From: jimruttshow8596
The ancient concept of Aristotle’s four causes provides a useful framework for understanding the nature of causation, a framework that has undergone significant shifts in philosophical and scientific thought, particularly with the advent of modern science and the study of complex systems [00:03:10].
Aristotle’s Four Causes
Aristotle’s theory of causation, possibly derived from observing a potter at a wheel, identifies four types of causes [00:03:27]:
- Material Cause: The raw material from which something is made (e.g., the clay for a pot) [00:03:40].
- Final Cause (Teleology/Purpose): The purpose or goal for which something is made (e.g., pouring water for a pitcher) [00:03:44]. Aristotle believed organisms and natural phenomena had an inherent entelechy or internal purpose, like an acorn containing the form of an oak tree [00:05:21].
- Formal Cause: The essence or fundamental identity that makes a thing what it is (e.g., what makes a pitcher a pitcher, not just its shape) [00:03:57].
- Efficient Cause: The actual force or energetic exchange exerted to make something (e.g., the potter’s hands shaping the clay) [00:04:13].
Historical Shift Towards Efficient Causality
Prior to the late 16th century, people generally considered all four causes when thinking about nature [00:04:30]. However, with the rise of modern science, there was a significant and perhaps accidental shift towards a heavy focus on the efficient cause [00:04:40]. Material cause was left to science, while formal and final causes were largely discarded [00:04:51]. This led to an over-reliance on “matter and motion bumping into each other,” often referred to as naive Newtonianism [00:06:04].
This perspective was epitomized by Laplace’s idea that knowing the position and velocity of everything in the universe would allow total prediction of the future and past [00:06:19]. Despite Newton’s own warnings about issues like the three-body problem [00:06:51], this reductionist view persisted for centuries [00:07:00].
The Problem of “Nothing But-ism”
The over-focus on efficient causality gives rise to what is termed “nothing but-ism” [00:07:30], the idea that a whole is nothing but the sum of its parts [00:07:38].
“The idea of that the whole is nothing but the sum of its parts and therefore anything that appears to be an emergent property is really an Epi phenomenon it is it is it is sort of froth that’s thrown up but it really has no causal power.” [00:07:38]
This perspective struggles to explain emergent properties and top-down causality [00:07:44]. For example, if causality is only efficient, then emergent phenomena like the synchronization of photons in a laser beam cannot align their components [00:08:08]. This leads to absurd conclusions [00:08:13].
The challenge of explaining human behavior, particularly intentions, within an efficient-cause framework, highlights this limitation [00:08:29]. An intention causing a wink, versus sand causing a blink, raises the question of how an intention, a higher-level phenomenon, can exert efficient cause [00:08:40]. This suggests the need for network properties and emergent properties that can loop back down and influence lower-level components [00:09:23].
Reconceptualizing Causality: The Role of Constraints
Instead of battling the Newtonian definition of cause, Alicia Herrero proposes focusing on the notion of “constraint” [00:10:21]. She argues that constraints are the contemporary version of formal and final causes in complex dynamical systems [00:05:55].
Myriology: Whole-Part and Part-Whole Relations
The concept of myriology explores the relationship between parts and wholes [00:11:50]. In a complex system like an economy, individual elements interact in constrained ways to produce an emergent phenomenon (the economy) with properties its components don’t possess [00:12:09]. Once this coherent whole forms (a “constraint regime”), the components acquire new properties and roles (e.g., traders, regulators) [00:12:42].
This dynamic illustrates how the whole can loop back down and affect its components [00:13:59]. Unlike the Newtonian view where “holes are nothing but the sum of the parts” [00:13:36], constraints allow for top-down causality where, for instance, a culture can influence individual behavior without being an efficient cause [00:14:44].
Taxonomy of Constraints
Herrero distinguishes between types of constraints based on their context dependence [00:22:11]:
- Context-Independent Constraints: These are conditions that take a system “far from equiprobability” or random noise, essentially setting the boundaries of possibility space and creating inhomogeneities [00:22:42]. Examples include gradients, polarity, charge (in the early universe), and principles like the Pauli Exclusion Principle [00:23:03] [00:28:57].
- Context-Dependent Constraints: These take a system “away from independence,” linking things together [00:23:41]. Examples include:
- Catalysts [00:24:43]
- Feedback loops [00:24:52]
- Epigenetics [00:25:05]
- Temporal Constraints: The timing of actions (e.g., a child kicking a playground swing at the right moment) [00:25:46]. Sequencing (A before B before C) is another example [00:27:26].
- Spatial Constraints: The architecture or arrangement of elements (e.g., the length of a seesaw plank, the design of a traffic roundabout) [00:26:28].
- Enabling Constraints: Context-dependent constraints that achieve closure, leading to the emergence of a coherent whole [00:28:04].
- Rules and Regulations: These set possibility space and determine likelihoods [00:29:35].
- Scaffolds: Temporary, external artifacts that guide construction or provide a temporary equilibrium point, like a ratchet, enabling the next step [01:03:00].
- Entrenchment and Buffers: Mechanisms that control relationships between inside and outside, resisting or moderating change [01:06:06].
Emergent Properties and Causal Power
When constraints lead to the emergence of a coherent dynamic, it results in a phase transition to a continuous function, potentially allowing for analog control [00:30:09]. This analog nature allows for homeostasis to maintain system integrity with timeliness and sensitivity to local conditions [00:30:38].
This view implies that emergent properties have causal powers [00:17:46]. For example, homeostasis (the whole maintaining its integrity) readjusts metabolism and neural systems, demonstrating a top-down causal influence on its components [00:17:59].
The Brain as a Complex System
The brain exemplifies a system that is both digital and analog [00:33:00]. While neuronal electrical exchanges are digital, the phase transition to mental events involves analog control [00:32:49]. Studies on artificial neural networks, such as those by Hinton and others, demonstrate how “semantic attractors” can emerge in middle layers, leading to semantic-based errors like deep dyslexia (reading “band” as “orchestra”) when networks are lesioned below feedback loops [00:38:05]. This emergence of a “semantic attractor” is an example of an emergent property with causal power [00:38:46].
Top-Down Causality Without Violating Physics
The perceived problem with top-down causality is that it seems to violate physical closure and conservation of energy if viewed solely as efficient causes [01:08:43]. However, when understood as “constraining dynamics,” top-down causality does not violate these fundamental physical laws [01:09:01]. For instance, a culture affects an individual by changing the likelihood of different behaviors, rather than through a direct efficient force [01:10:07]. Similarly, a molecule of water in a Benard cell behaves differently because of the constraint structure of the cell, not because of a new efficient cause [01:10:25].
Degeneracy and Many-to-One Transitions
The concept of “many-to-one transitions” or “degeneracy” is crucial [01:13:05]. In biological systems, degeneracy refers to multiple lower-level paths or forms that can achieve the same higher-level function [01:13:20]. For example, different amino acids can produce the same protein [01:13:18].
This challenges the idea of a one-to-one relationship between mental (supervenient) and neural (sub-vening) properties [01:12:27]. When a part of the brain is excised, another part might take over its function (multiple realizability), demonstrating that a specific function isn’t tied to a single neural pattern [01:12:58]. The phenomenon of “pluripotentiality” (e.g., stem cells) is the reverse: one lower-level entity has the potential for multiple different functionalities [01:15:37].
Higher-level properties of complex dynamical systems often exhibit degeneracy; an economy can maintain its identity despite varied configurations, as long as its overarching constraint structure remains within a certain range [01:14:58]. This also relates to “dimensional reduction,” where many inputs converge to a single decision or action [01:15:12].
Future Directions: The 4E Approach to Cognition
The “4E approach” to cognitive science (Embodied, Enacted, Extended, Embedded) aligns with this constraint-based view of causality [01:19:26]. This perspective acknowledges that the mind extends beyond the brain, interacting with tools and environments [01:20:12]. The mind is not just embodied, but also enacted through behavior within specific contexts [01:20:57].
This holistic view posits that coherent dynamics arise from an accumulation and interweaving of constraints over time, creating an “overarching interlocking set of constraints” [01:22:20]. These constraints are considered real, not just epistemological constructs [01:22:38]. The concept demonstrates that while basic physics laws hold true, there’s a more interesting structure built upon them—the relational and dynamical properties that are just as real as primary ones [01:08:29]. This approach shows how different levels of complexity (physics to chemistry, chemistry to biology, etc.) involve phase transitions that produce new emergent properties and “codes” [01:19:10].