HOLISM  VS  REDUCTIONISM  -  AN  AGE-OLD  METHODOLOGICAL  ISSUE

 

 

When looking around on the cybernet for information about where the Cognitive science of today stands regading Claude Lévi-Strauss´ ideas about "binary opposition" as some kind of fundanental and universal organising principle, I found this very enlightening paper by James Marcum at the Department of Philosophy, Baylor University, from which I have quoted some excerpts, but deleted all the footnotes and references for the text to be more easy to read. If you want to examine these notations more deeply you have to look it up in his paper here »

 

METHODOLOGICAL APPROACHES OF HOLISM AND REDUCTIONISM

 

2.3 Ontological Categories of Organicism and Physicalism

A scientist’s or a scientific community’s investigation of the natural world depends upon a particular methodological approach to that world.

 

During the first part of the twentieth century, two approaches— holism and reductionism—achieved notoriety for how best to examine and explain natural phenomena.

 

Although holism certainly has its roots in the Aristotelian adage, “the whole is greater than the sum of its parts”, the term was only introduced in the first half of the twentieth century.

 

Jan Smuts coined it in reaction to the prevailing approach of mechanistic reductionism. Smuts defined holism as:

 

“the ultimate synthetic, ordering, organizing, regulative activity in the universe which accounts for all the structural grouping and syntheses in it, from the atom and the physic-chemical structures, through the cells and organisms, through Mind in animals, to Personality in man”.33

 

Given the expansive nature of holism, Smuts realized that the scientific community might take the notion as “a mere assumption which may have a philosophical or metaphysical value, but that it has no scientific importance, as it cannot be brought to the test of actual facts and experiments”.34 Indeed, the scientific community did respond as he presaged and it did not widely accept the concept.35

 

However, in post-genomic and systems biology of the late twentieth and early twenty first century, the notion of holism has resurfaced as a means for addressing the complexity of biological phenomena.36

 

Contemporary holism involves an approach by which a scientific community investigates natural phenomena in their wholeness or integrity because “the properties and behavior of ‘whole’ systems or objects (cells, persons, societies, etc.) cannot be reduced to, or explained fully by reference to, the properties and behavior of their parts”.37

 

In other words, the simple summation of the properties of the individual parts constituting a phenomenon cannot account for its overall properties; rather, the properties of the phenomenon qua whole arise from its structure or organization, and the relationships among the different parts.

 

Thus, the behavior of a phenomenon is irreducible to the behavior of its parts and reflects its unique and specific organization.

 

Finally, causation is top-down in which higher-ordered structures regulate and temper the activities of lower-ordered structures.38

 

Although various types of holism—or antireductionism for some like Thomas Nagel who contrast it strictly with reductionism—are proposed in the literature, only epistemological or theoretical holism, methodological holism, and ontological holism are examined in the present section.39

 

Epistemological or theoretical holism pertains to the discovery or formulation of organizational laws that are required for explicating the behavior of a phenomenon.

 

In antireductionist terms, theories and laws pertaining to the holistic dimension are not reducible to lower level theories and laws governing the constitutive elements of it.

 

Methodological holism refers to empirical and experimental investigations that maintain a phenomenon’s integrity or coherence. In antireductionist terms, holistic protocols do not compromise organizational structure or reduce it to component parts in order to study the part and then synthesize the phenomenon based on the properties of its parts.

 

Ontological holism also pertains to the distinctiveness of a phenomenon vis-à-vis its components, i.e. a phenomenon qua whole is not simply representative of an aggregation of its components.

 

In antireductionist terms, a phenomenon’s properties are not reducible to properties of its parts.

 
Finally, ontological holism also pertains to causation, as mentioned earlier, in that higher-order structures can causally influence its parts and regulate them, i.e. top-down causation.

 

In antireductionist terms, causation is not simply bottom-up in terms of the traditional analytic- synthetic method.40

 

Holism is the approach of neuropsychologists advocating DPT to investigate and explicate the nature of cognition. In terms of epistemological or theoretical holism, DPT qua theory of cognition approaches cognitive activities as the complex interaction of two predominant processes, which are responsible for specific cognitive activities. In other words, cognition is not reducible to one particular process, such as rule utilization. In terms of methodological holism, DPT represents investigation of cognitive activities at multiple levels to maintain their integrity.

In other words, cognitive activities cannot be investigated without investigating them vis-à-vis other activities as well.
As for ontological holism, DPT provides a means for exploring the differences in cognitive activities among cultures.

 

For example, Richard Nisbett and colleagues demonstrated that Eastern cognition is often context-dependent and concrete compared to Western cognition, which is generally context-independent and abstract.41

 

According to Emma Buchtel and Ara Norenzayan, DPT provides a means for explicating these differences.42 Eastern cognitive activities reflect predominantly T1 processing, although not exclusively, while Western chiefly T2 processing, although again not exclusively.

 

Throughout the twentieth century, reductionism was the predominant approach towards investigating nature, especially biological phenomena. For example, biologists aimed to reduce inheritance in terms of classical genetics to molecular genetics.43

 

Briefly, reductionism refers to the investigation and understanding of natural phenomena, such as living organisms, in terms of their component parts.

 

“Reductionism”, according to Thomas Nagel, “is the idea that all of the complex and apparently disparate things we observe in the world can be explained in terms of universal principles governing their common ultimate constituents: that physics is the theory of everything”.44

 

The drive to reduce natural phenomena and the scientific theories accounting for them to one theory, i.e. a theory of everything, was part of a larger effort, especially by the early twentieth century logical positivists and their academic descendants, to unify the sciences—particularly in terms of the physical sciences.45

 

Although various types of reductionism have been identified and discussed in the literature, only epistemological or theoretical, methodological, and ontological reductionism—to parallel the types of holism—are considered in this section.46

 

Epistemological or theoretical reductionism refers to the articulation of knowledge at higher levels of organization within terms of lower levels. For example, an organism’s phenotype can be expressed in terms of its genotype, which, in turn, can be specified with respect to a particular nucleotide sequence. Thus, biological laws governing organismic activities can be articulated sufficiently—if not completely—in chemical and/or physical laws, although such reduction has yet to be achieved fully.

 

Next, methodological reduction involves investigating the constitutive parts of a phenomenon in order to characterize them and to synthesize the phenomenon based on their characterization.
Finally, ontological reductionism pertains to an indistinctiveness of the phenomenon vis-à- vis its components, i.e. a phenomenon simply represents an aggregation of its components. In other words, properties of the phenomenon qua whole are a summation of the properties of its parts.

 

Ontological reductionism also states that causation is from bottom-up in that the components constituting a phenomenon are sufficient to account for it synthetically.

For proponents of UMT, epistemological or theoretical reductionism refers to explicating cognitive activities in terms of rule-based activity alone. In other words, DPT can be reduced to UMT in that the cognitive processes involved in intuitive reasoning or T1 processing depend exclusively on rule-based activities comparable to those of deliberative reasoning or T2 processes.

 

Moreover, UMT relies on methodological reductionism in that cognitive activity like DPT’s T1 and T2 processes are similar or even identical vis-à-vis rule selection and utilization to perform a cognitive task, i.e. cognition represents employment of a rule or set of rules under a particular situation.

 

In other words, cognition can be investigated simply as an aggregation of the rules used to make a judgment. Importantly, according to UMT advocates, “the same rules can underlie both intuitive and deliberate judgments”.47

 

Finally, ontological reductionism for UMT involves bottom-up causation in that the rule or set of rules employed in cognitive activities, whether intuitive or deliberative, are sufficient to account for it synthetically, while for DPT advocates both intuitive and deliberative activities are ontologically distinct—not a trivial distinction.
Although the roots of organicism extend back to the Greeks with Aristotle’s adage about the relationship between parts and wholes, the notion achieved widespread attention in the late nineteenth and early twentieth centuries.50

Probably the most celebrated champions of organicism in the English-speaking countries were the members of The Theoretical Biology Club, which met in Cambridge and Oxford prior to World War II and afterwards in London until 1952, when it failed to obtain funding from the Rockefeller Foundation.51 Members of the club included Francis Huxley, Peter Medawar, Dorothy and Joseph Needham, Karl Popper, Conrad Waddington, Joseph Woodger, and Dorothy Wrinch.

 

The club’s goal was to tackle and possibly resolve the following problem: “What is the relation between those large particles which we call elephants, trees, or men, and those extremely small ones which we call molecules or electrons?”52

 

Alfred North Whitehead’s philosophy of the organism was influential among the club’s members, especially in terms of providing a moderate position between—or an alternative tothe extremes of mechanism and vitalism.53

 

Two of the major criticisms or complaints against organicism are ambiguity with respect to defining its nature and its association—albeit loose—with vitalism.54
Denis Phillips, in a historical analysis of the notion of organicism identifies five versions of it in the literature.

 

The first represents a rejection of biological mechanism as adequate for explicating organism qua whole, particularly in terms of the kinds of relationships required to constitute the organism. As Phillips notes, this version has served as the basis for the other four.

 

The second version is predicated on the Aristotelian adage that the properties of the organism qua whole are greater than the sum of the properties of its parts, while the next claims that the organism itself determines the properties of its parts.

 

The fourth version asserts that the properties of the parts cannot be “understood”, if isolated from the organism as a whole.
Phillips acknowledges that advocates of this version of organicism are unclear by what is meant by the term, “understood;” but, he offers this defense: “the intention of organicists probably was to state that the nature of the parts (i.e., their defining characteristics) cannot be known if the parts are considered in isolation from the whole”.55

 

The final version proposes that an organism’s parts dynamically interrelated and interdependent upon one another.
Although organicism was eclipsed temporarily during mid-twentieth century, it reappeared in the late twentieth to early twenty-first century.56 For example, Scott Gilbert and Sohortra Sarkar propose a version of organicism they call materialist holism, which they claim accounts for the emerging complexity seen during developmental processes.

 

This holism is materialist in order to distinguish it from vitalist holism.

 

To that end, they combine both bottom-up and top- down causation and identify the combination as a mechanistic property of living organisms.

 

“The properties of any level depend”, as Gilbert and Sarkar go on to explain, “both on the properties of the parts ‘beneath’ them and the properties of the whole into which they are assembled”.57 At a higher level, properties emerge from the interactions of the underlying parts.58 In addition, the emergent properties exhibit regularities that can be accounted for through organismic laws.

 

Gilbert and Sarkar recognize that some organicists claim emergent properties are unpredictable and so cannot be accounted for via such laws. However, they argue that emergent properties can be used to explain organismic behavior, even though it is unpredictable, and they cite evolutionary history as an example.

 
DPT’s ontological category is organicism, in the sense that cognition represents an emergent property that cannot be reduced to the property of anyone component comprising either T1 or T2 processes. In terms of Phillips’ versions of organicism, DPT represents an example of the final version in which the different cognitive processes are dynamically interdependent and interrelated to one another. In other words, T1 processes are intimately related or connected to T2 processes, e.g. via T2 oversight processing. T2 processing can decouple T1 processes if T1 output appears questionable, delaying a cognitive decision until a reasoned decision is made via T2 processing. Moreover, a change in the properties of T2 processes can have an impact on a cognitive outcome. For example, if T2 processing does not override a questionable T1 output when necessary then an error in judgment could be made. Moreover, James McClelland identifies various constructs within cognitive science that represent “emergents.” Specifically, he lists architectural constructs, such as attention and memory, and cognitive processes and their outcomes, such as beliefs and inferences.59 Certainly, DPT includes each of these constructs.60 In sum, the distinctive existence and/or identity conditions of the cognitive processes of DPT are organic, i.e. adaptable, in nature and not simply or strictly reducible—in contrast to UMT—to the physical.

 

Physicalism is the metaphysical category in which reality is explicated narrowly in terms of matter or material substance and forces acting upon it.61

 

It is considered the successor of materialism, which holds that the world is composed of matter only, since physicalism incorporates notions of physical forces and energy.62 “Physicalism”, according to Andrew Melnyk, “is roughly the thesis (1) that every entity is either itself a physical entity or is exhaustively composed, ultimately, of physical entities, and (2) that every property is either itself a physical property or is realized, ultimately, by physical properties”.63 The world—whether living or not—is made up of physical stuff, and this stuff interacts in a mechanical or mechanistically way.64 In other words, the world is a machine composed of individual parts that operates consistently and predictably for the most part, given a specific input that, in turn, determines an outcome. For physicalism machines cannot adapt, in contrast to organicism where organisms can.

 

UMT’s ontological category, especially as Kruglanski and colleagues define UMT, is physicalism.65 It is not physical in the sense of matter

 

According to some, cognition is analogous to digital computation in which rules provide the software to analyze experiential data in order to generate a cognitive or an epistemic output.67 In other words, cognitive agents are like “digital computers” in a very real and ontological way.68

3 Plurimodel Theory of Cognition

 

It appears that the debate between proponents of DPT and UMT of cognition has reached an impasse because of the limits associated with the metaphysical, methodological, and ontological notions informing the debate.69 The metaphysical presuppositions of DPT and UMT are too parsimonious to capture the extensive nature of cognition. And, their methodological approaches are often so constrained and myoptic that they distort cognitive phenomena. Finally, the ontological categories are not robust enough to account for the causal complexity of cognition.

 

In response to the apparent impasse associated with the debate between advocates of DPT and UMT, a plurimodel theory (PMT) of cognition—based on a metaphysical presupposition of pluralism, a methodological approach of pragmatism, and an ontological category of dynamical system—is introduced and discussed in this secti

4 CONCLUSION
The metaphysical presuppositions and methodological approaches, along with their attendant ontological categories, are critical for analyzing and moving towards resolving the cognition debate between advocates of DPT and UMT.

 

In particular, DPT’s metaphysical presupposition of pluralism captures the extensive nature of human cognition as compared to the metaphysical presuppositions of either DPT or UMT, which are too parsimonious.

 
And, its approach of pragmatism provides a comprehensive method for investigating human cognition as compared to the limited and overly myopic approaches of DPT and UMT. And lastly, its ontological category of dynamical system is sufficiently robust to account for the complexity of human cognition as compared to the ontological categories of DPT and UMT, which are too constrained in terms of representing the causal interrelatedness of cognitive processes.

 

Finally, the proposed PMT of cognition not only provides the philosophical resources for addressing the DPT-UMT debate and possibly resolving it but it also affords “a unified theoretical framework for cognitive science, as well as an understanding of the emergence of cognition in development and evolution”.87 As noted earlier, this is especially true for integrating various theories of cognition, ranging from the connectionist theory to the extended theory of cognition.

 

Department of Philosophy, Baylor University

 

Email: James_Marcum@baylor.edu