Talk:Emergence (biology)

Skepticism

Hi Anthony:

While it is beyond debate that systems can be described in language that has no meaning when applied to their constituent subsystems, things like temperature and possibly consciousness, the idea of emergence seems to be just another name for these system properties with the addition of some "magical" elements about the amazing appearance of new properties. To my mind the microscopic behavior of atoms in a gas is more fundamental than some average that we call "temperature", or some property that we call "heat". It is obvious that thermodynamics gets along with such concepts just fine (within its domain of applicability, which excludes things like fluctuations about the mean), and it doesn't have to refer to its underpinnings in statistical mechanics. However, the use of thermodynamics to explain a situation instead of a complete microscopic analysis based upon atomic motions or maybe the Standard model is simply an economy of thought, made necessary by the limited capacity of the human mind and its computer agents, and not the emergence of a whole greater than the sum of its parts.

I'd argue that the notion of " inexplicably unpredicted novel properties, functions and behaviors, ones not observed in the system's subsystems and their components, and not explainable or predictable from complete understanding the components' properties/functions/behaviors considered in isolation from the system that embeds them." is a set with zero members. John R. Brews 14:00, 27 August 2012 (UTC)

It's an argument that is in danger of descending into pedantry, around what we mean by "complete understanding". For computational biologists, emergent behaviour is complex "high level" behaviour that was not explicitly designed into the system, and which arises only above a critical level of complexity. There are lots of proposed examples - and the examples are didactically important as narratives of how apparent 'design' in a complex system can arise/evolve/emerge 'spontaneously' once a system gets complex enough. In addition, for biologists generally there is a common problem with inferring causality. In health sciences we are used to thinking of risk factors rather than causes of disease in acknowledgement of the fact that predicting who will get a disease, when they will get it and its likely course may be possible for populations but not for individuals. Complex organismal phenotypes, including those that confer susceptibility or resistance to disease, often look (for the moment) like emergent properties of hugely complex gene networks. Whether these can ever be subsumed into a conventional causal narrative is unclear - the "causes" may be causes only in the sense of the butterfly's wings causing a tornado.Gareth Leng 08:34, 29 August 2012 (UTC)

Gareth: Your remarks seem to me to straddle both sides of this argument, suggesting on the one hand that "high level" behavior can be shown to arise spontaneously from models involving only subsystem interactions and no additional interactions introduced by "global" connections, and on the other hand that (perhaps out of inability) we are forced to make predictions based upon global considerations that have no cause traceable to subsystem interactions. So is it your view that the issue is undecidable, suggesting a perhaps a different phrasing of the issues? John R. Brews 14:51, 29 August 2012 (UTC)

Here is a discussion of "emergent" gravity as a macroscopic version of a more microscopic theory. Here is a more general presentation of the relation between macro-theories and micro-theories.

in contrast we have this from Answers.com ( a reputable source?):

"Macroevolution, in all its possible meanings, implies the emergence of new complex functions. A function is not the simplistic sum of a great number of "elementary" sub-functions: sub-functions have to be interfaced and coherently integrated to give a smoothly performing whole. In the same way, macroevolution is not the mere sum of elementary microevolutionary events."

I am afraid that this kind of "thinking" flourishes in the mystical environment of emergence. John R. Brews 14:32, 27 August 2012 (UTC)

And we have this:

"In the absence of a dialectical understanding of human history, the question whether individuals create social reality or vice versa continues to shape sociological theory construction. Is society a sui generis, transcendent reality which coercively shapes human behavior or is it, instead, simply equal to the sum of individual actions? Can social facts be explained only by other social facts or does explanation require, to be valid, that social facts be reduced to micro-level explanations?"

which has the merit of posing a question, not making an assertion. John R. Brews 15:06, 27 August 2012 (UTC)

Why emergence?

I am concerned that the section Why emergence?, and perhaps this entire article, is phrased too much in the manner of presenting an established predominating view, when it should be presented as simply one point of view, and contrasting views should be presented as well. Although a vast number of proponents can be cited, in my opinion this topic suffers from the exploitation by some of ambiguity and vague terminology to advance mysticism. John R. Brews 14:49, 27 August 2012 (UTC)

This section is far too assertive about what is mere conjecture. The statement:

"One reason: the intrinsic properties of a system’s components cannot themselves determine those of the whole system; rather, their 'organizational dynamics' does — how the components interact coordinately in time and space. Those organizational dynamics include not only the interrelations among the components themselves, but also interactions among the many different organizational units in the system."

is a bald assertion without foundation. To phrase what is said above a bit differently, the assertion is that the properties of the parts deciding their reaction to inputs are an incomplete description, and that when the parts are juxtaposed they both generate and respond to novel forms of interaction. Maybe there is an example of such a system of entities, but I don't think so. The source describing "organicism (materialistic holism)" is hokum, I'm afraid. John R. Brews 16:56, 27 August 2012 (UTC)

The query:

"Why do not all of the properties/behaviors of a living system predictably result from the properties of its components?"

is not just a question, but a presupposition of facts not in evidence, definitely disprovable in some cases, and perhaps in general. John R. Brews 15:17, 29 August 2012 (UTC)

Role of closed systems

The example provided of studying a protein separated from the cellular system that embeds it in a cell, and the proposed inadequacy of this study to "explaining any of the properties it has in the context of the system that embeds it" may point simply to the failure to properly identify the "system". This idea of how to identify a closed system may be a better entry point into "emergence" than the identification of magical properties.

If two systems are coupled, each may impact the other, and the analysis of the pair must involve this coupling. We have a feedback arrangement, and although each system may be completely described in terms of its response to arbitrary inputs, when these inputs are generated by another system that responds to the first, the solution to this coupled-system behavior may be much easier to describe in new terminology.

An example might be the vibrations of a crystal lattice, usually described in terms of phonons, entities describing vibrations of the crystal as a whole. It remains the case however, that a description in terms of individual atomic properties can be recovered from the phonon picture, and the two descriptions are simply different perspectives on the same phenomena, of which one or the other may be more practical in a given situation, somewhat analogous to an architect's sketch being more useful in presenting a proposal to a layman than a mechanical drawing of its floor plan. John R. Brews 15:27, 27 August 2012 (UTC)

Some changes

I've made a few changes to separate the concept of emergence from the characterization as "unpredictable", to leave that as a possible but not required property of emergence. John R. Brews 16:40, 27 August 2012 (UTC)

Water as an example

The question posed in the section Water as an example are biased toward the "surprising" nature of water when viewed simply as hydrogen and oxygen. However, I suspect that most properties of water, including its dielectric behavior (transparency) and its hydrodynamics (e.g. vortex formation) all are entirely predictable using modern physics and chemistry. These questions should be rephrased to allow that an entirely reductionist explanation is not only possible but already extant. John R. Brews 17:05, 27 August 2012 (UTC)

Recommend change of title

Why have we an article called Emergence (biology)? I suspect that an article Emergence would be better and could include any biological applications as a subsection. That would remove the possible proliferation of many "emergence" articles, like Emergence (traffic control), Emergence (phase transformations) and so forth. I suggest a renaming of this article. John R. Brews 15:54, 29 August 2012 (UTC)

Renaming this article could lead to a salutary rewrite that is much needed. John R. Brews 16:53, 29 August 2012 (UTC)

Section: Examples of emergence

The section Examples of emergence repeatedly makes the criterion identifying emergent properties the condition that they cannot be explained based upon subsystem behavior alone. I believe this approach to emergence is too restricted.

For one thing, what we can explain today and what we might explain tomorrow does not particularly distinguish emergence from other areas of experience. For a second thing, it is not necessarily complexity that leads to unpredictability.

The proper (IMO) identification of an emergent property is as a phenomenon described using global terminology, independent of whether the phenomenon also is explicable by a theory based upon subsystem properties alone that do not invoke the global terminology. An emergent property is described as a cooperative activity in cooperative terms, and it is neither here nor there whether it has only a collective description or may have multiple descriptions using different vocabularies. John R. Brews 14:51, 31 August 2012 (UTC)

What is controversial about emergence?

Although the article says emergence is controversial, the controversy is not pointed out explicitly. It may be that this paragraph pinpoints the controversy:

"Emergent properties are viewed by some as novel properties, functions and behaviors, ones not observed in the system's subsystems and their components, and neither explicable nor predictable from even a complete understanding of the components' properties/functions/behaviors considered in isolation. Others take the view that these novel properties are the outcome of interactions between the constituents understandable from microscopic behavior, but more readily envisioned by introduction of novel organizational concepts."

Probably most would recognize emergent behavior when they saw it, although there might be some undecidable cases. The controversy is over whether this behavior is explicable or predictable from the understanding of the component parts. I'd guess two camps are formed: the mystics who say no not even in principle (a basically unprovable position) and the pragmatists who say we don't have the explanation just yet, or it has some failings, but explanation will come with time (also an unprovable position). Inevitably, as sometimes an explanation will be found, the phenomena considered "emergent" by the first camp will continue to be nibbled away at as explanations are discovered, while the second camp will not be so affected.

If this argument over predictability is actually the controversy, it is puerile and should be identified that way. If the controversy lies elsewhere, it should be spelled out. John R. Brews 05:10, 21 September 2012 (UTC)

A third position is that, in some cases, explanation is in principle possible but in practice not (computationally infeasible, or requiring impractical experiments). Also unprovable.

But what's with provability anyway? Neither science nor philosophy is about that. Peter Jackson 09:22, 22 September 2012 (UTC)