04-relations-synergy-emergence

4. Relations, Synergy, and Emergence

Source: ebook ch. 7–8 (“Synergies and Relations”, “Emergence”); slides 16–17 (“Emergence”, “Emergent Thinking”).

This chapter goes deeper into the Reduction → Emergence paradigm shift introduced in 01-worldview-and-paradigm.md. It is also, per the ebook, arguably “the central idea within systems theory.” The dedicated Emergence Theory and System Theory guides in this folder were reviewed and their load-bearing, non-redundant ideas folded in below; their more technical/applied material remains out of scope (see README.md).

Relations

“The beauty of a living thing is not the atoms that go into it, but the way those atoms are put together.” — Carl Sagan

A relation is a simple but abstract concept: a connection or interaction between two or more components. Through this connection there is an exchange of matter, energy, information, or ideas that binds the elements into a state of interdependency, where the gains and losses of any one component are correlated with those of the others in the relationship.

Relations between a system’s constituent elements come in two fundamentally different kinds:

• Constructive relations → Synergy
• Destructive relations → Interference

Synergy

A synergy is an interaction or cooperation of two or more components that produces a combined effect greater than the sum of their separate effects.

• Classic example: the honeybee and the flowers it pollinates. Bee and plant exchange pollen and nectar; each has a need it cannot fulfill alone — the bee needs a resource it cannot produce itself, the immobile plant needs transportation for its pollen. Both get more out of the interaction than they put in — the sum is greater than the parts in isolation.
• Another example: gains from a business merger, attributable to combined talent, economies of scale, and cost reduction.

With synergies, the value added by the system as a whole — beyond what’s contributed independently by the parts — is created primarily by the relationship among the parts, i.e. how they are interconnected, not by anything intrinsic to the parts themselves.

Interference

Interference is the prevention of a process or activity from being carried out properly due to interaction between elements or systems — a relation that is destructive in nature, reducing the combined output of the system to less than the sum of its parts.

• Example: interference between two drugs, where one affects the activity of the other negatively when both are administered together, reducing the overall positive effect below the benefit of either drug alone.
• Example: destructive interference between sound waves that are out of sync, leading them to cancel each other out.

The degree of synchronization (or a-synchronization) between elements is an important factor in determining whether a relation tends toward synergy or interference. Game theory gives this a formal backing: a relation structured as a positive-sum game (both parties can gain, as in trade) tends toward synergy; one structured as a zero-sum game (one’s gain is exactly the other’s loss) tends toward interference or, at best, neutrality. Some of the more damaging patterns in real systems are social dilemmas — interactions that are individually rational but collectively destructive, such as the tragedy of the commons — where a shared resource is depleted because no individual actor bears the full cost of their own extraction. (Full game-theoretic treatment in 10-complex-adaptive-systems.md.)

Not all relations carry equal structural weight, either. A single infrequent weak tie bridging two otherwise-separate clusters of a system can matter more for the system’s overall resilience and diversity than many strong ties confined within one cluster — see 09-networks.md for the structural (graph) vocabulary this draws on.

Differentiation and integration

Synergies often arise as a product of differentiation — a process, occurring in many kinds of systems as they develop, defined by the proliferation of subsystems and specialized elements internal to the system, making it more capable of responding to a greater diversity of states in its environment.

• Most notable in multicellular organism development: starting as a single cell, through division the organism develops a multitude of differentiated (specialized) cells capable of performing many different functions.
• The same pattern is observed in the development of technologies and social organization.

Differentiation also involves a proliferation of relations through which the now-specialized components can draw on each other’s services — specialization followed by exchange is a key source of synergistic relations, not just a byproduct of them.

But differentiation is only half the story. Its necessary counterpart is integration: the process of recombining differentiated, specialized parts back into a coordinated, functioning whole. The two form a dialectic — each enables and drives the other over a system’s development — and a system’s overall synergy depends on keeping them in balance:

• Over-differentiation without integration produces fragmentation — parts so specialized and disconnected from each other that they can no longer be coordinated (an academic field split into subfields that no longer talk to each other).
• Over-integration without differentiation produces crowding-out — so much central coordination or standardization that local diversity and specialization are lost.

Negative synergy (interference) can result from imbalance in either direction, not only from insufficient differentiation.

Simple, complicated, and complex

A useful typology, based on the symmetry of a system’s pattern, sits between “set vs. system” (ch. 2) and emergence proper:

• Simple systems have a high degree of symmetry — a small number of repeating rules generate the entire pattern (a honeycomb tiling; a crystal lattice).
• Complicated systems have little to no symmetry — no shortcut rule generates them; they must be described (or built) part by part (a tangled ball of string; most machines).
• Complex systems interleave symmetry and asymmetry — recognizable order coexists with underlying randomness or variation (an ecosystem; a city; a language).

This is a different axis from “how many parts” — a complicated system can have very few parts (a knot) and a simple one very many (a crystal). It is complex systems, not merely complicated ones, that are the typical home of the emergence and self-organization discussed below.

Emergence

“A process whereby larger entities, patterns, and regularities arise through interactions among smaller or simpler entities that themselves do not exhibit such properties.” — Wikipedia (quoted in the ebook)

Where synergy says the combined effect is quantitatively greater than the sum of the parts, emergence says something stronger: what’s created is qualitatively different — none of the contributing elements possess, even in isolation, the qualities that emerge at the level of the whole.

• Water: neither hydrogen nor oxygen has the property of “wetness.” Wetness emerges purely from how the two combine, and exists only at the macro level of the resulting substance.
• Life: a plant cell is made of inanimate molecules, none of which is alive in isolation. It’s the particular arrangement of these elements into structures and processes that enables the emergent phenomenon of a living system.
• Also cited: ant colonies, galaxies, cultures, traffic jams, social movements, flocks of birds, the human body (trillions of cooperating cells), hurricanes, financial crises.

These systems are the products of the process of emergence — unassociated elements interact, synchronize into synergies, and out of this emerges some new, previously non-existent, phenomenon. Emergence can be understood as a form of nonlinear pattern formation, driven by an evolutionary-like dynamic. The specific mechanism by which this bottom-up patterning happens with no central coordinator — self-organization — along with the dynamics (attractors, the “edge of chaos”) that govern it, is treated in 10-complex-adaptive-systems.md.

Reductionism:  top-down hierarchy — a whole broken into a tree of parts
Emergence:     bottom-up — many small, unconnected parts self-organize into a new pattern

Phase transitions

To produce something qualitatively new, a system must go through a phase transition — an often rapid or accelerated period during a system’s development, on either side of which the fundamental parameters describing the system change qualitatively.

• Physical example: solid → liquid when a substance is heated.
• Biological example: the metamorphosis of a caterpillar into a butterfly — not just the morphology but the whole set of parameters used to describe the creature changes so drastically, before and after, that we give it an entirely new name.

Weak vs. strong emergence

The sharpest way to draw this line is by whether the emergent, macro-level phenomenon exerts downward causation — causal influence back down onto the very micro-level elements that produced it in the first place:

• Weak emergence: only upward causation. The macro pattern is, in principle, fully derivable from and simulatable by the micro elements — it’s just often too computationally expensive to actually do so. Conway’s Game of Life is the clean example: every macro-level pattern (gliders, oscillators) is 100% computable from the simple per-cell rules, yet practically unpredictable without running the simulation.
• Strong (“irreducible”) emergence: genuine downward causation. The macro level constrains or governs its own substrate — an organism regulates its organs; an institution constrains the individuals who constitute it — and the phenomenon requires genuinely new descriptive vocabulary at the higher level, not merely a harder computation at the lower one. Consciousness and quantum entanglement are the examples most often cited.

A related distinction: epistemological emergence is a limit on our knowledge or computing power — in practice we can’t derive the macro from the micro, though nothing in principle prevents it (this is what weak emergence amounts to). Ontological emergence claims something stronger about the world itself — that the macro level is in principle irreducible to the micro, independent of any observer’s knowledge or computational limits (this is what strong emergence claims).

This distinction connects directly to the next chapter’s theme of hierarchy — micro vs. macro levels, and bottom-up vs. top-down causality (see 05-hierarchy-and-abstraction.md).