The original publication is available at www.springerlink.

com

Constructive Research and Info-Computational

Knowledge Generation
Gordana DODIG CRNKOVIC

School of Innovation, Design and Engineering, Computer Science Laboratory, Mälardalen

University, Sweden [email protected]

Abstract. It is usual when writing on research methodology in dissertations and thesis work
within Software Engineering to refer to Empirical Methods, Grounded Theory and Action
Research. Analysis of Constructive Research Methods which are fundamental for all know-
ledge production and especially for concept formation, modeling and the use of artifacts is
seldom given, so the relevant first-hand knowledge is missing. This article argues for intro-
ducing of the analysis of Constructive Research Methods, as crucial for understanding of
research process and knowledge production. The paper provides characterization of the
Constructive Research Method and its relations to Action Research and Grounded Theory.
Illustrative example of Blue Brain Project is presented. Finally, foundations of Constructive
Research are analyzed within the framework of Info-Computationalism which provides
models of knowledge construction by information processing in a cognizing agent.

1. Understanding of Research Methodology in Computing on the Background of

Philosophy of Engineering and Philosophy of Science

Information age is fundamentally dependent on engineering whose knowledge largely rests

on the research within Computing / Informatics. Demonstrably, artefacts built on research
results of Computing work remarkably well and in that sense the research within Compu-
ting yields a reliable and trustworthy knowledge about the behavior of computational (en-
gineered) systems. One of the relevant methodological and epistemological questions is
what sort of knowledge is it and how it relates to our understanding of the world, including
our understanding of humans. It is of interest to know what aspects of Computing could be
labeled “Science” in a rigorous classical sense, and how knowledge produced within Com-
puting enriches both our understanding of the world and our potential for further learning
about the world and acting in the world, (see [1] and [2]).

As we are building Knowledge Society, our understanding of knowledge production, both

on the level of Cognitive Science and on the level of technological enhancements provided
by ICT is becoming strategically important. In this context a very general view of Compu-
ting as Science and Computing as Engineering is needed wherefrom also necessity follows
of understanding of Computing in both Philosophy of Science and Philosophy of Engineer-
ing contexts. In short, engineers have discovered the innovative potential of broad insights
into the discipline, which gives tangible comparative advantage. Knowing about underlying
mechanisms and structures for knowledge production helps us look out of the box and find
novel approaches, which is especially important in constructive and design research.
2

2. Constructive Research Based on Ontological Realism

The key idea of Constructive Research (or the Constructivist1 knowledge production), is the
construction, based on the existing knowledge used in novel ways, with possibly adding a
few missing links. The construction proceeds trough design thinking that makes projection
into the future envisaged solution (theory, artifact) and fills conceptual and other know-
ledge gaps by purposefully tailored building blocks to support the whole construction. Arti-
facts such as models, diagrams, plans, organization charts, system designs, algorithms and
artificial languages and software development methods are typical constructs used in re-
search and engineering. Constructivist solutions are designed and developed and not in the
first place discovered, even though a lot of inspiration for many artifacts comes from na-
ture, today especially evident in Natural and Organic Computing.

A construction, be it theoretical or a practical one, when it differs profoundly from anything

previously existing, constitutes a new reality against which preexisting one can be ex-
amined and understood, so it has an undeniable epistemological value. E.g. constructing of
a non-Euclidean geometry helps understand everyday Euclidean world as a special case.
Many theories are constructivist in nature, and Einstein is known for his constructive ap-
proaches to Physics with the emphasis on the creative side of theory building. Cognitive
Science in its Info-Computational approach to human mind is one of the best examples of
epistemological significance of constructive approaches in production of scientific know-
ledge.

The article will first describe Constructive Research Method, Constructivist Epistemology
and Constructivist Learning Theory and make necessary distinction between Constructiv-
ism and Action Research. The awareness of the significance of Constructive Research ap-
pears to be lacking in both research education, Research Methodology and Philosophy of
Computing.

To start with, let us first make some distinctions. The form of constructivism found in
sciences and engineering is a moderate constructivism or realist constructivism, based on
the realist ontological foundations. Realism in engineering consists in the conviction that
the world exists independently of the observer and that it is knowable. Or, in the words:

There is ample evidence that we still can adopt a critical realist outlook, even if
every part of our world view is a construction. Saalmann [3]

Constructivism assumes creation of knowledge through interaction between the observer

and the observed, and thus also recognizes the dependence of ontological constructs (all
which exists and can exist) on the ways of interaction of the observer with the world. As a
consequence, it also acknowledges social character of knowledge i.e. the fact that know-

1 The meaning of terms Constructivist and Constructionist vary vastly, and one of often
quoted distinctions (but far from the only one) is from Ackermann [4]. Also the content of
Radical Constructivism is highly ambiguous such that for example Maturana is sometimes
placed under Radical Realism, Critical Realism as well as Radical Constructivism.
 3

ledge production occurs in networks of interacting agents2. Even though taking into account
social aspects of construction of knowledge, Realist Constructivism does not go as far as
some Radical Constructivisms as to deny the relevance of reality as the ultimate authority:

Appealing to reality as the ultimate arbiter of (scientific) disputes gives rise to the
belief that there exists a mind-independent reality (MIR) which defines what is
true and what is not. Riegler [5]

Realist Constructivism insists on the existence of stable world with which we interact and
thus build increasingly rich knowledge about. The problem with the denial of the relevance
of reality is that it conflates two claims, the ontological and the epistemological one. Reali-
ty exists independently of any mind, but it is mind-dependent in the epistemological sense
and (re)constructed by cognizing agents through their interactions with the physical world
and with other agents. The requirement for reproducibility of experiments under corres-
ponding conditions (measurements, methods, interpretations) provides unambiguous and
repeatable connections between an observer and the observed reality. The fundamental fea-
ture of reality of natural phenomena including humans as natural beings, is its remarkable
stability, which is the basis of Realist Constructivism. That agrees with the following von
Foerster’s analysis:

The most we can say, therefore, is that the observer generates a description of the
domain of reality through his or her interactions (including interactions with in-
struments and through instruments), and that the observer can describe a system
of systems (a system of consensus) that leads to the emergence of systems that can
describe: observers. As a consequence, because the domain of descriptions is
closed, the observer can make the following ontological statement: The logic of
the description is isomorphic to the logic of the operation of the describing sys-
tem. von Foerster [6]

When it comes to Kant’s [7] dichotomy between “phenomena” (things as they appear
3
which constitute the world of common experience, which is illusion ) and “noumena”
(things in themselves, constituting a transcendental world to which we have no empirical
access, and which is reality) one can wonder: were X-rays in ancient Greece part of nuo-
mena? Isn’t the space of nuomena shrinking as we learn to observe the world in the do-
mains to which our common experience does not have access?

Historically, we learned that reality is a resource much richer than what one human epoch
can realize. Our knowledge about the world including ourselves is constantly growing. In
the reality of ancient Greece there were no cell phones, no quantum mechanics, and micro-
scopes, no quasars and DNA, no movies and TV, no process algebra, but the laws of me-
chanics and astronomy were the same and even human physiology and psychology, for all

2 In this context the distinction must be made between Social Constructionism which con-

cerns the phenomena related to social contexts (sociological constructs) and Social Con-
structivism with focus on individuals knowledge within a social context (psychological or
cognitive construct).
3Later on I will give the example of the observation of a very distant quasar that appears to
us as it was millions of years ago.
4

we know. If we look back and see how all those new discoveries and inventions came into
being, we find that it is through efforts of generations of scientists and engineers who
worked under assumptions of the mind-independent existence of knowable reality. Even
though our interactions and new conceptualizations change our relationships with the
world, all the constructive work of research is always constrained by the laws of the world
itself through its elements that exist independently of mind and from which all the con-
structing work starts.

Devices and model systems are what socio-cultural studies of science refer to as
the “material culture” of the community, but are also what cognitive studies of
science refer to as “cognitive artifacts” participating in the representational, rea-
soning, and problem-solving processes of a distributed system. Our data lead to
their interpretation as cognitive-cultural artifacts. They are representations and
thus play a role in model-based reasoning and problem-solving; they are central
artifacts around which social practices form; they are sites of learning; they con-
nect one generation of researchers to another; they perform as cultural ratchets (..)
in an epistemic community (..), enabling researchers to build upon the results of
the previous generations, and thus move the problem solving forward. Nercessian
[8]

In a similar way as physical reality presents fundamental constraints for knowledge con-
struction, a community of practice (epistemic community) puts additional constraints
through the process of interaction. Indeed, scientific work is in many ways defined by the
interactions between members of a research community, but its role is often misunderstood
by postmodern critics. Science is not just another narrative or just a different myth. It is not
less stringent because it develops as a result of a collective effort, just on the contrary. It is
a strictly regulated social system for knowledge production, and the result is, even though
not absolute, still our best existing knowledge.

3. Characteristics of Constructive Research

The whole activity of manipulation is devoted to build various external epistemic

mediators that function as an enormous new source of information and know-
ledge. Therefore, manipulative abduction represents a kind of redistribution of the
epistemic and cognitive effort to manage objects and information that cannot be
immediately represented or found internally (for example exploiting the resources
of visual imagery). If we see scientific discovery like a kind of opportunistic abili-
ty of integrating information from many kinds of simultaneous constraints to pro-
duce explanatory hypotheses that account for them all, then manipulative abduc-
tion will play the role of eliciting possible hidden constraints by building external
suitable experimental structures. Magnani [9]

Constructive research method implies building of an artifact (practical, theoretical or both)

that solves a domain specific problem in order to create knowledge about how the problem
can be solved (or understood, explained or modeled) in principle. Constructive research
gives results which can have both practical and theoretical relevance. The research should
solve several related knowledge problems, concerning feasibility, improvement and novel-
ty. The emphasis should be on the theoretical relevance of the construct. What are the ele-
ments of the solution central to the benefits? How could they be presented in the most con-
densed form? [10] [11]
 5

4. Research within Computing

In the beginning, Computing was accepted with skepticism among sciences but successive-
ly it is becoming increasingly important and presently in many respects it is seen as the pro-
totype of an Ideal Science, thus replacing Physics [1]. The reason is the ability of Compu-
ting to provide a common language for not only sciences but also other forms of
scholarship and arts. It tends to reflect the totality of human interests as information mi-
grates from the physical world into the virtual world of computer networks.

No other field today presents such a lingua franca for all of our knowledge and agency as
Computing. Concepts and paradigms from Computing are spreading “like a wild fire”
(Cantwell Smith) through other fields. As Computing is a very wide area, including com-
puter Science, Computer Engineering, Software Engineering and Information Systems, in
what follows, we will concentrate on Software Engineering, SE. Classification of the issues
of the SE discipline into Engineering and Scientific was proposed by [12] and [13] with
two different kinds of objects of study: problems of building new software artefacts in En-
gineering, and the theoretical basis of the SE in Science.

Therefore, Engineering research differs greatly from traditional scientific research

because while Sciences deal with the study of existing objects and phenomena, be
it physically, metaphysically or conceptually, Engineering is based on how to do
things, how to create new objects. Lázaro and Marcos [12]

In the above sense, constructive research is what suits both Engineering and Science, the
former in the creation of new objects from the existing ones and the latter in the formation
of new concepts and models.

5. Software Engineering Research

The received view is that methodology of engineering as a research field and Software En-
gineering in particular is fundamentally constructive. Most Software Engineering research
groups work predominantly in 'construction mode', inventing new models and tools. Never-
theless, little effort has been invested into understanding of mechanisms of knowledge pro-
duction in Constructive Research by means of distributed cognition and use of cognitive ar-
tifacts.

Constructive research takes off from the existing well understood ground and that is why
research in Software Engineering often starts with empirical investigations where quantita-
tive (Controlled experiment, Survey) or qualitative (Grounded Theory, Case studies) me-
thods are used prior to the constructive work. Only when sufficient understanding of the re-
search problem and the domain is obtained, one can start addressing a Software
Engineering problem by Constructive Research method.

Both Grounded Theory and Action Research can be applied to the phase in engineering re-
search/ design research/ economic research when artifact which is central for the first de-
sign or prototype phase is already done, when a new construct is employed in a social con-
text (group of users applying a programming or theoretical tool or other artifact previously
developed).
6

However the main knowledge production within engineering usually happens in the con-
struction phase of the research. The major task of an engineer or a scientist who constructs
an artifact is the process of construction itself, and it needs to be studied with the same rigor
and should be given at least as much attention from the methodological and Philosophy of
Science point of view as Grounded Theory and Action Research perspective of the same re-
search project which address the implementation of the artifact (construct).

The idea of predominantly constructive nature of engineering exists but it is not clearly un-
derstood and the common view often identifies general methodology of Constructivism
with radical post-modernist Social Constructivism which is not well received among scien-
tists and engineers who take it to be hostile to sciences and engineering. Moreover, the role
played by constructive research is an under-researched topic in the Theory of Science. New
studies such as Nercessian [8] are important for recognition of research practices with con-
structive approaches and it is evident that a lot of work remains to be done. On the level of
Philosophy and Methodology of Science, understanding is needed of what sort of know-
ledge can be produced by constructive methods and how it relates to knowledge produced
by other research traditions, what the contribution of a constructive knowledge is and how
it is justified.

6. Design Research as Constructive Research

Design research is often present as an important part of research within Engineering, Com-
puter Science and Information Systems. It involves the analysis of the use and performance
of designed artifacts (constructs) in order to understand, explain and improve designed sys-
tems. The outputs of Design Research are constructs, models, methods, theories, instantia-
tions, algorithms, human-computer interfaces, system design methodologies, languages and
other artifacts, [14].

Design Research as constructive research presents a bridge between natural and human
spheres as it produces artifacts which are both natural and intentional. That implies under-
standing both of the workings of basic mechanisms (as found in sciences) and the role
which a given construct may play in the broader context (societal aspects).

7. Constructive Research vs. Action Research

Constructivist epistemology emphasizes the fact that scientific knowledge is constructed by

scientists with help of cognitive tools. It is the opposite of the positivist epistemology which
sees scientific knowledge as discovered in the world. For a classical positivist, scientific
facts are discovered and the connection between the world and the fact is unique. On the
other hand constructivism entails that there is no single valid methodology for construction
of scientific knowledge, so no unique prescription to establish “the facts” or provide the da-
ta, and no guarantee for a consensus. One can say that constructivism is more interested in
the mechanisms of theory building [15] while positivism describes the steady state of
theory where one dominant framework has been established among competing approaches.

Often mentioned is also the alleged opposition between Constructivism and Realism. Both
come in two flavors – ontological and epistemological. The confusion among varieties of
claims is vast. In short, constructivist approaches are primarily applied in the process of
discovery where many possibilities are still open, in the sense of ontological choices, dur-
 7

ing concepts formation [16] and in the sense of epistemological approaches. In the new re-
search, the reality appears as malleable and negotiable as an uncharted territory. Similar to
the process of reinforcement learning in the brain, learning in the network of agents causes
that those paths that are taken repeatedly get reinforced, and those that are not followed
fade out – so finally in a long time perspective some approaches win and some get forgot-
ten. That is a territory of steady state in which it can appear as if there is only one clear-cut
approach to the phenomenon or a solution to the problem. As already mentioned, a received
view in sciences and engineering is ontological realism in conjunction with epistemological
constructivism.

8. Constructivism vs. Constructionism

Among numerous terminological confusions, the one around the distinctions between
Constructivism vs. Constructionism is often found. We follow Ackermann who succinctly
characterizes the two. Unlike some other authors she uses both terms when it comes to
research practices as well as learning theories.

Piaget’s constructivism offers a window into what children are interested in, and
able to achieve, at different stages of their development. The theory describes how
children’s ways of doing and thinking evolve over time, and under which cir-
cumstance children are more likely to let go of—or hold onto— their currently
held views. Piaget suggests that children have very good reasons not to abandon
their worldviews just because someone else, be it an expert, tells them they’re
wrong. Papert’s constructionism, in contrast, focuses more on the art of learning,
or ‘learning to learn’, and on the significance of making things in learning. … In-
tegrating both perspectives illuminates the processes by which individuals come
to make sense of their experience, gradually optimizing their interactions with the
world. Ackermann [4] (emphasis mine)

9. Action Learning and Action Research

Action learning is a method where learners improve their knowledge and skills through ac-
tions, it is learning-by-doing and by means of examples and exercises. Often, action learn-
ing is performed in small groups (learning sets). In the same way as Constructive research
is related to Constructive learning, Action research is related to Action learning.

action-research, a comparative research on the conditions and effects of various

forms of social action and research leading to social action that uses “a spiral of
steps, each of which is composed of a circle of planning, action, and fact-finding
about the result of the action. Lewin [17]

What distinguishes this type of research from general professional practices, consulting, or
ordinary problem-solving is the emphasis on scientific approach which includes systematic-
ity and interventions based on theoretical considerations. During the process, researcher is
refining the methodological tools while collecting, analyzing, and presenting data on an on-
going basis simultaneously getting concerned people involved into research. The research
takes place in real-world situations, and aims to solve real-world problems so it has a social
dimension too. Action researchers do not try to remain objective, but recognize their bias to
the other participants. Oquist [18] analyses epistemological positions of Action Research
8

and presents the reasons for its rejection by Empiricism, Logical Positivism and Structural-
ism and the reasons for its acceptance by Pragmatism.

Action Research may be one of the tasks of an engineer or scientist constructing an artifact.
It may be used for validation of the results, and in the study of the research processes them-
selves. See for example Shaw [19] who exemplifies validation of a construct in Software
Architecture Research. Benavides et al. [20] suggest using Action Research in Software
Engineering as a method of resolving what they call “the triple schizophrenia of the Soft-
ware Engineering researcher” – their split between researching, teaching and learning activ-
ities. This however represents a meta-level with respect to knowledge generation studies of
constructive research mechanisms.

10. Grounded Theory

Borgatti defines Grounded Theory as a method of using empirical data by inductive struc-
turing of observations without preconceived theories, in contrasts to theory derived deduc-
tively from the existing theories.

Constant comparison is the heart of the process… Theory emerges quickly. When
it has begun to emerge you compare data to theory.
http://scu.edu.au/schools/gcm/ar/arp/grounded.html

Notice that the above process of structuring of data is inherently constructive. In sciences
and engineering this process is present in a construction of a model or an artifact, and is
seldom done ex nihilo. Typically there is a starting point in the existing background know-
ledge, models and artefacts which present important constraints in the process of construc-
tion. In the research process of the design of an artifact of central interest is how this de-
signed construction evolves and how the artifact relates to its users.

11. Computational Models and Simulations as Knowledge Construction Tools

Mathematical modeling and numerical simulation are extensions of the traditional empirical
and experimental approaches. They provide effective ways for virtual experimentation
when real experiments are impractical or just too expensive. They enable study of the exist-
ing data, identify critical areas where data are missing, facilitate hypothesis generation, and
similar. Simulations can be used to predict the behavior of the system for specific choices
of parameters and/or initial conditions, [21].

An experiment with an artifact model provides an extension of the scientific prac-

tice of thought experimenting (…) – more complete and complex and potentially
less subject to individual bias and error. It affords a more detailed representation,
a wider range of manipulations, and more control than is possible for a thought
experiment. In a manner similar to how microscopes and telescopes extend scien-
tists’ capacity to see, artifact simulation models extend their biological capacity to
perform simulative reasoning. As with thought experimenting, the model system
simulation contributes to predictions about how the in vivo system might perform
under specified conditions. Nersessian [8]

Best computational methods are results of an exhaustive theoretical analysis. At the same
time, the analysis of results is similar to the analysis of experimental data. Computational
 9

methods/simulations in science are very useful when the problem is too difficult to handle
analytically, an approximate theoretical result might not be reliable or an experiment is ex-
pensive or not feasible to perform. Ultimately it is of course experiment which decides on
the nature of the real-world phenomena, but computational methods present very effective
cognitive tools.

The problem-solving practices of engineering scientists provide an excellent re-

source for advancing the agenda of environmental perspectives because their si-
mulation technologies are simultaneously cognitive and cultural artifacts. Within
the confines of this paper attention is directed primarily to one aspect – the func-
tion of physical models as cognitive artifacts in distributed cognitive processes –
but in fact the socio-cultural dimensions are always present if not attended to di-
rectly in this brief analysis. The epigraph quote from Daniel Dennett provides a
pithy summary of a major premise of range of research under headings such as
“distributed cognition,” “the extended mind thesis,” “situated cognition,” and “ac-
tivity theory.” (…) Within this framing, to understand how problem solving is
achieved requires examining the generation, manipulation, and propagation of sa-
lient representations within the system; that is, examining how representational
states flow across media and accomplish cognitive work.” Nersessian [8]

The central claim is that reasoning in this distributed cognition framework involves co-
processing of information in human memory and in the environment, where artifacts are in-
tegrated into cognitive processing.

12. Research as Learning. Knowledge Generation as Information Processing

Furthermore, all researchers are learners. As we have found, the researchers and
their technologies have intersecting developmental trajectories. Nersessian [8]

The Constructive Research Paradigm we suggest as closest to actual practices in problem

solving and research within natural sciences and engineering is Realist Constructivism, or
Grounded Constructivism, grounded in a sense that it ontologically presupposes the mind-
independent existence of reality. (N.B. that is not to say that Grounded Constructivism pre-
supposes existence of mind-independent reality which is a different and wrong claim.) In
sum, Realist Constructivism is based on ontic realism and epistemic constructionism.

Following are characteristics of Realist Constructivism:

ƒ The world exists and is knowable.

ƒ All knowledge is embodied in agents.

ƒ Knowledge is constructed via computational processing of information.

ƒ Reality is informational in nature (Informational Structural Realism, Floridi)

ƒ Dynamics of information is computation.

10

13. Info-Computationalist View of Knowledge Generation and Constructive Re-

search Paradigm

Within Info-computational naturalism [22] knowledge is seen as a result of successive

structuring of data, where data are understood as simplest information units, signals ac-
quired by a cognizing agent through the senses/sensors/instruments [23]. Information is
meaningful data, which can be turned into knowledge by an interactive computational
process going on in the agent. Potential information exists in the world even if no cognizing
agents are present. The world (reality) for an agent presents potential information, both out-
side and within an agent. Knowledge, on the other hand, always resides in a cognitive
agent. The interplay of an agent with the environment which leads to adaptation, clearly
presents developmental advantage which increases agent’s ability to cope with the changing
environment.

An agent is a physical system, living organism or a robot possessing cognitive structures

capable of adaptive behaviors. Living agents evolved successively starting with prebiotic
evolution of increasingly complex molecules able to adapt to its environment and repro-
duce. Semantics of information is relative to an agent. Meaning that proto-information will
take for a biological organism is in the use the agent (organism) has for it. Information is
always embodied in a physical signal, molecule, particle or event which will induce change
of structure or behavior in an agent. Semantics develops as data /information/knowledge
structuring process, in which complex structures are self-organized by the computational
processing from simpler ones. Meaning of information is thus defined for an agent and a
group of agents in a network (ecology) and it is given by the use information has for them.
Knowledge generation as information processing in biological agents presupposes natural
computation, defined by MacLennan [24] as computation occurring in nature or inspired
by that in nature, which is the most general current computing paradigm.

Taking information and computation as basic principles in a dual-aspect Info-

Computational model, a common framework is formulated in [22] [25], providing the ar-
gument that knowledge generation should be understood as natural computation. Know-
ledge is a characteristic of intelligent agents. Intelligence in a living organism as well as its
cognitive capability is a matter of degree. Artificial intelligent agents can be ascribed know-
ledge as long as they function as agents, exchanging information with the world, in analogy
with an organism. In the same way as we do not ascribe knowledge to a dead living organ-
ism, we do not ascribe knowledge to a non-functional robot. Information, on the other hand
can be stored in non-living systems, but for information to be knowledge, process of inte-
raction between different pieces of information is necessary.

When we talk about computation in living organisms, models of computation beyond Tur-
ing machine are needed to handle complexity of phenomena such as creation, maintenance
and reproduction of cognizing agency, perception, learning, intelligence, consciousness, so-
cial phenomena, etc. Nowadays, the source of inspiration for new models of computing is
found in natural processes, and especially in organic computing seen as evolutionary devel-
opment of intelligent self-organized processes and structures necessary for living organisms
to survive in the complex and dynamical world.

All of our knowledge of the world is based on information we gain from the world, be it
from direct experience (interaction with the world) or by learning (getting directly or indi-
rectly information) from other people. Physicist Zeilinger [26] suggests possibility of see-
ing information and reality as one. This is in accord with Informational Structural Realism
 11

which says that reality is made of informational structures [27] [28] as well as with info-
computational ontology [25] based on Informational Structural Realism (ISR). What Floridi
assumes to be mind-independent data corresponds to proto-information (information in the
world, world as information) of Info-computationalism. Reality is informational and mind-
independent and consists of structural objects, which brings together metaphysical views of
Wiener (“information is information, not matter or energy”) and Wheeler (“it from bit”).

14. The Computing Universe

The complementary basic concept of information is computation, which is the dynamics of

informational structure. Floridi’s Informational Structural Realism with information as the
fabric of the universe, has as a logical consequence that the process of dynamical changes
of the universe makes the universe a huge computational network. Here is how Chaitin de-
scribes the computing universe:

And how about the entire universe, can it be considered to be a computer? Yes, it
certainly can, it is constantly computing its future state from its current state, it's
constantly computing its own time-evolution! And as I believe Tom Toffoli
pointed out, actual computers like your PC just hitch a ride on this universal com-
putation! Chaitin [29]

Within Info-computationalism, computation is both symbolic and sub-symbolic4 informa-

tion processing, such as natural computation or organic computation. As it corresponds to
the dynamic of processes that exist in the universe, it is necessarily both discrete and conti-
nuous. In the Info-computational naturalist framework, information and computation are
two fundamental and inseparable elements necessary for naturalizing meaning, cognition
and mind, Dodig Crnkovic [22].

In sum: information is the structure, the fabric of reality. The world exists independently
from us (realist position of structural realism) in the form of proto-information or ur-
(in)formation, the richest potential form of existence corresponding to Kant’s Ding an sich.
That proto-information becomes information (“a difference that makes a difference” ac-
cording to Bateson) for a cognizing agent in a process of interaction through which specific
aspects of the world get uncovered. In science this process of successive transformation of
manifold potential proto-information in the world into particular actual information in an
agent becomes evident when for example observing the same physical object in different
5
wavelengths . The world as it appears to an agent is dependent on the type of interaction
through which the agent acquires information. Potential information in the world is ob-
viously much richer than what we observe, containing invisible worlds of molecules, atoms
and sub-atomic phenomena, distant cosmological objects and similar. Our knowledge about
this proto-information which reveals with help of scientific instruments will surely continue

4 Sub-symbolic computations go on in neural networks.

5 Results of observations of the same physical object (celestial body) in different wave-
lengths (radio, microwave, infrared, visible, ultraviolet and X-ray), see
http://bb.nightskylive.net/asterisk/viewtopic.php?f=8&t=14561
12

to increase with the development of new devices i.e. the new ways of interaction with the
world, [2].

Computation and information are the backbone of sciences. They are inseparable – there
can be no structure without a process (a structure actualizes through a process) and the vice
versa, there is no process without a structure to act on, [22] [23]. Formalization of info-
computational approach within category theory may be found in Burgin [30].

15. Information and Computation in Biological and Intelligent Artificial Systems

Recent studies in biology, ethology and neuroscience, which have increased our knowledge
of biological cognitive functions, have led to the insight that the most important feature of
cognition is its ability to deal efficiently with complexity, [31]. Insights into natural intelli-
gence, together with the increase in power of electronic computing bring us closer to the
modeling of intelligent behavior. Artificial intelligence (AI) is based on the belief that intel-
ligent behavior can be understood in such a way that a machine can be constructed able to
simulate it, [23]. In the same paper it is argued that Info-computationalism is the most ap-
propriate theoretical framework for understanding of the phenomena we call intelligence.
From the computationalist point of view intelligence may be seen as based on several levels
of data processing [32] [33] in a cognizing agent. As AI attempts to construct non-living in-
telligent agents, we include even AI agents in the discussion. Information (produced from-
sensory data processed by an agent) can be understood as an interface between the data
(world) and an agent’s perception of the world, [34]. Patterns of information should thus be
attributed both to the world and to the functions and structures of the brain. In an analogous
way, knowledge can be understood as an interface between perception and cognition. Struc-
tures of knowledge are related both to percepts (sensory information) and to the brains cog-
nitive functions and organization. Meaning and interpretation are the results of the
processes of temporal development of information, and its refinement by relating to already
existing memorized information. The meaning of an object in the external world is recog-
nized through the process of perception of sensory signals, their processing through nerv-
ous system, comparison with memorized objects, anticipation from memorized experiences
of possible scenarios in which the object played different roles, etc.

Data, information, perceptual images and knowledge are organized in a multiresolutional

(multigranular, multiscale) model of the brain and nervous system, [32]. Multiresolutional
representation has proven to be a good way of dealing with complexity in biological sys-
tems, and they are also being implemented in AI [33].

Cognitive robotics research presents us with a sort of laboratory where our understanding
of cognition can be tested in a rigorous manner. From cognitive robotics it is becoming evi-
dent that intelligence is closely related to agency. Anticipation, planning and control are es-
sential features of intelligent agency. A similarity has been found between the generation of
behavior in living organisms and the formation of control sequences in artificial systems.

Self-organization, self-reproduction and self-description (or self-representation) are funda-

mental intrinsic properties of natural intelligent systems. More and more of intelligent ro-
botics builds on similar basic principles called organic computing. Learning is an essential
part of each of the above three capabilities and it requires among others the development of
a symbolic system which is easy to maintain and use. It is possible to build intelligent con-
trol systems that can collect and process information, as well as generate and control beha-
 13

vior in real time, and cope with situations that evolve among the complexities of the real
world.

In intelligent biological systems based upon a hierarchy of functional loops, each of these
loops can be treated as a control system per se [32]. Generation of structures resulting from
sensory processes (data), and information organized into knowledge to be used in decision
making are built in a multiresolutional way, with many pattern recognition and control me-
chanisms hardwired.

Along with the above constructive approach to AI with roots in engineering, logic and pro-
gramming, there is a different (also constructive) one which tries to implement in silico in-
fo-computational biological functions translating biological structures starting on molecular
level – notably the Blue Brain Project of ETF which will be discussed later on. Common to
both approaches is the belief that intelligent behavior can be implemented in a machine
(non-biologically). Up to now there are only specific aspects of intelligence (weak AI) that
machines can possess, while general AI seems still a far-reaching goal, after more than fifty
years of research.

16. Knowledge Generation as Natural Computation

Present day’s computers perform syntactic mechanical symbol manipulation which in the
future has to be extended to include information processing with semantic aspects. Burgin
[35] identifies three distinct components of information processing systems: hardware
(physical devices), software (programs that govern its functioning), and infoware which
represents information processed by the system. Infoware is a shell built around the soft-
ware-hardware core which is the traditional domain of automata and algorithm theory. Se-
mantic Web is an example of infoware. Bio-computing and especially Organic computing
are expected to substantially contribute to the infoware by adopting strategies for handling
of complexity found in the organic world.

Classical theoretical models of computers are mathematical objects equivalent to abstract

automata (Turing Machines) which are equivalent to algorithms/effective procedures, re-
cursive functions, or formal languages. Turing Machine is a logical device, a model for ex-
ecution of algorithms. However, if we want adequately to model biological structures and
processes understood as embodied physical information processing, highly interactive and
networked computing models beyond Turing Machines are needed. In order to develop
general theory of the networked physical information processing, we must also generalize
the ideas of what computation is and what it might be. For new computing paradigms, see
[36], [37], [38] and [22]. Compared with new computing paradigms, Turing machines form
the proper subset of the set of information processing devices.

Naturalized epistemology according to [39] [40] is an idea that knowledge should be stu-
died as a natural phenomenon. The subject matter is not our concept of knowledge, but the
6
knowledge itself as it appears in the world . An evolving population incorporates informa-

6 Maturana was the first to suggest that knowledge is a biological phenomenon. He argued
that life should be understood as a process of cognition which enables an organism to adapt
and survive in the changing environment.
14

tion about its environment through changes by natural selection, as the state of the popula-
tion matches the state of the environment. The variants which adapt and survive in the envi-
ronment increase in frequency in the population. Harms [39] proves a theorem showing that
natural selection will always lead a population to accumulate information, and so to 'learn'
about its environment. Okasha [41] points out that

any evolving population 'learns' about its environment, in Harms' sense, even if
the population is composed of organisms that lack minds entirely, hence lack the
ability to have representations of the external world at all.

That is correct, and may be seen not as a drawback but as strength because of the generality
of naturalistic approach. It shows how mind is a matter of degree and how it slowly and
successively develops with evolution. There is fresh evidence that even simple 'lifeless'
prion molecules are capable of evolutionary change and adaptation [43].

However, this understanding of basic evolutionary mechanisms of accumulating informa-

tion at the same time increasing information processing capacities of organisms (such as
memory, anticipation, computational efficiency) is only the first step towards a full-fledged
evolutionary epistemology, but the most difficult and significant one. From bio-computing
we learn that in living organism biological structure (hardware) is at the same time a pro-
gram (software) which controls the behavior of that hardware.

By autopoetic process [42] [43], biological system changes its structures and thus the in-
formation processing patterns in a self-reflective, recursive manner. Natural selection of or-
ganisms, responsible for nearly all information that living systems have built up in their ge-
notypes and phenotypes, is a simple but costly method to develop knowledge capacities.
Higher organisms (which are “more expensive” to evolve) have grown learning and reason-
ing capability as a more efficient way to accumulate knowledge. The step from “genetic
learning” (typical of more primitive life forms) to acquisition of cognitive skills on higher
levels of organization of the nervous system (such as found in vertebrata) will be the next
step to explore in the project of naturalized epistemology.

17. Why Our Perception of the World is an Illusion

Of all information processing going on in our bodies, perception is only a tiny

fraction. Our perception of the world depends on the relative slowness of con-
scious perception. Time longer than one second is needed to synthesize conscious
experience. At time scales shorter than one second, the fragmentary nature of per-
ception reveals. The brain creates a picture of reality that we experience as (and
mistake for) “the actual thing” Ballard [45]

As already mentioned, Kant argued that “phenomena” or things as they appear and which
constitute the world of common experience are an illusion. Kaneko and Tsuda discuss why.

Hence the brain does not directly map the external world. From this proposition
follows the notion of “interpreting brain”, i.e. the notion that the brain must in-
 15

terpret symbols generated by itself even at the lowest level of information

processing. It seems that many problems related to information processing and
meaning in the brain are rooted in the problems of the mechanisms of symbol
generation and meaning. Kaneko and Tsuda [44]

Consciousness provides only a rough sense of what is going on in and around us, in the first
place what we take to be essential for us. The world as it appears for our consciousness is a
sketchy simulation. Belief that we ever can experience the world 'directly' is the biggest il-
lusion [46].

What would that mean anyway to experience the world 'directly as it is', without ourselves
being part of the process? Who would experience that? It is important to understand that, as
Kaneko and Tsuda [44] emphasize, the brain maps the information about the (part of the)
world into itself, but the mapped information is always affected by the activity of the brain
itself. This seems to be the view of Maturana [47] as well. The question of what is reality
“an sich” in the sense of proto-information and understanding of our interactions with the
world outside through the (conscious and sub-conscious) exchange of information is fun-
damental, and we have to keep searching for a deepened understanding of that reality in re-
lation to ourselves. To that end, the awareness of the presence of models (simulations) in
understanding (and perception) of the world is essential.

18. Science. The World and a Model. Real and Virtual

If what we perceive of the world is a simulation our brain plays for us in order to manage
complexity and enable us to act efficiently, then our understanding of the world must also
be mediated by this modeling nature of cognition. Not even the most reliable knowledge
about the physical world as it appears in sciences is independent of the modeling frame-
works which indirectly impact what can be known. The common positivist optimism about
observations independent of the observer proved problematic in many fields of physics
such as quantum mechanics (wave function collapse after interaction), relativity (speed de-
pendent length contraction and time dilatation) and chaos (a minor perturbation sufficient to
switch to a different attractor). In general, observer and the systems observed are related
and by understanding their relationship we can gain insights into limitations and power of
models and simulations as knowledge generators [6].

Models are simplified representations, made for a purpose and they ignore aspects of the
system which are irrelevant to its purpose. The properties of a system itself must be clearly
distinguished from the properties of its models. All our knowledge of systems is mediated
by models. Engineers using different models often get so familiar with the model and its
functions that they frequently act as if the model was the actual reality itself [48].

Awareness of the modeling character of knowledge and the active role of the cognizing
agent in the process of generation of knowledge is specifically addressed by second order
cybernetics. Cybernetic epistemology is constructivist in recognizing that knowledge can-
not be passively transferred from the environment, but must be actively constructed by the
cognizing agent based on the elements found in the environment in combination with in-
formation stored in the agent. The environment eliminates inadequate models, which are
getting into conflict with it. Model construction thus proceeds through variation and selec-
tion. This agrees with von Glasersfeld’s [50] two basic principles of constructivism:
16

Knowledge is not passively received either through the senses or by way of com-
munication, but is actively built up by the cognizing subject.

The function of cognition is adaptive and serves the subject's organization of the
experiential world, not the discovery of an ‘objective ontological reality’.

Finally in words by Kaneko and Tsuda:

The key question that arises in this ‘constructive’ approach lies in the relationship
between the virtual world (the model) and reality. The virtual world should not
just be an imitation of reality, but a sort of abstraction from reality, and be con-
structed from our side by utilizing some abstracted essential features of reality.
Understanding the relationship between the virtual world and reality is a funda-
mental issue in the study of complex systems with a constructive approach. [44]

19. The Blue Brain Project as an Example of Constructive Research in Fundamen-

tal Science

The Blue Brain Project http://bluebrain.epfl.ch/ studies functions of the brain modeled as
consisting of neocortical columns, NCC which are structural units in the cerebral cortex.
Millions of NCCs compose the brain. At present stage Blue Brain Project is attempting to
accurately model a single NCC of a young rat brain. Project has succeeded to create an ar-
tificial column that responds to electrical impulses in the same way as the biological one.
The model of an NCC is built down to the molecular level and simulates the real-world
electrical activity of rat brain tissue. The Blue Brain Project group hopes eventually to use
the same simulation approach just scaled up to model the whole rat brain and ultimately
even the human brain. Currently the Blue Brain runs on IBM’s Blue Gene supercomputer.
The present computing power is just enough to simulate a single cortical column, while
modeling millions of columns is far beyond the scope of computational capacity of a Blue
Gene, so new solutions must be found to increase the computational power.

An interesting question is what phenomena besides simultaneous firing of neurons may be

expected to emerge in the simulation. It is speculated that among emergent properties of the
whole brain simulation consciousness will emerge when rat brain simulation is connected
in sensorimotor loops with a robotic rat. This presents an eminently constructivist program
for understanding consciousness and the brain itself by re-constructing it in silico.

The Blue Brain project is in many respects an example of constructive science at its best
with classical elements such as reproducibility, third person perspective and meticulously
performed reconstruction of biological neurons computationally simulated in a great detail
on the Blue Gene supercomputer.7 The big triumph of the Blue Brain project was a sponta-
neous firing of simulated neurons when given electrical stimuli.

The way knowledge is generated in Blue Brain Project is typically info-computational, with
simulation representing epistemological laboratory of a kind in which computational mod-

7The same constructive strategy in developing scientific understanding is described by Ka-

neko and Tsuda in the case of study of chaotic systems [44].
 17

els are tested against biological “reality”. The prominent role of computerized visualization
is also characteristics of Info-computationalism as well as a mix of knowledge from many
different disciplines – from Biology, Neuroscience to Computer Science implemented in in-
fo-computational tools. Among others an advanced patch clamp robot technique is used
contributing to huge effectivization by automatization of experimental procedure, through
combining functions of a microscope with computer-assisted reconstruction of neuron
structure.

20. Conclusions

When writing chapter on methodology in dissertations and thesis work within Software En-
gineering it is common practice to refer to Empirical Methods, Grounded Theory and Ac-
tion Research. Attention is seldom paid to the analysis of Constructive Research methods
which are fundamental for all engineering and sciences when it comes to concept forma-
tion, modeling and the use of artifacts. This article argues for the necessity of changing of
this practice and focusing on the analysis of constructive research practices, as they lie at
the core of the knowledge production.

Justification for the claims about necessary constructive character of all knowledge is found
in Info-Computational approach which provides models of knowledge generation on a fun-
damental level of information processing in a cognizing agent.

References

1. Dodig-Crnkovic G.: Shifting the Paradigm of the Philosophy of Science: the Phi-
losophy of Information and a New Renaissance. Minds and Machines 13 (2003),
521-536 http://www.springerlink.com/content/g14t483510156726

2. Dodig-Crnkovic G. and Müller V: A Dialogue Concerning Two World Systems:

Info-Computational vs. Mechanistic. In: G Dodig-Crnkovic and M Burgin eds:
Information and Computation, World Scientific Publishing Co. Series in Infor-
mation Studies, forthcoming (2010)

3. Saalmann G.: Arguments Opposing the Radicalism of Radical Constructivism,

Constructivist Foundations 3 (2007) 16–18

4. Ackermann E.: Piaget's constructivism, Papert's constructionism: what's the dif-

ference? In: Constructivism: uses and perspectives in education, Vol 1 and 2,
Conference Proceedings, Geneva, Research Center in Education (2001) 85-94

5. Riegler, A.: Towards a Radical Constructivist Understanding of Science. Founda-

tions of Science 6(1) (2001) 1-30

6. von Foerster, H.: Understanding Understanding: Essays on Cybernetics and

Cognition, Springer (2003)

7. Kant, I.: Critique of pure reason. New York: St Martin’s Press. (1965)
18

8. Nersessian, N.J.: How do engineering scientists think? Model-based simulation

in biomedical engineering laboratories, Topics in Cognitive Science, 1:3, in press
(2009)

9. Magnani L.: Model-Based and Manipulative Abduction in Science, Foundations

of Science, 9 (2004) 219-247

10. Lukka, K.: The constructive research approach. In Ojala, L. & Hilmola, O-P.
(eds.) Case study research in logistics. Publications of the Turku School of Eco-
nomics and Business Administration, Series B 1: (2003) 83-101

11. Kasanen, E., Lukka K. and Siitonen A.: The Constructive Approach in Manage-
ment Accounting Research. Journal of Management Accounting Research, 5 (1)
(1993) 243-263

12. Lázaro M., Marcos, E.: Research in Software Engineering: Paradigms and me-
thods. In: J Castro, E Teniente, editors, Advanced Information Systems Engineer-
ing, 17th International Conference, CAiSE 2005, Porto, Portugal, Proceedings of
the CAiSE 05 Workshops, 2 (2005) 517-522

13. Lavrishcheva E. M.: Software Engineering As A Scientific And Engineering Dis-

cipline, Cybernetics and Systems Analysis. 44 (2008)

14. Vaishnavi, V. and Kuechler, W.: “Design Research in Information Systems”

http://ais.affiniscape.com/displaycommon.cfm?an=1&subarticlenbr=279 (2004)

15. Magnani, L., Nersessian, N., and Thagard, P. eds.: Model-based reasoning in
scientific discovery. New York: Plenum. Chinese translation published by China,
Science and Technology Press (2000)

16. Thagard, P.: Conceptual revolutions. Princeton University Press (1992)

17. Lewin, K.: Resolving social conflicts; selected papers on group dynamics. Ger-
trude W. Lewin (ed.). New York: Harper & Row (1948)

18. Oquist, P.: The Epistemology of Action Research. Acta Sociologica, 21, (1978)
143-163

19. Shaw, M.: The Coming-of-Age of Software Architecture Research. Proceedings

of ICSE-2001, Los Alamitos, CA: IEEE Computer Society Press. (2001) 657-664

20. Benavides D., et al. : The Triple Schizophrenia of the Software Engineering Re-
searcher. In: J Castro, E Teniente, edts, Advanced Information Systems Engineer-
ing, 17th International Conference, CAiSE, Porto, Portugal, Proceedings of the
CAiSE 05 Workshops, 2 (2005) 529-534

21. Stauffer, D., de Oliveira, S.M., de Oliveira, P.M.C. and Sa Martins, J.S.: Biology,
Sociology, Geology by Computational Physicists. (Monograph Series on Nonli-
near Science and Complexity). Elsevier, Amsterdam. (2006)

22. Dodig-Crnkovic G.: Information and Computation Nets, VDM Verlag (2009)
 19

23. Dodig-Crnkovic, G.: Knowledge Generation as Natural Computation, Journal of

Systemics, Cybernetics and Informatics. 6 (2008)

24. MacLennan, B.: Natural computation and non-Turing models of computation.

Theoretical Computer Science 317 (2004) 115 – 145

25. Dodig-Crnkovic, G.: Investigations into Information Semantics and Ethics of

Computing, Mälardalen University Press (2006)

26. Zeilinger A.: The message of the quantum. Nature 438 (2005) 743

27. Floridi, L.: Against digital ontology. Synthese 168 (2009) 151-178

28. Floridi, L.: A Defence of Informational Structural Realism. Synthese, 161 (2008)
219-253

29. Chaitin, G.: Epistemology as Information Theory, In Computation, Information,

Cognition – The Nexus and The Liminal, Dodig-Crnkovic G. and Stuart S., eds.
Cambridge Scholars Publishing. (2007) 2-18
http://www.cs.auckland.ac.nz/~chaitin/ecap.html

30. Burgin, M.: Information Dynamics in a Categorical Setting. In Dodig-Crnkovic

G. and Burgin M. eds.: Information and Computation. World Scientific Publish-
ing Co. Series in Information Studies, forthcoming (2010)

31. Gell-Mann, M.: The Quark and the Jaguar: Adventures in the Simple and the
Complex. Owl Books. (1995)

32. Minsky M.: Interior Grounding, Reflection, and Self-Consciousness. In Dodig-

Crnkovic G. and Burgin M. eds.: Information and Computation. World Scientific
Publishing Co. Series in Information Studies, forthcoming (2010)

33. Goertzel, B.: The Evolving Mind. Gordon and Breach (1993)

34. Hoffman D.: The Interface Theory of Perception: Natural Selection Drives True
Perception to Swift Extinction. In: Object Categorization: Computer and Human
Perspectives. Cambridge University Press (2009)

35. Burgin, M.: Super-Recursive Algorithms. Springer Monographs in Computer

Science (2005)

36. Wegner, P.: Interactive Foundations of Computing. Theoretical Computer

Science 192 (1998) 315-51

37. Kampis G.: Self-Modifying Systems in Biology and Cognitive Science: A New
Framework for Dynamics, Information and Complexity. Pergamon Press (1991)

38. Kurzweil, R.: The Singularity is Near. New York: Viking (2005)

39. Harms W. F.: Information and Meaning in Evolutionary Processes. Cambridge

University Press (2004)
20

40. Harms, W. F.: Naturalizing Epistemology: Prospectus 2006, Biological Theory 1

(2006) 23–24

41. Okasha, S.: Review of William F. Harms, Information and Meaning in Evolutio-
nary Processes. Notre Dame Philosophical Reviews 12 (2005)

42. Maturana, H. and Varela, F.: Autopoiesis and Cognition: the Realization of the
Living. In Cohen R. S. and Wartofsky M. W., eds.: Boston Studies in the Philos-
ophy of Science 42. (1980) Dordecht: D. Reidel Publishing

43. Maturana, H. and Varela, F.: The Tree of Knowledge. Shambala (1992)

44. Kaneko K. and Tsuda I.: Complex systems: chaos and beyond. A constructive ap-
proach with applications in life sciences. Springer, Berlin/Heidelberg (2001)

45. Ballard D.: Our perception of the world has to be an illusion, Journal of Con-
sciousness Studies, 9 (2002) 54-71

46. Nørretranders T. : The User Illusion: Cutting Consciousness Down to Size. New
York: Viking (1999)

47. Maturana, H.: Systemic versus genetic determination. Constructivist Foundations

3 (2007) 21-26

48. Heylighen F. and Joslyn C.: Cybernetics and Second Order Cybernetics. In: Mey-
ers R.A. ed.: Encyclopedia of Physical Science & Technology. 4 (3rd ed.). Aca-
demic Press. New York. 155-170 (2001)

49. Scripps Research Institute (2010, January 3). 'Lifeless' prions capable of evolutio-
nary change and adaptation. ScienceDaily. Retrieved January 12, 2010, from
http://www.sciencedaily.com/releases/2009/12/091231164747.htm

50. von Glasersfeld, E. Radical constructivism: A way of knowing and learning. Fal-
mer Press London (1995)