Why western science and philosophy cannot deal with the relations between parts and wholes

This post consists of excerpts from Juarrero’s forthcoming book, Context Changes Everything: How Constraints Create Coherence (MIT Press 2023).

In the fable of the Scorpion and the Frog the scorpion asks the frog to ferry it across the river safely. When the frog hesitates, fearful the scorpion would kill it, the scorpion protests that doing so would mean that both would drown. The frog accepts that reasoning. Halfway across the river the scorpion stings the frog. As they are both drowning the dying frog asks the scorpion, “Why did you do that?” To which the scorpion replies, “I couldn’t help it. It’s my nature.”  Thinking of identity in this way suggests that certain properties internal to the creature define what it is to be a scorpion –  that is, those properties capture its fundamental essence, and therefore its identity. The basic idea conveyed in this fable is that interactions and relations with others do not influence one’s basic nature and identity. This is philosophy and western science’s approach to identity to this day. It reappeared in the thesis that inherited genes determine the traits of living things: many were quite surprised when sequencing the human genome did not fully account for all diseases.

In contrast, identity politics today point to contextual factors such as race, colonialism, class, culture, sexual orientation, and age in determining personal identity. Environmental and historical influences are implicated in making all living things what and who they are. In biology, the new field of epigenetics has shown that the dichotomy between nature and nurture – between internal and external – is too simplistic. The significance of protein folding is that it reveals that the functionality of the configuration is due much more to what is connected to what than to the sequence of base pairs. The recent research on the metagenome in the gut microbiome, not to mention the constitution of lichens and mycorrhizae, raises even further questions about the identity of symbiotic wholes.

Here I will argue that the seeming intractability of today’s debates over identity are framed by centuries-old philosophical and scientific presuppositions about identity; they prevent us from even considering other ways to imagine identity. The problem of identity, that is, is noteworthy because it encapsulates in a single controversythree subjects that are egregiously responsible for misleading scholars and the public alike. These are:

• How do wholes differ from aggregates, if at all?  
• Are context and history part of the fabric of reality?
• How do cause and effect work? 

Historical answers to these questions are deeply interwoven in the way we see things even today. The problem of identity arises because (1) relations and interactions, context and history have for centuries been judged to be at best derivatively real. (2) As a result, seemingly coherent wholes are considered to be nothing but impotent aggregates. And to make matters worse, (3) the very idea of cause and effect is restricted to energy-transferring processes.  The problem of identity, in short, arises because of our deeply flawed views about the possibility and sources of coherent wholes with “causal” power on their components and their behavior.

Modern science and philosophy first consolidated this idea by identifying “primary properties” as those essential properties internal to an entity that make it what it is – like atomic weight and mass. From this perspective, essence and identity are what might be called internalist concepts: they pick out objectively measurable and observer independent properties that are in the thing in question. Essential traits underpinning identity, in other words, are independent of context. Compare this with the properties of color and sound that arise from interactions with a perceiver. Their context-dependence makes them accidental properties, ontologically secondary, their reality derivative.

It is important to note that, even in classical times, Types, Kinds, and Universals were understood to be multiply realizable: different actual tokens can realize the same Type. Whether isosceles, equilateral, scalene, acute, or obtuse, actual triangles represent tokens of Triangularity; they manifest its essential Platonic Form. According to the received understanding of Types and Kinds, then, individual tokens of a given Kind differ only in their secondary properties. Their essence, those primary properties that identify the individual specimen as being the same Type of thing, remain universal and unchanging throughout.

Philosophers as far back as Plato were sophisticated enough not to confuse a thing’s Form (its core Type identity) with its shape. Justice was a Platonic Form, for example, even though it obviously has no shape. But if Forms are a transcendent principle of coherence generally that confers a thing’s essential nature, exactly how that coherence with matter comes into being in the natural world remains unexplained. In opposition to Plato, Aristotle maintained that Form exists only as coherent embodied Substance; alas, the way the 4 Aristotelian Causes are supposed to integrate and serve as the sources of that coherence is far from satisfactory.

Hypotheses about essences and primary and secondary properties intersect with views about the status of interactions and relations. The field of chemistry is all about relations and transformations; chemists study interactions among molecules and charged particles. But if only primary properties subtend essence and identity, and if relations and interactions with other entities in the environment are accidental and ontologically derivative, chemical compounds and chemical synthesis must be nothing more than clumps of constituent physical elements, along with their secondary interactions.  Chemical properties are merely effluvia, the summation of primary properties of massed fundamental particles.  If they are mere masses of elementary parts, the apparent unity of chemical compounds must be illusory. Despite presenting apparently qualitatively novel capacities and properties like oxidation, and metabolism, these processes are mere castoffs of massed particles. Specifically, clumpings elementary particles must possess no properties or causal powers in virtue of any seeming coherence.

The thesis that seemingly coherent totalities are merely massed clumps is presumed to apply generally, not only with respect to chemistry. It underlies Reductionism, the thesis that because wholeness is nothing but aggregation, any properties – especially apparently causal powers – of wholes can be derived and therefore predicted from laws pertaining to the parts. This thinking allegedly applies to living things: properties like sensation and awareness, are mere castoffs of elementary particles. They possess no causal powers in virtue of the emergent property of coherent interactions.

According to Carl Gillett (Reduction and Emergence in Science and Philosophy, 2016), today physicists have gone beyond philosophers’ insistence that qualitatively novel emergent properties of wholes are logically derivable from their constituent parts. But even physicists remain as committed as philosophers to the thesis that “complex collectives” are causally powerless as coherent wholes; they are impotent as totalities. They are epiphenomenal; they do not directly influence or change their components or own behavior. End of story.   

I submit that much of this controversy is due to the fact that, after the 16^th century, the term causality tout court and its cognates came to mean exclusively efficient cause, the transfer of kinetic energy from an agent-cause to a body. In combination with the theses that interactions with the environment are accidental, and that wholes are nothing but aggregates, restricting all cause-effect relations to Aristotelian efficient causes made the possibility of top-down causal influence from wholes to parts a nonstarter.

Going back as far back as Aristotle, the thinking went as follows: Efficient causes are logically and spatiotemporally distinct from their effects. Since causes and their effects cannot simultaneously exist both before (as cause) and after (as effect of) themselves, causes must be other than their effects.

Proposing as Rene Descartes did that Mind and Body are two separate and distinct substances, each capable of independent existence and limning non-overlapping essences, raised fundamental questions. How can a non-physical event like an intention make my material body move such that my behavior carries out the intention (see my Dynamics in Action, 1999) – that is, so that the behavior enacts the content of the intention? Descartes identified the pineal gland as the site where beliefs and intentions set in motion purposive action.

 This maneuver runs up against conservation laws and the principle of physical closure. 

Combine an exclusive reliance on efficient causality with the First Law of Thermodynamics, which holds that that the total amount of matter and energy in the universe is fixed. Matter can neither be created nor destroyed; it can be transformed (into energy) and back, but the total amount is always conserved. In combination, these two principles lead to the principle called causal closure of the physical. (Causal, as in efficiently causal). This principleholds that 1) all states are physical states, and 2) all physical states are the effects of physical causes. The reasoning goes as follows: For non-material or non-kinetic-based “causes” to introduce themselves (as efficient causes) into the physical realm, their otherworldly “power” must be converted into an energetic force that activates the body. This implies that those “powers” inserted something into the physical world, which adds to the total sum of physical matter and energy – and violates conservation. Conclusion: (allegedly) Cartesian thoughts, intentions, and other such emergent mental events or properties cannot bring about effects in the physical realm without violating causal closure.  

Mental causation as such would also violate the prohibition against overdetermination implied by causal closure.  This argument goes as follows: Causal closure implies that events (such as my arm rising) occur as a result of physical causes, whether I formulate the intention to raise it or not. Now assume for the moment that intentions can bring about effects in the physical world. In contrast to my arm rising because of an involuntary neuromuscular spasm, mental causation implies that my arm rises because of my intention to raise my arm, turning my behavior into a purposive action (Juarrero 1999).

How would that causality work? Since, on the standard view, efficient causes and their effects must be spatiotemporally and logically distinct, intentions must be other than neurophysiological processes, which would violate causal closure. So if intentions bring about change in virtue of their intentional content – by virtue of being the intention to raise my arm – my arm going up would be doubly caused, by the intention to raise my arm as well as by the neuromuscular processes involved, which causal closure forbids. The requirement of logical distinction between efficient causes and effects would also be violated since intentions are specified as the intention to do X. Conclusion: allegedly irreducible and emergent properties of living things such as intentionality, consciousness, etc., cannot bring about effects as such – as the content of the intention to do X. Top-down causation by emergent properties is thus disallowed; systemwide properties as such are epiphenomenal.^[1]  Even in the 20^th century, Jaegwon Kim’s notes that the Principle of Supervenience that attempts to salvage functionalism is vulnerable to this objection, which he calls “Descartes’ Revenge.”

Because Western philosophy’s exclusive reliance on efficient cause bars naturalist accounts of downward causation at the root, overdetermination and causal closure disallow top-down causality from wholes to parts generally. By insisting that identity rests on primary properties of fundamental material particles (while simultaneously insisting that efficient causality is the sole mechanism of bringing about effects and that relations and interactions are secondary and epiphenomenal), the process whereby a coherent dynamic with emergent properties could be generated, bottom-up, becomes intractable. Moreover, how those coherent wholes might then control their components top-down such as to execute purposive actions (both of which they would have to do solely as efficient causes) is even less tractable.

What makes coherent wholes in general, not only living things, organize and persist as the same type of entity despite dramatic alterations? Suggesting, as did Prigogine, Maturana and Varela, Pattee, that in open conditions far-from-equilibrium constraints self-organize complex coordination dynamics with strongly emergent properties even in the abiotic realm went completely against the received framework. With very few exceptions, complexity was disregarded by mainstream philosophers of science.

The possibility that, in open far-from-equilibrium conditions, context-dependent constraints might self-organize and generate dynamic wholes with emergent properties and powers (not merely as aggregations); that the constraint regimes of those novel dynamic wholes might confer emergent new properties and powers to those coherent wholes; and that among these powers might be a novel capacity to control and regulate their own behavior as well as influence and affect components top-down, in virtue of their collective systemwide and emergent properties, was simply not countenanced.  The reductionist principle that because aggregates are nothing but the sum of their parts, that seeming coherent wholes such as organisms like you and me are epiphenomenal – causally powerless, that is – ruled the day.

Context Changes Everything (MIT Press, forthcoming 2023) argues that dismissing context, as just described, is the original sin that permeates these debates, including discussions about identity and individuation. The book argues that coherence-making by context-dependent constraints can change everything in open far-from-equilibrium conditions. In particular, it articulates a mechanism whereby nature organizes into increasingly general Types of entities, with emergent properties and powers.

Constraints are multifaceted; they operate, simultaneously, across different scales and dimensions. They can be enabling, governing, multiply realizable, and both context dependent and independent. Bottom-up, enabling, context-dependent constraints such as positive feedback, catalysts, and autocatalysis, take systems to a threshold of instability and precipitate phase transitions to novel and coherent interdependencies (to emergent and collective coordination dynamics). Closure of constraint as described by Montévil, Mossio and Moreno, plays a unique role in binding together parts and wholes into coherent dynamics. Once those interdependencies achieve closure, acting as top-down governing constraints implemented as cascades of negative feedback, emergent order parameters preserve coherence by stabilizing the interdependencies among the components; in turn, acting top-down, the overarching constraint regime modulates and regulates the system’s behavior within the attractor basin of those interdependencies. Coherent wholes can therefore persist.  

Since constraints not directly involved in energy transfer, governing constraints of coherent interdependencies can do all that without violating causal closure, conservation laws, or overdetermining the natural universe.

In combination, context-independent and context-dependent constraints are responsible for phase transitions to novel forms of coherence and control. The workings of constraints reopen a path to rehabilitating coherence and mereological causation – from parts to wholes and wholes to parts. Taking context seriously therefore requires taking a different form of causality as seriously as we take forceful causes. Causality as constraint can reconceptualize the notion of wholeness and explain not only how constrained interactions among parts generate wholes with emergent properties. More significantly, it can also account for the ways in which constitutive constraints that hold those wholes together regulate, modulate, direct, and channel their components and behavior top-down to preserve the emergent properties embodied in the intertwined constraints and support the persistence of the whole’s overarching constraint regime.

[1] It is also for that reason that so-called “special sciences” such as economics, sociology, and ecology, all of which are focused on systemic properties, are often considered “non-science.”  This issue is also at the heart of the discussion on Supervenience.