initiatives. The EC should try to experiment the role of the client for
globalization issues.
The EC experimental program - even if focused
on the S&T contribution - will itself be multipolar. We propose the
following types of intervention:
Play the role of the client by making use of
the current R&D Framework Programme to incentivate as much as possible the
attention to globalization issues, such as, e.g.:
- provide a financial premium for R&D projects that not only respond to each
action lines terms of reference, but also foresee the participation of LDC's
institutions or firms to the research ,
- add
a specific item on each line of action that calls for ideas and preliminary
investigations on problems having a globalization dimension.
Stimulate the emergence of new clients by
incentivating the convergence of public and private financing to set up Trusts
and Foundations, each one aiming at a specific objective or approach on
globalization matters, such as, e.g.:
- develop new products and production processes that respond to the local needs
and capacity in LDCs,
- incentive creativity and innovation in new products design (suited to new
needs coming from globalization issues) by means of instruments such as product
contests and prizes,
- assure the accessibility of advanced technology to design products (and related
manufacturing processes) for local needs, by favoring the concept of flexible
design using high technology components (both for products and processes) which
could be put together to local ones (hybrid technology solutions),
- promote and support networks of R&D institutions.
Enlarge the client dimension by developing
international initiatives (in co-operation with governments, GO's and NGO's) to
launch the first phase (problem identification and preliminary solutions
development) of S&T for major globalization issues, such as, e.g.:
- attack environmental global issues,
- exploit global commons
-develop low cost, socially acceptable shelter technology for massive inhurbated
area,
-experiment marginal agriculture technologies,
- develop a multimedia new encyclopedia of technology and know how's.
Act as a client in search of good ideas for
ill defined problems, by setting up an open Counter initiative to incentive
with financing any kind of unsolicited proposals that can contribute to develop
a portfolio of ideas and globalization issues.