KARL JASPERS FORUM

TA40 (Dykstra)

Commentary 4

A BETTER WORLD ?
by Alexander Riegler
27 March 2002, posted 2 April 2002

<1>
Dewey Dykstra’s target article, together with his response R2, forms an important treatise on why radical constructivism (RC) has more potential than realism. It is an interesting paper for several reasons. It examines the solipsism-reproach radical constructivism is often confronted with, it sheds light on how RC performs in science education, and explains how it relates to science. While I do support many of his arguments, I am skeptical with regard to some other points he makes, especially concerning whether a scientific world dominated by RC would unavoidably turn into a better world.

<2>
SOLIPSISM

A major point in Dykstra’s article is splitting the realism vs. constructivism debate into its ontological dimension p1 ("reality exists vs. doesn’t exist") and its epistemological dimension p2 ("we can figure it out vs. we cannot") [40-TA:3]. John Searle (1995) makes a similar distinction when he claims that "realism" doesn’t say anything about epistemological aspects. Rather, it is an ontological statement and neither a theory of truth, nor epistemology, nor a theory of language. This distinction between the ontological and the epistemological aspect is also reflected in Rollins’ definition of solipsism quoted by Herbert Müller in his commentary [40-C1:1]. Rollins, among others, distinguishes a reality solipsism from a knowledge solipsism. What is so puzzling about the ontological variant is that it is not clear why somebody would want to describe oneself as a solipsist as this per se excludes the person from any (scientific) community. Consequently, I reject the formulation that someone could possibly "accept" (ontological) solipsism the same way he or she accepts a matter of fact against which he or she is defenseless. Solipsism is not a passive attitude, it is a matter of actively claiming it. And only this distinguishes an alleged solipsist from others. Therefore, using the word "solipsism" is a mere polemic technique in order to discredit somebody else and his or her position - a philosophical or academic offence, if you want - rather than something you would want to call yourself.

<3>
VARIOUS ROOTS

As Dykstra points out, RC is totally different from ontological solipsism. He claims that it arises from a long skeptical tradition [40-TA:4]. This statement, of course, borrows heavily from Ernst von Glasersfeld’s perspective (e.g., 1995). I would disagree with this argument insofar as skeptical philosophy has never been a popular branch—perhaps because it has this negative, deconstructive connotation of questioning everything—a connotation which, in its negative form, should better not be transferred to and identified with RC. But of course, Glasersfeld is not explicitly wrong with his argument about the relationship between RC and skepticism. Therefore, I would prefer saying that RC is compatible with skepticism rather than being its historical consequence or even successor.

<4>
However, there is another, much stronger reason to moderate Dykstra’s and Glasersfeld’s claim. While Glasersfeld is undoubtedly the one who set RC moving - especially on the terminological level - there are many more who have greatly influenced the shaping of RC, such as Foerster, Maturana, and Varela, to name but a few. They arrived at their RC-like conception from very different angles which are equally appropriate starting points. What they basically maintain is that the cognitive apparatus is a closed system. Cybernetician Heinz von Foerster refers to the neurophysiological principle of undifferentiated encoding in the nervous system while biologist Humberto Maturana arrives at the closure idea from the his studies of color perception and the apparent lack of correlation between physical lightwave frequencies and subjective color experience.

<5>
Choosing closure as the starting point provides you with a different perspective on why to prefer RC to realism. No longer is there the skeptical doubt whether we can ever know an external objective reality; rather, the necessity arises to abandon this idea all together. Maintaining the existence of a reality suddenly becomes an invention - whether this makes you a liar, as Foerster (1998) put it, or not, is of secondary importance. Of course, closure doesn’t permit rejecting reality either. The organizational closure of the cognitive apparatus puts us in an agnostic position from where all we can say is that whether a reality exists or not is a useless question to ask. The most important consequence of this "agnostic" position is that from now on we will have to find explanations for observed phenomena without any anchor in a "secure" objective reality.

<6>
ARBITRARINESS

Of course, I cannot decide which of those who participated in the formation of RC have had the greatest impact on its development. This is a question they might want to decide for themselves. But while I might not feel called to shed light on the historical aspect of RC, accepting the dominance of skepticism might turn out disadvantageous for yet another reason. What the skeptical way to RC cannot (or at least does not) address properly is the problem of arbitrary constructions. Dykstra says, again in agreement with Glasersfeld, that "explanations of the world are constrained by the world" [40-TA:7] This poses the following subtle problem. Our constructing of the world does not imply anything about a pre-fabricated reality (see also Müller’s commentary [40-C1:3]; this is in contradiction to Anthony Roark’s opinion [40-C2:2] that RC affirmed the existence of a mind-independent reality [40-C2:2]). Thus we cannot say anything about it. If this argument applies, how can we then justify the inclusion of a piece of the "pre-fabricated reality" in our defense against arbitrariness - the piece that prevents us from constructing whatever we want? (cf. also Roark’s rejection of what he calls the "fit-to-experience" truth surrogate [40-R2:14]) In other words, we agree that by the very fact that we cannot transcend our experiences and world constructions we cannot make statements about a mind-independent reality - we cannot even make statements such as it exists or it does not exist (Riegler 2001b). But at the same time we should use that mind-independent reality in order to secure our world-view against the reproach of being unconstrained? In my opinion this does not hold water.

<7>
In Riegler (2001b), I offer a set of basic axioms that addresses the problem. As it is compatible with various RC authors, it can be considered the foundation of RC. The core postulates are (1) The Radical Constructivist Postulate of the organizational closure the cognitive system (mind). (2) The Epistemological Corollary that a mind-independent reality is neither rejected nor confirmed but rather considered irrelevant—the agnostic stance as described above. The corollary makes it also impossible to maintain that constructs are all there is since the agnostic stance cannot exclude the existence of a mind-independent reality. (3) The Methodological Corollary that explanations are necessarily circular since there is no outside point of reference. And (4) the Postulate of Limitations of Construction which says that the internal complexity of a cognitive system itself (being composed of experiences) gives rise to canalizations and therefore makes arbitrary constructions impossible. Postulate 4 rests on the idea that experiences are clustered in compound constructs which in turn may take part in another, even more complex construct thus restricting the freedom of the more complex compound. In the literature, the self-constraining property of complex systems has been formulated in different ways, such as Haken’s (1977) slaving principle within synergetics or Kauffman’s (1993) autocatalytic set theory. Evidently, the fourth postulate addresses the problem of arbitrariness in a way that forgoes any reference to a mind-independent reality as the system poses restrictions on itself rather than being restricted from "outside." The degree of reversibility, i.e., the revocation of canalizations, depends very much on the age of experiences involved in a particular construct. More recently developed, "young" intellectual capacities such as mathematical problem-solving are less prone to unavoidable canalizations than behavioral patterns and habits. And these are still easier to revert than perception and action patterns which we acquired in very early years.

<8>
Furthermore, "younger" capacities such as mathematical problem-solving require more conscious thought than older ones. Therefore, the degree of reversibility might also be proportional to the amount of conscious activity involved - a claim which partly contradicts Glasersfeld’s idea that consciousness is one of RC’s theoretical presuppositions needed for segmenting, remembering and reflecting on experience [40-C3:7].

This framework of postulates also accounts for surprises. We feel surprised as soon as an experience does not agree with an (implicit) prediction made by our younger conscious intellectual capabilities.

<9>
On this conscious level, we can reflect on the components of the compound constructions and do as if we could deal with each component separately in order to assemble them to new hypothetical compounds. Only from this process of reflection do we gain the impression that we could have carried out a construction also in another way—an a posteriori reflection that leaves us with the misbelief that constructions are arbitrary. When Roark [40-C2:4] tells first-order (= knowing about the world) from second-order knowledge (= knowing about the correctness of our knowledge about the world) he might be a victim to the same illusion. While we can always put in another meta-level in our scientific explanations it is (as for any formal system) always possible to flatten the hierarchy such that all meta-levels are dissolved into one single layer.

<10>
Apart from this line of argumentation, the apparent need to "meta"-know about the correctness of a physical theory - the second-order knowledge in Roark’s terminology - only arises from skepticism which doubts the correctness of constructs in the first place. For example, Roark’s statement, "we can know without knowing that we know" could also be reversed into a "we can (believe to) know that we don’t know anything." This is possible because the state of "knowing" cannot be clearly separated from "believing", e.g., in a religious sense. Of course, we can maintain that science tries to secure knowledge against believing in one-time illusions and against fraud using the scientific method (see also next section) but sooner or later also in science we arrive at a set of axioms which are no further questioned. So what is then the point in skeptically worrying about correctness? We "simply" know regardless of the content of our knowledge. Knowing is experienced as the satisfaction that the content of knowledge fits into a wider framework of knowledge. All the fuss about whether a theory is correct comes from a different corner, as I shall investigate in the next section.

<11>
SCIENCE

In his attempt to show that an RC view on science does not at all slip into the "vesania" of solipsism, Dykstra points out that science is based on two robust mechanisms that defend it from solipsistic arbitrariness [40-TA:11] (Hence the post-modern approach that not only deconstructs theories, as Dykstra believes [40-R1:4], but also seeks to deconstruct the scientific method is not at all friends with RC). These core mechanisms are reproducibility and shareability of experimental results. Both concepts do not relate to any ontological claim. However, as early science had to defend itself against other forms of "knowing", for sociological-psychological reasons it had to establish a similar technique as religions, namely to claim the possession of the unique way to knowing what "really" is (Riegler 2001b). Francis Bacon’s "Scientia est potentia" clearly reflects this attitude. In a world of uninformed laymen where each world-view claims its unique power you simply have to play the game in order to catch the attraction. Science has been successfully accepted by society because of its claim to be able to reveal the truth about reality much better than any other method, be it superstition, be it magic. While the foundation of the scientific method is clear to every scientist, it is, however, much less evident to the layman, even today. People simply need to know "for sure", like an architect needs to know for sure that the bridge will comply with the burden of the heavy traffic that will roll over it. The quicker you can establish this security of knowledge, the better for the world-view. This poses a severe problem for RC which for reasons of self-contradiction must not claim that it is the only philosophy that got it right (Glasersfeld 1995).

<12>
So the criterion of reality as the ultimate arbiter regarding the correctness of a scientific theory has had psychological and sociological rather than scientific reasons. It’s main purpose was to force people in past centuries to get rid off superstitious knowledge. In Riegler (2001b), I collected four reasons why pushing one’s own theory with a unique ontological claim is still popular. (1) Claiming authority by referring to an external truth makes one’s own point of view unassailable (Mitterer 1994). This is a technique especially popular in religious communities. Also when Roark criticizes the view that "realism is harmful to science" [40-R2:11] it may be worth pointing out that "realism" ought to refer to the patronizing truth standards put forward by an authority. Not realism itself, as Roark put it, but the realists subdue the incorrigibles. (2) From a political-sociological perspective, we may want to justify research expenses, as the true description of reality "…is what we are working for and what we spend the taxpayers’ money for" (Weinberg 1998). (3) On the psychological level, claiming the possession of unique objectivity is a means to force others to do what they would not otherwise do themselves (Maturana 1988). (4) Finally, philosophically (or better metaphysically) speaking, realism is equated with seriousness and rationality. This is reflected in Searle (1995) for whom maintaining ‘external realism’ is a "first step in combatting irrationality".

<13>
In this context, Müller’ claim that "as-if-ontologies" are required for the "stabilization of thinking" [40-C1:7] sounds intriguing. The drive to make things less complicated by doing as if a constructed entity really existed could be an important (historical) factor for realism to become a dominating world-view. At least, such as-if-ontologies might have an indirect effect on thinking as clinging to a firm reality has a strong emotional component. As soon as we accept the skeptical perspective of RC we seem to loose sight of the shore of certainty, and this might disturb (rational) thinking.

<14>
Let us investigate the two basic scientific mechanisms in more detail. The claim that scientific knowledge must be reproducible matches perfectly with the constructivist concept of inductively working organisms: "A living system, due to its circular organization, is an inductive system and functions always in a predictive manner; what occurred once will occur again. Its organization (both genetic and otherwise) is conservative and repeats only that which works." (Maturana and Varela 1980).

The requirement that knowledge must be shareable, however, leads to a problem if we say with Dykstra that "in order for experience to be shareable there must be other cognizing beings capable of disagreeing about experience" [40-TA:12]. What status do "other beings" have? Is this claim to be understood in an ontological way and thus leading to self-contradiction as Roark points out [40-C2:6]? Or can we exploit the following definition? An experiment is "shareable" if the others (in your experience) react to the presentation of the experimental results in a particular way (that most likely doesn’t surprise you since you would react similarly). Not observing any reaction means that you obviously failed to share the insight. If, later on, others share their insights which refer to your experimental results, then you successfully shared it in the first place. Such a definition does not resort to any ontological postulate.

<15>
CONSEQUENCES

Dykstra also speaks about the consequences of constructivism for education. While it seems plausible that a constructivist perspective restricts far less individual scientists compared to a realist background, I am hesitant to agree with the idea that a constructivist education would turn out many more people who are "good enough" to do science [40-TA:14]. There are two reasons. Firstly, this has the touch of J. B. Watson’s attitude who became infamous for his behaviorist assertion that an infant’s future can be dictated, be it a doctor, lawyer, artist, merchant, or thief. The ultimate objection against Watson’s claim was there is still a good portion of hereditary influence on one’s intellectual future. In addition to that I believe that the authority or personality of the teacher is a far more crucial factor in the education of further scientists than any (realist or constructivist) curriculum. Choosing one’s own future profession is more of an emotional than rational matter. And this also applies for scientists. No matter how well-balanced a curriculum is, the success of a student will always depend on the student’s own motivation (see also Glasersfeld 1983). RC applies here straightforwardly. Perceiving a problem, developing an approach towards a new topic, and creating problem-solving strategies depend on the student’s constructs rather than on the epistemological orientation of the curriculum. Secondly, there is also a social component to science in the sense that the standards for science will always self-adjust in order to have a constant percentage of people turning scientists. A society simply can’t consist of scientists only. Would an increased number of scientists in a world with an unchanged need for them not leave many frustrated?

<16>
Dykstra claims that within the RC perspective, "one is free to try out new theories" [40-TA:15]. While in theory this seems to work, it turns out that science cannot proceed that way. If Dykstra’s proposal was correct then each scientist would work on his or her own theory neglecting other theories and without building on the insights and theories of previous scientists. However, science is a highly hierarchical enterprise otherwise its achievements wouldn’t go beyond the achievements of individuals. And this poses restrictions with regard to how the individual designs his or her scientific work.

<17>
Furthermore, the two basic mechanisms of generating reproducible and shareable knowledge unavoidably lead to knowledge structures that build on each other. Scientists repeat experiments. Based on the results they improve the layout of further experiments and theories which are linked with the experiment. Scientists read about other experiments and their results and use these shared insights to go on with their own research. However, entirely new ideas have to be more thoroughly evaluated than new aspects of older trusted theories. This aspect comes into play when Dykstra sketches the picture of a constructivist scientific community which specifically disallows dogmatism [40-R2:33]. While such a picture is intriguing in that it could animate people to listen to each other since nobody is "permitted" to posses the truth anymore, it might not work for psychological reasons such as mastering the mental effort of studying the other’s world-view, and the unwillingness (or at least lack of motivation) to pursue any scientific goal with perseverance. It all sounds like planned economy which ultimately failed due to lack of personal dedication. If there is no trophy (the truth) to win, why play? Note that not only truth itself is the trophy. It is also used as the arbiter to decide whose theory will improve the reputation of its inventor. Furthermore, it is the inductive organization of living beings which gives rise to the fact that established scientists become trusted authorities (obviously, so far they have had good ideas, good theories, so we inductively assume his or her future output will be of similar quality) as compared to newcomers who might also lack the eloquence of the experienced ones. This is no explicit, no intended dogmatism. Dykstra’s claim, "there is no canon of truth to tell what to do" [40-TA:15] sounds promising but it simply doesn’t reflect the interactive canalization of scientific activity. Thus, whether an RC foundation will change the world into a better one remains questionable.

---------------------------------

<18>
NOTE

Several of the insights presented here come from my editorial work for a special issue (vol. 6) of the Kluwer journal Foundations of Science. Twelve authors investigated "The Impact of Radical Constructivism on Science" in much greater detail than I could do here. For more details see http://www.univie.ac.at/constructivism/books/fos/

------------------------------------

REFERENCES

Alexander Riegler

Free University of Brussels