Talk:Effective theory

"all theories are effective"

This can be either a deep, or just a smart-ass observation, depending on context. It isn't very useful to say "every theory is effective", because it immediately makes the term superfluous.

In reality, there is a meaningful distinction at every stage of scientific progress. The epicycles are an "effective theory" of planetary motion, but only the heliocentric view gives an actual model which in a meaningful sense explains observational data. But likewise, the simple heliocentric model, with planets wandering around the sun in a certain way simply because God told them to is an "effective theory", while only Newtonian gravity provides an actual model explaining on why the planets move like this. And obviously, Newtonian gravity is an "effective theory" for the low energy limit of general relativity (explaining how Newton's gravitational force "is just" a pseudoforce caused by inertia).

In this sense, an "effective theory" is a term used in contrast to a more satisfying theory which provides a model for the effect which happens to be under investigation.

Also, the point is not that the effective theory is "approximate". You can model the solar system as accurately as you like using epicycles. The step forward is not increased accuracy, but rather a deeper understanding of underlying mechanisms, even if this step forward could possibly mean sacrificing accuracy of calculation (say, because the math becomes less manageable).

A related term is probably "black box". An effective theory contains a black box of some kind, even if the input and output can be modelled perfectly using a sufficient number of unexplained parameters. This is exactly what the standard model does, it has lots of unexplained parameters (particle masses, coupling constants), and is therefore not satisfying even if it is perfectly accurate.

The Turing test is a test of "effective intelligence". Its model of intelligence is "anything indistinguishable from intelligence". This relates to the idea of separating the concepts of form and function, as I find put intelligently here, comparing the Turing test to the statement that "if you can build something indistinguishable from a bird, it must definitely fly, which is true but spectacularly unuseful in building an airplane." --dab (𒁳) 14:39, 15 December 2011 (UTC)[reply]

In QFT, particularly the standard model

Modern doctrine looks the effective character of the Standard Model as an explanation of its renormalizability. This is related to the decupling theorem; for a low energy effective theory to decuple, it must be renormalizable. Then the renormalizability of the SM is just a reflection on the fact that all the high energy parameters have been hidden away, into renormalisable parameters, not measurable at all at this scale. This avoids -bypasses, or makes irrelevant- any discussion about the fundamentality of renormalisation of a theory. Note the difference with Fermi theory, which was non renormalisable and then signaled its own end of validity, at the Fermi scale. — Preceding unsigned comment added by 87.216.217.210 (talk) 00:05, 29 September 2015 (UTC)[reply]

Merge proposal

I propose merging effective field theory (EFT) into effective theory (ET). I think the content in EFT can complement the current ET which is a stub. Another proposal is to do it the other way around, EFT into ET as EFT is much more well discussed topic, however ET is the more general one.ReyHahn (talk) 12:05, 13 November 2024 (UTC)[reply]

Support. Google Books search for "effective theory" mostly returns field theoretical results, so merge seems appropriate. Even those books/articles which do not focus solely on field theory but take a broader view (e.g. [1]) seem to be informed by the field theory philosophy. The current article Effective theory is not based on sources, so "merge" would really mean redirecting ET to EFT and renaming the article. Some less technical introductory material could also be added. Jähmefyysikko (talk) 12:50, 13 November 2024 (UTC)[reply]

Support with two caveats. This page is too short compared to the other, so I would find a home for the Newton piece and skip the rest. But...

There needs to be careful cross-referencing to Effective medium methods, Constitutive equations and similar. At the moment these pages do not have a broad enough view. Please think Physical sciences not just Physics.
I think significant care is needed to clarify the difference between approximation and linearization/quasiparticles in these approaches. For instance, to me quasiparticles are "real" Hamiltonian eigensolutions with momentum etc, whereas this page currently calls them "effective particles".

N.B., a WP search for "Effective" brings up a list of other pages to manage. Ldm1954 (talk) 14:02, 13 November 2024 (UTC)[reply]

Support Based on the sources we have, the primary topic is "Effective field theory" with "Effective theory" being a history of science or philosophy of science generalization. I suggest starting with the primary title and adding a section based on the broader view book cited by Jähmefyysikko. Thus "Effective theory" would redirect to that section. Johnjbarton (talk) 17:31, 13 November 2024 (UTC)[reply]

Weak support of the first choice. An effective field theory is one major type of effective theory, but by no means the only type. Harmonic oscillators and Newtonian gravity are also considered effective theories, see for example the book Effective Theories in Physics. Effective theories are also used to help understand complex neural networks, e.g., The Principles of Deep Learning Theory. So my impression is that this concept is used more broadly than just continuum and quantum field theories. Merging EFT into this article makes good sense as a prominent example, but merging this into EFT would just be adding off-topic EFs to the EFT article. I also in principle support a this article as standalone, if sufficient sources conferring notability could be found. --{{u|Mark viking}} {Talk} 19:11, 13 November 2024 (UTC)[reply]