viewtopic.php?f=10&t=28460
I wanted to write up something about emergence to discuss. Reviewing things I read over the years, I looked around for something more recent. I found a new paper by Dequet et al. that I thought was interesting. From the abstract:
… Our goal in this paper is to give a broad survey of emergence definitions, to extract a shared definition structure and to discuss some of the remaining issues. We do not know of any comparable surveys about the emergence concept. …
I thought that was very inceitful actually and I’ll piggyback on that thought for this thread. I’m not going to follow Dequet’s paper, I’ll jump over the discussion about the “notion of detection”. I didn’t see that as particularly interesting. However, two slightly different conceptualizations of emergence that Dequet identified are:
1) Weak versus strong emergence
2) Levels of nature
I’ll add a third which I’ll touch on here.
3) Nonlinear or dynamical systems
Bedau is a philosopher at Reed college and has written extensively on emergence. Here’s what he says about trying to define emergence:
There are a variety of notions of emergence, and they are contested. We can provide some order to this controversy by distinguishing two hallmarks of how macro-level emergent phenomena are related to their micro-level bases:
(1) Emergent phenomena are dependent on underlying processes.
(2) Emergent phenomena are autonomous from underlying
processes.
These two hallmarks are vague. There are many ways in which phenomena might be dependent on underlying processes, and there are also many ways in which phenomena might be autonomous from underlying processes. Any way of simultaneously meeting both hallmarks is a candidate notion of emergence.
For the term emergence to be useful, we must relate the concept to what we believe occurs in nature. When we use the word emergence as an adjective to describe something such as emergent levels of nature, what we generally mean is that there’s something about a phenomenon that we can’t express by simply talking about the underlying processes. The question then is, “What makes a phenomenon emergent?” That question isn’t one about numbers of angels dancing on pin heads. The language used to describe emergence has to reflect what is happening in nature. The concepts of weak and strong emergence describe something that is happening in nature and they provide a conceptual framework that limits what we believe can and can not happen.
Levels of nature can be either weakly or strongly emergent, but weakly or strongly emergent phenomena need not be levels of nature. Dynamical or nonlinear physical systems are other examples of how some folks have categorized emergent systems. I think the right way to look at emergence however, is to categorize concepts of levels of nature and concepts of dynamical systems as being concepts that display either weak emergence or strong emergence.
Bedau describes weak emergence as follows. I’ve tried to boil it down to the bare minimum. For brevity, it might be best to skip the long quote and come back to it:
The system's global behavior derives just from the operation of micro-level processes, but the microlevel interactions are interwoven in such a complicated network that the global behavior has no simple explanation. The central idea behind weak emergence is that emergent causal powers can be derived from micro-level information but only in a certain complex way. As Herbert Simon puts it, given the properties of the parts and the laws of their interaction, it is not a trivial matter to infer the properties of the whole (1996, p. 184). ...
The strengths and weaknesses of weak emergence are both due to the fact that weak emergent phenomena can be derived from full knowledge of the micro facts. Weak emergence attributes the apparent underivability of emergent phenomena to the complex consequences of myriad non-linear and contextdependent micro-level interactions. These are exactly the kind of micro-level interactions at work in natural systems that exhibit apparent emergent phenomena, so weak emergence has a natural explanation for these apparent emergent phenomena. Weak emergence also has a simple explanation for the two hallmarks of emergence. Weakly emergent macro phenomena clearly depend on their underlying micro phenomena. So weak emergent phenomena are ontologically dependent on and reducible to micro phenomena; their existence consists in nothing more than the coordinated existence of certain micro phenomena. Furthermore, weakly emergent causal powers can be explained by means of the composition of contextdependent micro causal powers. So weakly emergent phenomena are also causally dependent on and reducible to their underlying phenomena; weak emergence presumes causal fundamentalism. (More on this below.) At the same time, weakly emergent macro phenomena are autonomous in the sense that they can be derived only in a certain non-trivial way. In other words, they have explanatory autonomy and irreducibility, due to the complex way in which the iteration and aggregation of context-dependent micro interactions generate the macro phenomena. …
The fundamental micro-level causal dynamics of the system - its “physics” - is captured in a set of explicit rules for how the state of a micro entity changes as a function of its current state and the current states of its local neighboring entities. Macro entities and their states are wholly constituted by the states and locations of their constituent micro entities, so the causal dynamics involving macro objects is wholly determined by the underlying micro dynamics. Thus, causal fundamentalism reigns in such a system; macro causal powers are wholly constituted and determined by micro causal powers. The micro dynamics is context sensitive since a micro entity's state depends on the states of its micro-level neighbors. The context sensitivity of the system's underlying causal dynamics entails that understanding how a micro entity behaves in isolation or in certain simple contexts does not enable one to understand how that entity will behave in all contexts, especially those that are more complicated. Locally reducible systems are those that meet all the conditions spelled out in this paragraph.
The notion of weak emergence concerns the way in which a system's micro facts determine its macro facts. A system's micro facts at a given time consist of its micro dynamic and the states and locations of all its micro elements at that time. If the system is open, then its micro facts include the flux of micro entities that enter or leave the system at that time. Its micro facts also include the micro-level accidents at that time, if the system's micro dynamics is nondeterministic. Since causal fundamentalism applies to locally reducible systems, the micro facts in such systems determine the system's subsequent evolution at all levels. Given all the system's micro facts, an explicit simulation could step through the changes of state and location of each micro element in the system, mirroring the system's micro-level causal dynamics. Since macro entities and states are constituted by the locations and states of their constituent micro entities, this explicit simulation would reflect the evolution over time of the system's macro facts. Such an explicit simulation amounts to a special kind of derivation of the system's macro properties from its micro facts. It is an especially “long-winded” derivation because it mirrors each individual step in the system's micro-level causal dynamics. A locally reducible system's macro properties are always derivable from the micro facts by a simulation. However, in some situations it is possible to construct a quite different “short-cut” derivation of a system's macro properties, perhaps using a simple mathematical formula for the evolution of certain macro properties arbitrarily far into the future. Such short-cut derivations are the bread and butter of conventional scientific explanations. They reveal the future behavior of a system without explicitly simulating it.
It is now easy to define weak emergence. Assume that P is a nominally emergent property possessed by some locally reducible system S. Then P is weakly emergent if and only if P is derivable from all of S's micro facts but only by simulation. …
A core concept of weak emergence concerns properties that in principle are underivable except by finite feasible simulation. A slightly weaker notion of emergence concerns properties that in principle are derivable without simulation, but in practice must be simulated. A slightly stronger notion of emergence concerns properties that are underivable except by simulation, but the requisite simulation is unfeasible or infinite. …
It is important to recognize that my notion of weak emergence concerns how something can be derived, not whether it has been derived. It concerns which derivations exist (in the Platonic sense), not which have been discovered. Perhaps nobody has ever worked through a short-cut derivation of some macro property. Nevertheless, if there is such a derivation, then the macro property is not weakly emergent.
Bedau is focused on the following points regarding weak emergence. Weakly emergent phenomena are characterized as follows:
1. Macro level properties of the system are derivable from the micro-level facts (ie: physical states). So the macro level is derivable from the micro-level.
2. The micro-level causal dynamics of the system are governed by an explicit set of rules.
3. The rules governing the micro-level causal dynamics are based on local interactions between micro-level ‘parts’.
4. I would add that because rules governing the behavior of parts are local in nature, the interactions between parts can only propogate at some rate dictated by some rate of propogation.
When talking about the “micro-level”, Bedau only means some physical level which is smaller than the larger “macro-level”. If the macro-level is that of a neuron, the micro level could be the compartments of axons or dendrites. If the macro-level is that of a brain, the micro-level might be the entire neuron. If the macro-level is that of a galaxy, the micro-level might be that of a solar system.
Another feature of weak emergence is that these micro-level parts only interact locally. Weak emergence does not allow for non-local interactions between the parts.
This conception is very similar to, and in complete agreement with, separability.
Although Bedau doesn’t actually state that weak emergence only applies to phenomena that can be described at the level of classical physics, and although I don’t know if Bedau means for it to only apply at that level, it would seem fairly obvious that weak emergence is only applicable to classical physics. At least for classical physics, weak emergence is a perfectly valid conception.
Notice that Bedau goes off in another direction that may not be obvious. He mentions the following:
1. Weakly emergent phenomena are underivable and can only be predicted by simulation.
2. Many phenomena today are not yet derivable except by simulation, but future work might create a “short cut” suitable derivation. Such phenomena are not weakly emergent.
By derivation, I believe he means that you will find many examples in the sciences of equations and methods of analysis that can predict a phenomenon at some macro-level without needing to analyze everything that is happening at some lower micro-level. A good example might be what happens to a gas when subjected to work or heat. The macro-level in this case regards the thermodynamic descriptions of gas properties such as temperature, pressure, internal energy, entropy, etc… The micro-level in this case would be the individual gas molecules which might have properties of mass and some distribution of momentum. The thermodynamics of a gas would NOT be weakly emergent from molecular properties as near as I can tell. Another quick example might be one dimensional pressure loss due to flow through pipe as calculated by using the Hagen-Poiseuille equation which can be derived from the more general, three dimensional Navier-Stokes equations. So pressure loss through a piping network is not weakly emergent since we can predict the macro-phenomenon without resorting to simulation of the underlying micro-level parts. Note that Bedau also suggests that if some time in the future, some new derivation between micro and macro levels arise, that phenomenon would not be considered weakly emergent. We might call these phenomenon emergent, perhaps nominally emergent, but when a simulation is not require, the phenomenon won’t be weakly emergent.
I wonder if Bedau has bridge laws in mind when refering to these short cut derivations. We’re highly unlikely to find bridge laws for example between economics and the underlying physics, so perhaps all of economics is weakly emergent. The same could be said for most of the higher level sciences such as psychology, sociology, and other sciences relating to human behaviors but not necessarily to the sciences that study large scale phenomenon such as meteorology or cosmology. I suspect Bedau simply wants to stay away from bridge laws and the controversy associated with them, but I don’t know.
The concept of simulation seems to refer to the development of software in the past roughly 50 to 60 years which relies on finite elements. Bedau talks about the Game of Life extensively which is essentially a 2 dimensional system of finite elements. Finite element analysis (FEA) as used in the sciences however, takes some small volume of space which Bedau might call the micro-level. FEA assumes the essential properties in that element are finite as opposed to varying, and linearizes otherwise nonlinear equations that describe the element properties. By doing this, a numerical analysis can be performed on the system (ie: macro-level). This basic philosophy of how systems can be broken down into their micro-level constituents has been used for phenomena as small as tiny portions of a neuron’s dendritic tree to phenomena as large as weather systems. The same basic philosophy allows interactions between different types of phenomena such as heat transfer, fluid motion, stresses in solid bodies, electromagnetic interactions and many others. Such software is often called multi-physics software because of its ability to integrate widely disparate causal influences. So the concept of simulation as defined by weak emergence is widely used, well understood, and follows from the basic concept of separability of classical physics.
Bedau and others have suggested that the alternative to weak emergence is strong emergence which is characterized by downward causation.
The most stringent conception of emergence, which I call strong emergence, adds the requirement that emergent properties are supervenient properties with irreducible causal powers. These macro-causal powers have effects at both the macro and micro-levels, and macro-to-micro effects are termed “downward” causation. We saw above that micro determination of the macro is one of the hallmarks of emergence, and supervenience is a popular contemporary interpretation of this determination. Supervenience explains the sense in which emergent properties depend on their underlying bases, and irreducible macro-causal power explains the sense in which they are autonomous from their underlying bases. These irreducible causal powers give emergent properties the dramatic form of ontological novelty that many people associate with the most puzzling kinds of emergent phenomena, such as qualia and consciousness. In fact, most of the contemporary interest in strong emergence (e.g., O'Conner 1994, Kim 1992, 1997, 1999, Chalmers 1996) arises out of concerns to account for those aspects of mental life like the qualitative aspects of consciousness that most resist reductionistic analysis.
The supervenient causal powers that characterize strong emergence are the source of its most pressing problems. One problem is the so-called exclusion argument emphasized by Kim (1992, 1997, 1999). This is the worry that emergent macro-causal powers would compete with micro-causal powers for causal influence over micro events, and that the more fundamental micro-causal powers would always win this competition. … The exclusion argument aside, the very notion of strong emergent causal powers is problematic to some people. By definition, such causal powers cannot be explained in terms of the aggregation of the micro-level potentialities; they are primitive or brute natural powers that arise inexplicably with the existence of certain macro-level entities. … strong emergence should be embraced if it has compelling enough supporting evidence. But this is where the final problem with strong emergence arises. All the evidence today suggests that strong emergence is scientifically irrelevant. Virtually all attempts to provide scientific evidence for strong emergence focus on one isolated moribund example: Sperry's explanation of consciousness from over thirty years ago (e.g., Sperry 1969). There is no evidence that strong emergence plays any role in contemporary science. The scientific irrelevance of strong emergence is easy to understand, given that strong emergent causal powers must be brute natural phenomena. Even if there were such causal powers, they could at best play a primitive role in science. Strong emergence starts where scientific explanation ends.
Note that in defining strong emergence and downward causation, Bedau and others do so in reference to parts. Whether they are micro-level parts or the parts go by some other name, the conception is first that of a system made of parts. In fact, there’s an unwritten axiom which must be inferred; a strongly emergent system must have clearly definable parts. To do that, the parts must be separable. If the parts are not separable, we can’t define the system as having distinct parts.
Strong emergence makes sense when we consider phenomena which supervene on systems which can be defined by classical physics because those systems are separable and thus have clearly defined parts. For those phenomena, strong emergence is a concept that simply doesn’t fit because of the role downward causation must play. Imagine for example, a neuron that does not react to inputs from its immediate neighbors but instead, reacts because of some overall physical state that the brain is in. Neurons change state due to local inputs to the neuron which are fully sufficient to define how the neuron will change state, so per causal exclusion, any emergent property of the brain would be a macro causal power which would have to compete for causal influence over micro-level causal powers but it would be the micro-level causal powers which would clearly win that battle each and every time.
The problem of causal exclusion is a serious issue but in addition, problems such as violations of conservation of energy, violations of conservation of mass or momentum and other seemingly ‘magic’ phenomena might appear due to downward causation. There actually are a few papers that try to defend against this second problem with downward causation by appealing to points of “bifurcation”. However, these arguments only deserve a passing glance as the authors appear utterly devoid of an understanding of the natural sciences.
Strong emergence however, would not apply to nonseparable systems since they have no distinct parts. We wouldn’t for example, suggest that a pair of entangled photons have clearly defined states. They are not clearly defined parts. When the state of one photon is observed, the second photon will collapse to a physical state dictated by the first photon’s state. Any quantum mechanical system which exhibits nonseparability will similarly not be definable in terms of strong or weak emergence because there are no distinct parts. The concept of strong emergence as it applies to quantum mechanics however seems a poor choice in my opinion, primarily because it takes as it’s basis the concept of parts as discussed earlier.
Per Dequet, another conception of emergence regards levels of nature, so I’m going to switch gears here and go down this road for a moment. There’s been quite a bit of discussion around levels of nature being emergent but I think it still boils down to whether or not levels of nature are weakly emergent or strongly emergent.
Imagine some higher level phenomenon such as economics depending on underlying physical processes but in some way being autonomous from those processes. We talk about things in economics such as the law of supply and demand or Gresham’s law for example that help define how economic systems work. In some way, it would seem economics is independent of underlying physical processes. It’s hard to imagine how the value of money and the resulting monetary exchanges depend on quantum mechanics. This type of nonreductionism is pervasive to the point that Kim says:
Expressions like “reduction,” “reductionism,” “reductionist theory,” and “reductionist explanation” have become pejoratives not only in philosophy, on both sides of the Atlantic, but also in the general intellectual culture of today. They have become common epithets thrown at one’s critical targets to tarnish them with intellectual naivete and backwardness. To call someone “a reductionist,” in high-culture press if not in serious philosophy, goes beyond mere criticism or expression of doctrinal disagreement; it is to put a person down, to heap scorn on him and his work.
Kim of course, is a reductionist who I generally agree with. Looking at levels of nature as being emergent in any way that is more than weak is, in my opinion, only to attempt an explanation per strong emergence.
Kim claims emergentism is now commonly called “nonreductive materialism” and that in the past few decades, the concept has been that there are emergent levels in nature that are autonomous in some way. I think that’s a very big part of it. Consider some of the lower level physical sciences such as chemistry, thermodynamics and fluid mechanics. In those cases, it might be relatively easy to see phenomena described at those levels being dependent on the underlying physics such as quantum mechanics. But going up a level to biology, the underlying physics becomes a bit more tricky. By the time we get up to psychology, sociology or economics, it may take a leap of faith that phenomena seen at these levels are completely determined by the lower level physics, especially quantum mechanics. I would however contend that these higher levels of science are weakly emergent on the lower levels.
There are some brilliant physicists, especially condensed matter physicists such as Anderson and Laughlin, who point out that some phenomena such as superconductivity, “like the fractional quantum Hall effect, is an emergent phenomenon – a low-energy collective effect of huge numbers of particles that cannot be deduced from the microscopic equations of motion in a rigorous way and that disappears completely when the system is taken apart.” I have no quarrel with that. The examples you’ll find are all nonseparable. But to extend the concept of emergence as these brilliant physicists have done to higher levels of nature is a mistake. Once the phenomena in question have risen above the borderline where classical physics can take over from quantum mechanics, we have separable systems that can no longer exhibit strong emergence nor any kind of downward causation.
Summary:
The concept of emergence has 2 basic forms, weak and strong. These concepts can be applied to levels in nature and also to dynamical systems, so the concept of weak and strong emergence is more fundamental than the concept of emergent levels of nature or emergence as it applies to dynamical systems. However, weak and strong emergence are only applicable to classical physics as they require clearly defined micro-level parts and those parts must be seprable. Nonseparable systems are not weakly emergent but they are not really strongly emergent either. There is a clear lack of philosophical discussion around nonseparability as it pertains to emergence. However, for separable systems, it seems clear that we can apply those concepts of weak and strong emergence.
Anderson, P. W. (1972). More is different. Science, 177(4047), 393-396.
https://www.tkm.kit.edu/downloads/TKM1_ ... nt_PWA.pdf
Bedau, M. (2002). Downward causation and the autonomy of weak emergence. Principia: an international journal of epistemology, 6(1), 5-50.
https://periodicos.ufsc.br/index.php/pr ... 7003/15556
Deguet, J., Demazeau, Y., & Magnin, L. (2006). Elements about the emergence issue: A survey of emergence definitions. Complexus, 3(1), 24-31.
http://citeseerx.ist.psu.edu/viewdoc/do ... 1&type=pdf
Laughlin, R. B. (1999). Nobel lecture: fractional quantization. Reviews of Modern Physics, 71(4), 863-874.
http://www.nobelprize.org/nobel_prizes/ ... ecture.pdf