From: jimruttshow8596
The discussion with Alicia Herrero, author of Context Changes Everything: How Constraints Create Coherence, delves into how constraints, rather than just efficient causes, shape and define complex systems [02:53:00]. This perspective challenges the traditional, reductionist views that have dominated scientific thought since the modern era.
Aristotle’s Four Causes and the Rise of Efficient Causality
Historically, thinkers like Aristotle considered four types of causes when analyzing phenomena, often illustrated by a potter making a pot [03:27:54]:
- Material Cause: The raw material (e.g., clay) [03:36:09].
- Final Cause (or teleology/purpose): The goal or purpose of the thing being made (e.g., pouring water for a pitcher) [03:44:06].
- Formal Cause: The essence or fundamental identity that makes a thing what it is (e.g., what makes a pitcher a pitcher) [03:57:04].
- Efficient Cause: The actual force or energy exerted to bring about the effect (e.g., the potter’s hands shaping the clay) [04:13:56].
Prior to the late 16th century, all four causes were typically considered when thinking about nature [04:30:23]. However, with the advent of modern science, there was a heavy focus on the efficient cause, leading to material cause being left to science, and formal and final causes largely discarded [04:40:27]. This over-reliance on efficient cause is often referred to as “naive Newtonianism” [06:04:18], contributing to the “Laplacian error” of believing that total prediction is possible if all positions and velocities are known [06:19:07].
In the context of complex dynamical systems, formal and final causes implicitly re-emerge [05:11:07]. Constraints, in this view, serve as the contemporary version of formal and final causes, arising from interactions with the environment [05:54:57].
Myriology and the Whole-Part Relationship
Myriology refers to the study of whole-part relationships [11:54:59]. The idea of “nothing-but-ism” (or reductionism) posits that the whole is nothing but the sum of its parts, and any apparent emergent property is merely an “epi-phenomenon” without causal power [07:32:00]. However, this view quickly leads to absurd conclusions, as it struggles to explain phenomena like the synchronization of photons in a laser beam or the causal power of an intention on action [07:50:07].
In complex dynamical systems, interactions among individual elements occur in a constrained way, leading to emergent phenomena that possess properties their components do not [12:12:35]. Once a “coherent whole” emerges—organized by constraints—the components themselves acquire new properties and roles (e.g., traders in an economy, citizens in a society) [13:01:14]. This illustrates top-down causation, where the whole influences its parts [13:59:03]. For example, a culture affects individual behavior, not as an efficient cause, but by constraining possibilities [14:53:08]. The idea that emergent properties have causal powers, as seen in homeostasis, which readjusts bodily systems to maintain integrity, is crucial [17:37:37].
Defining and Taxonomizing Constraints
Constraints are conditions or factors that influence a system, taking it away from randomness or independence [22:42:07].
Types of Constraints:
- Context-Independent Constraints: These conditions take a system far from equiprobability, essentially setting the boundaries of a “possibility space” and creating inhomogeneities [22:45:30]. Examples include gradients, polarity, charge, and fundamental principles like the Pauli Exclusion Principle [22:56:04].
- Context-Dependent Constraints: These take a system away from independence [24:40:34]. They link things together and include:
- Catalysts [24:48:19]
- Feedback loops [24:52:57]
- Epigenetics: A clear example of context-dependent constraints influencing gene expression [25:02:40].
- Sequencing: Where the order of events makes a huge difference in the outcome [27:26:24].
- Architectural design: Like the layout of a roundabout, which influences the behavior of drivers and pedestrians [27:00:26].
- Temporal and Spatial Constraints:
- Temporal: The timing of actions, as seen in a child learning to pump a playground swing effectively [25:57:04].
- Spatial: The physical arrangement, as demonstrated by a seesaw, where the length of the plank and the position of the children determine balance [26:28:18].
- Enabling Constraints: A subset of context-dependent constraints that, together, achieve closure, allowing a coherent whole to emerge [27:48:07].
Constraints can also include rules and regulations that set possibility spaces and determine likelihoods [29:27:05]. Higher up the “stack” of complexity, these are more about probabilities than black-and-white outcomes [29:43:24].
Analog vs. Digital Control in Complex Systems
The brain serves as an excellent example of a system that employs both digital and analog control [32:57:48]. While neuronal electrical exchanges are digital, the transition to a mental event, such as facial recognition, involves analog control based on typology [32:38:09]. Analog processes are significantly more computationally efficient than digital ones for comparable tasks, which is why the brain doesn’t overload [32:12:06]. This suggests that higher-level, top-down control in complex dynamical systems might be analog, allowing for continuous adjustment and sensitivity to local conditions, similar to a dimmer switch [30:06:58].
Identity as Interdependent Constraints
Identity, whether personal or collective, can be understood not as an internal essence but as a set of interdependent constraints [34:51:24]. These constraints create a high-dimensional identity that is influenced by various factors like culture, society, and family [34:55:40].
Semantic Attractors and Top-Down Causality
The concept of “semantic attractors” illustrates how emergent properties can have causal power. Research into artificial neural networks by Hinton, Plaut, and Shallice, which simulated dyslexia, demonstrated this [35:34:39]:
- Surface Dyslexia: Occurs when networks are “lesioned” (damaged) above feedback loops, leading to transposition errors (e.g., “cat” read as “tac”) [37:22:24].
- Deep Dyslexia: Occurs when networks are lesioned below feedback loops, resulting in semantic errors (e.g., “band” read as “orchestra,” “bed” read as “cut”) [37:42:07].
This suggests that the network, through its middle layers, creates semantic attractors that influence the output, providing evidence for causally powerful emergent properties [38:31:37].
Constraints in Ecosystems: Mutual Adjustment
Ecological systems demonstrate how constraints lead to mutual adjustment and co-evolution. The prairie grassland ecosystem example, where grasses “invite” horses to feed on them by growing “mary stems” that are easy to eat, showcases this [42:33:04]. This seemingly paradoxical relationship helps the grasses compete more effectively against flowering plants, transforming a competitive dynamic into an enlarged, mutually beneficial ecosystem [43:10:29]. This highlights that “fit” in evolutionary terms implies mutual adjustment, where components and the whole constantly constrain and adapt to each other [44:11:46].
Time and Constraints: Ordinality and Indexicality
Beyond simple cardinality (amount), complex systems involve specific temporal constraints:
- Ordinality: Implies a sequence or order of events (first, second, third) that cannot be derived from Newtonian mechanics alone [45:34:26]. The social evolution of cassava processing in South America is a prime example; specific preparation steps must be followed in a particular sequence to remove poison [49:11:15].
- Indexicality: Refers to how perspective or position within a complex dynamical structure influences properties and causality [46:10:50]. This means causality must be interpreted in terms of emergent dynamics, where the view from inside an attractor basin differs from the view from outside [47:56:56].
Classic Complex Systems Examples
Key examples illustrate the principles of complex systems and emergence:
- Bénard Cells: When a viscous fluid is uniformly heated from below, beyond a certain temperature gradient (a context-independent constraint), the system undergoes a phase transition and self-organizes into rolling hexagonal convection cells [50:58:05]. The emergent cells then top-down constrain the individual molecules, making them behave as if they “knew” what others were doing [51:55:05]. This exemplifies how order emerges from chaos through dissipative structures, creating coherent wholes that persist [53:25:27].
- Kaufman’s Buttons: Tying buttons together transforms them into a mesh, illustrating a phase transition due to interdependent components [24:24:21].
- Huygens’ Pendulums: Multiple pendulums suspended from a common support will eventually synchronize their swings, a classic example of self-organization [50:35:48].
These systems often become more efficient at burning energy, even with increased structure, without violating the Second Law of Thermodynamics [55:29:21].
Autocatalytic Networks and the Secret of Life
The distinction between non-living and living complex systems lies in “closure of process” [57:12:12]. In autocatalytic and hypercycles (like those found in cells), the constraints themselves become self-perpetuating and self-reinforcing [57:40:17]. A prime example is how autocatalytic reactions within a cell are responsible for building and maintaining the cell membrane, which in turn allows for the necessary concentration of reactants [58:31:02]. This “self-setting” of boundary conditions enables reproduction and is a defining characteristic of life [58:16:35]. The emergence of a new dynamic often corresponds to a “new code”—a new set of rules and settings governing the system’s boundary conditions [01:00:10].
Scaffolds and Other Constraining Processes
Constraints are not always static; they can be dynamic processes that facilitate change:
- Scaffolding: Like architectural scaffolds, they provide temporary equilibrium points or “ratchets” that lower the activation energy for taking the next step in development or construction [01:04:53]. Some scaffolds, like bone growth lattices, can even become integrated into the final structure [01:06:49].
- Entrenchment and Buffers: These control the relationships between internal and external elements, influencing resistance to change or providing stability [01:05:37].
These processes demonstrate effects that are not solely efficient causes, showing how relational and dynamical properties are as real as primary ones [01:07:31].
Hierarchical Interaction: Structure, Function, and Regulation
Complex dynamical systems often exhibit three interacting layers: structure, function, and regulation [01:03:10].
- Structure: The physical or organizational form (e.g., the vasculature of the circulatory system) [01:00:33].
- Function: The dynamic processes (e.g., homeostasis) [01:01:54].
- Regulation: The “set points” or “dimmer switches” that control the functional system, influencing its responsiveness to context [01:02:05]. Conditions like PTSD or chronic inflammatory disease might stem from issues with these regulatory settings, rather than just dysfunctional structure or function [01:02:29]. These levels interact through constraints, not merely efficient causes [01:03:30].
Many-to-One Transitions and Degeneracy
“Many-to-one transitions” and the concept of “degeneracy” are crucial to understanding complex systems.
- Degeneracy: Multiple lower-level forms can achieve the same higher-level function [01:13:10]. For example, different amino acid sequences can produce the same protein [01:13:15]. This contrasts with the older “one-to-one” notion of supervenience, which struggled to explain how different neural patterns could result in the same mental event [01:11:40].
- Multiple Realizability: The same emergent property can be realized through many different lower-level paths or configurations [01:13:34]. This is particularly true for higher-level properties of complex dynamical systems, where the overarching constraint structure can remain stable despite varied underlying configurations (e.g., an economy) [01:13:48].
- Pluripotentiality: One lower-level entity has the potential to become many different functionalities, as seen in stem cells [01:15:18].
These concepts highlight that top-down causality does not violate physical laws but operates by constraining dynamics, changing the likelihood of behaviors within a possibility space [01:09:41].
The Future: 4E Approach to Cognitive Science
The “4E” (Embodied, Enacted, Extended, Embedded) approach to cognitive science aligns with this understanding of constraints and context [01:19:23]. It posits that the mind is not solely in the brain but is:
- Embodied: Rooted in the agent’s physical body [01:20:41].
- Enacted: Shaped by the agent’s behavior within specific contexts [01:20:50].
- Extended: Incorporating tools and artifacts (e.g., feeling a car as an extension of one’s body when parking) [01:20:05].
- Embedded: Situated within and inseparable from a larger, coherent ecosystem [01:21:08].
This view emphasizes that cognitive processes emerge from and are shaped by the accumulation and interweaving of constraints over time, forming “channelized” and interlocking sets of relationships [01:21:59]. This perspective highlights the reality of these constraint dynamics as fundamental to how reality works, moving beyond purely epistemological interpretations [01:22:21]. This aligns with complexity science in organizational context and other fields, suggesting homologous and analogous constraint dynamics across different levels of organization, from physics to chemistry to biology [01:18:42].