# Scientific evidence of reductionism

There are claims that the standard model is a theory which explains almost all phenomenon that we see in the world. I am wondering what scientific evidence backs this claim?

Specifically, there is a notion that the universe is made of subatomic particles/ fields, (I will use the words “particle” and “field” interchangeably here,) and the interactions of these fields — though not known completely — determine all behavior of matter. To what degree is this claim scientific?

It seems that there are a few parts to this claim:

1. All matter is ultimately made up entirely of subatomic particles.
2. All interactions between matter happen between individual particles.
3. There are no fundamental bulk effects, i.e. there are no fundamental interactions between groups of particles.
4. Any macro-scale properties are emergent properties coming from interactions at a smaller scale.

What physical evidence is there for these beliefs? Just because we find laws governing the relationships between particles, and find them to be constituents of matter at larger scales, what evidence do we have for this claim of total reductionism? What evidence do we have that there exist “building blocks” of matter, and that all matter is entirely determined by those building blocks?

• I think what you are asking for is too broad, because all experimental evidence is consistent with the atomic model — that is, that bulk matter is composed of atoms. And all experimental probes of atoms are consistent with a model of atoms made up of electrons and nuclei. And so on, on down in length scale. If you could explain what specifically you take issue with, we could perhaps point out more specific experimental evidence.
– d_b
Jun 25 at 1:04
• I have never seen anyone explain how a car engine works, or why a heart beats, using particle physics. The idea is that there is an apparent causal chain between scales all the way down to a smallest level. But, the thinking goes, the links in the chain are only illusory. Only links at the smallest level exist, and all the other chains are “made up” of chains at the smallest level. Therefore, if one knows in detail the workings at the smallest level, they can reverse-engineer knowledge of all levels. This seems like a pretty strong claim. I want to know if it has scientific backing.
– Eoin
Jun 25 at 1:10
• Haven't you ever taken a biology course where some phenomena were explained by resorting to chemistry? Or a chemistry course where some phenomena were explained by resorting to physics? Of course reductionism has been passing tests since ancient times. Someone might take the time to post detailed examples here. But what do you think is the most likely response when you ask a bunch of scientists whether their work is really religion? Jun 25 at 2:44
• Are you asking for examples of phenomena which are not fully understood from a reductionist viewpoint, or are you asking for positive evidence that there exist collective phenomena which cannot be understood reductively? Essentially every open problem in every scientific field is an example of the former, and I suspect the latter would be Nobel-worthy to say the least. Jun 25 at 3:03
• The evidence is from solving microscopic equations and investigating the resulting emergent phenomena at larger scales. A comparison of these solutions with natural phenomena generally shows good agreement. Cases that are not in agreement usually point to missing microphysics or poor approximations in obtaining solutions. However, it may not be bad to begin making a survey of constraints on violations of reductionism, and to design experiments to extend those bounds. Jun 25 at 5:34

I think there are two different questions here:

1. Do we have the proof that everything is made up of subatomic particles and all the processe/interactions are reducible to them
2. Can all the phenomena be explained in terms of the interaction of subatomic particles

The answer to the first question is clearly affirmative, as has been shown in physics experiments on all levels of organization of matter: subatomic particles, nuclei, atoms, molecules, solids/liquids/gases, etc.

The second question is trickier, since we are dealing here with the emergent phenomena where the matter exhibits behaviors very different from what one could have predicted mechanistically from the knowledge of the lower levels of organization. On the other hand, as already stressed above, the matter engaging in these complex new behaviors is still composed from elementary particles and obeys the fundamental laws of physics. This was very much the subject of the debate around the question Does physics explain why the laws and behaviors observed in biology are as they are? - people seem to disagree strongly about this subject.

Note however that one need not even go as far as biology to encounter difficulties with reductionist view: statistical physics and the theory of critical phenomena (such as phase transitions) were pretty much developed in response to the impossibility of explaining some phenomena using microscopic description. Depending on their background, some would argue that this is simply the consequence of limited computational capacity, which can be overcome in the future, whereas others would say that this limitation is unsurmountable or that it would add little in terms of explaining/understanding these complex behaviors.

Update:
As an example of reducing vs. explaining one could consider a written text.

• Such a text can be reduced to the letters, punctuation marks and spaces composing it, and the interactions between those could be studied in details - e.g., the statistical distribution of letters or their sequences, which may obey some very strong rules for a specific language, where some sequences are prohibited, others are common. Using these rules one could try to construct a text, which would look like English... but still be gibberish. Examples can be found, e.g., in the Shannon's seminal paper.
• One could work on a higher level - that of words. We cannot say for sure which allowed combinations of letters make a word - we cannot explain words, even though we have detailed knowledge of what they are made of. However, once we have a dictionary, we can study words, their interactions, etc. We could try to use this knowledge to write a text... and it still would be gibberish.
• We could study sentences - there are rather strict rules governing srntence structure, known as generative grammar. Programming (artificial) languages are fully explained by these rules and new languages can be constructed at will. But they will still be a far cry from a real human language, even though we understand their components in details.
• You can't actually prove anything with a universal quantifier in a scientific theory. Only math can do this with its pure objects. Jun 25 at 13:39
• @Ruslan Only if you use universal quantifiers in a mathematical sense :) Jun 25 at 14:16
• If the answer to number (2) is not affirmative, then there may be phenomenon that cannot be explained in terms of interactions between subatomic particles. If this is the case, it seems a leap of faith to assume that the phenomenon is still reducible to the interactions between subatomic particles (1). “I can’t explain it using this toolset, but lets just say this toolset explains it anyways.” Therefore, not having (2) undermines (1). Therefore, I think the answer to (1) is “not clearly affirmative”.
– Eoin
Jun 25 at 15:03
• @Eion Have you looked at the discussion around the question linked? E.g., a computer is made of atoms, but can you explain or reduce your code to the interactions between these atoms? Or can you explain the content of your PhD thesis in terms of the letters composing it? Jun 25 at 15:09

A brief answer could be: all stuff is made of particles, but we don't know (fully) what particles are.

In a slightly more full answer, what I called 'particles' just now are already quite subtle because the mathematics describing their physical behaviour is the mathematics of quantum field theory, and so it is not quite right to say the constituents of the cosmos are just 'particles'; they are something more subtle, something like a collection of quantum fields.

When it comes to collections of such 'particles' there can emerge, in their collective behaviour, types of behaviour which we do not know how to connect to the underlying field theory of the particles taken one at a time or in small groups. Faced with this, we don't need to abandon reductionism right away, because it is reasonable to think that such collective behaviours are consistent with, or the outworking of, whatever is the correct description of the basic constituents. But it is not reasonable to think that we already know that correct description in all detail and precision; we certainly do not. And it is also not reasonable to think that reductionism always succeeds in all respects, because the phenomena relating to quantum entanglement give counter-examples.

That last claim has to be considered carefully. What is true is that if we take 'reductionism' to entail that physical predictions concerning spacelike-separated parts can be correctly obtained by considering the parts as if they carried their properties with them separately, then reductionism is disproved by Bell-inequality-type observations. But this should not be taken to mean that reductionism is completely undermined. It is a subtle limit to reductionism, whose impact on wider phenomena is as yet unclear. It seems to be involved in some many-body phenomena, and it can be argued, convincingly I think, that it lies behind the computing power of quantum computing.

I notice that some comments to the original question have reacted as if reductionism equates to science, as if to question the first is to question the second. I guess one is right to be nervous of appearing to open a door to unscientific reasoning. But to underline the point I am making here about entanglement, I would invite further consideration of either the example of quantum computing, or the example of Cooper pairs in BCS superconductivity, or comparable examples in other collective phenomena. The evolution over time of two registers in a quantum computer performing Shor's algorithm cannot be described by any description which adopts the language 'register A is thus; register B is thus'. One can only say 'registers A and B are thus'. The Cooper pair is a component of a reductive description, but when one says that such a pair consists of 'two electrons' it is not fully correct to say that there is some X of which the Cooper pair is 'two X' in any ordinary meaning of the word 'two'. The mass and charge are 'two $$m$$' and 'two $$e$$' but the momentum is not 'two' anything. Almost all energy eigenstates of electrons in atoms are also highly entangled, and only by loose use of language do we say there are 'two electrons' rather than 'one entangled electron pair' in helium.

Finally, then, the issue becomes what a more complete fundamental theory would be like. Would it keep open the subtle door to non-reductive possibilities that has already been opened in quantum entanglement? Are there even more rich non-reductive possibilities actually at work in the cosmos, right at the level of the descriptions of the basic constituents? We do not know, but I would hazard a guess that the answer is yes.

• Interesting and good example of non-reductionism of entangled pairs of particles.
– Eoin
Jun 25 at 14:57

There is no such thing as "evidence of reductionism" in your sense, any more than there is "evidence of falsificationism" or "evidence of positivism". Reductionism is a philosophical standpoint, not a property of the world we could ever be sure of.

You can have evidence that reductionism works in specific cases - e.g. the theory of the Standard Model which in turn generates the theory of residual nuclear forces - and all natural sciences are full of cases like that. But in any case where you do not have a reductionist explanation, you have a philosophical choice - you can say this is a challenge for reductionist approaches to tackle, or you can believe it cannot be explained in more reductionist terms. In the latter case, you have two more choices - either this is another "fundamental" thing to add to the set of explanations in terms of which reductionist approaches explain everything else, or you can declare victory over reductionists and say reductionism is dead.

There is no way to prove something cannot be explained in terms of some as of yet unknown reductionist theory. You can only prove this for specific reductionist theories that already exist. There is equally no way to prove something can be explained in reductionist terms without actually providing that explanation. And even if all known phenomena have been explained in reductionist terms we cannot be certain we won't one day discover a new phenomenon that cannot be.

For the project of natural science, whether you are trying to "prove" or "disprove" reductionism when you investigate possible explanations of a phenomenon is entirely irrelevant - in either case you're trying to produce scientific explanations of the world.

However, note that the above uses the word "evidence" as it is used in the question, but not in the way science usually does. When we talk about evidence for a scientific law, we don't say we don't have evidence that it holds just because we have only observed it to hold in specific cases. The more cases we observe where the law holds, the more certain we become the law is "true", until we observe an instance where it does not. The problem with reductionism is that it is not falsifiable in this sense - we cannot observe an instance where it does not hold, since we can always believe we just haven't figured out the correct reductionist explanation yet. This is why reductionism is a metaphysical belief rather than a scientific property of the world.

I'll divide this answer into two parts. In the first, I wish to dispel/prevent any possible category errors here.

Ground rules from the philosophy of science

There are claims that the standard model is a theory which explains almost all phenomenon that we see in the world. I am wondering what scientific evidence backs this claim?

To say a theory explains observed phenomena is to say those phenomena are among its predictions. The evidence for this claim is, of course, the phenomena themselves.

To what degree is this claim scientific?

By "scientific", you might mean "amenable to scientific analysis" or "well attested in the light of such analysis". These respectively apply to falsifiable claims and claims whose predictions concord with observation. So, as aforesaid, the SM is scientific on both counts.

It seems that there are a few parts to this claim

"This claim" meaning what the SM claims, rather than the previously discussed claim that the SM explains observed phenomena. The parts you identify in dissecting the SM are how we make the aforementioned predictions. To take a much simpler example, the claim that people have minds with certain internal states, responsive to and influencing their bodies in certain ways, explains people's observed behaviour. You can't see minds or quantum fields directly, just their observable effects.

SM-related phenomena specifically

Warning: I'll mention some lines of evidence, but each is mentioned only very briefly.

There are no fundamental bulk effects, i.e. there are no fundamental interactions between groups of particles. Any macro-scale properties are emergent properties coming from interactions at a smaller scale.

That seems a fair summary of what's claimed. In other words, (i) interactions between composite particles are an emergent consequence of those between fundamental ones, and (ii) we respectively call these interactions emergent and fundamental. While (ii) is a definition, (i) is a claim about Nature.

What physical evidence is there for these beliefs? Just because we find laws governing the relationships between particles, and find them to be constituents of matter at larger scales, what evidence do we have for this claim of total reductionism? What evidence do we have that there exist “building blocks” of matter, and that all matter is entirely determined by those building blocks?

This splits into two issues we should address in the reverse of the order you raised them: how do we know (in the evidence-fits-predictions sense of know) there are such building blocks, and how do we know fundamental interactions are sufficient to explain interactions between composite structures?

For the first question, we gradually work down to smaller components (e.g. Brownian motion attests to the existence of atoms, chemistry attests to the electronic configurations and NMR spectroscopy to their nuclear structures, observation of beta decay shows neutrinos exist, and the classification of hadrons supports their quark-gluon structure). While one can never prove a particle doesn't have a substructure, the claim that it doesn't predicts it has no discernible size, and that's true so far of electrons and quarks, down to length scales of $$10^{-18}$$ m. Any substructure they may have hasn't been shown to have better predictions of observation.

For the second question, it takes detailed calculations to show fundamental interactions predict atoms' ionization energies and electronegativities, and van der Waal forces. From these follow predictions in chemistry and bulk predictions, such as materials' viscosity, boiling point etc. So far, while many predictions of "bulk forces are emergent" have been successful, none have proven unworkable. We will acknowledge non-reductionist interactions when we find them. (They will only be "not reductionist yet", just in case reduction is eventually possible; for example, this is what happened with friction, elastic forces, vibrations etc.)

Science does not "prove" anything, it can only disprove false models of reality.

We create mathematical models which predict the results of experiments. Models which work well are no more than that - whether they might represent some "underlying reality" is a purely philosophical proposition. The Standard Model is one such, moreover its underlying reality is famously a matter of wildly differing interpretations and interminable debate.

But models which make incorrect predictions may safely be discarded by philosophers, as representing only some false idea of any underlying reality. The particular nature of the falsehood concerned may be open to debate, but the overall failure is unarguable.

Thus, as long as a theory does not produce false predictions, there is no way to show it as either true and accurate or false and a fluke of luck. For example Newton's model of gravity proved accurate for centuries, until science advanced and showed it to be an oversimplification. The local realist model of matter as physical particles, with the waves as pure statistics, is another such failure, victim of Bell's Theorem and subsequent experimental disproof of the theory's predictions. Now we have General Relativity and quantum entanglement instead, and nobody can say for sure whether these models of reality are philosophically right or wrong.

However, there is a small but crucial body of evidence demonstrating that the Standard Model is wrong, or at least is not the whole story. Many physicists are tinkering around, trying to improve it while still explaining the vast body of evidence consistent with the field/particle aspect.

What we do not have is any model of holistic or other non-standard entities with any useful predictive power, nor even the slightest hint of how such a model might be formulated. Without such a mathematical model, philosophers have nothing to get to grips with and, consequently, no reason to take the idea seriously.

This started as a comment, but it's just too long and I turned it into an answer.

Reductionsism as a approach to understanding physical phenomena is currently under challenge from "emergence", i.e. the idea that a complex system cannot be reduced to the study of its invidual parts. Basically, entanglement and quantum correlations between parts of a large system may prevent reduction to the individual subparts.

There's an interesting recent discussion of this in

Aharonov Y, Cohen E, Tollaksen J. Completely top–down hierarchical structure in quantum mechanics. Proceedings of the National Academy of Sciences. 2018 Nov 13;115(46):11730-5.

In this paper (which is not so easy to read because some of the language and background is not familiar to most physicists), the authors present a model of 3 particles in 3 boxes, so that correlations between the 2 particles can be inferred from the 3-particle correlations, but not the other way around (that the "top-down" of the title). If you reduce the system to the study of 2-particle subsystems, you cannot infer the properties of the whole system. There is no issue of computational power here, and it would therefore seem to be fatal to strict reductionism.

Thankfully the authors also show that quantum mechanics is compatible with any experimental predictions of this model, thus displacing the problem to one of compatibility between quantum mechanics (which can incorporate such correlations - local or not - at least phenomenologically) and reductionism. As others have pointed out, quantum mechanics appears to be where the reductionism buck stops: many-body collective pheonena (v.g. Bose-Einstein condensates) or even few-body effect (photon bunching effects, which can be measured for very few photons) cannot be completely explained in terms of individual-particle effects, yet suggestions that these are evidence against reductionism have never been taken terribly seriously.

• I am not sure that this "strict" approach to reductionism is a good understanding. It is interesting to note in this context the Bohm-Hiley hidden-variable model of particles and pilot waves, and Bohm's concept of the "implicate order" which pervades spacetime and determines the behaviour of the pilot waves; the particles are reducible, the implicate order is not. The main formulation is mathematically equivalent to standard QM, but the implicate order is metaphysics. Jun 29 at 7:50
• @GuyInchbald I am admittedly not an expert at this stuff but I don’t quite understand your comment. The paper of Aharonov clearly shows that you cannot infer the properties of the 3-particle system from those of any subsystem alone (at least that’s how I read the paper), yet the properties of the system can be understood from rules applied to its constituents. (I hope my use of language is correct). I could be wrong but I do not see how this is in opposition with your short discussion of the Bohm-Hiley apparoch (of which I know not much) Jun 29 at 12:17
• My point is that there are many forms/degrees/definitions of reductionism. The existence of a three-particle state which cannot be reduced to interacting couplets invalidates only the very strictest forms. The OP notes that in the Standard Model, which explains the three-body correlation, "There are no fundamental bulk effects, i.e. there are no fundamental interactions between groups of particles." The strictest forms of reductionism are, as you have pointed out, not consistent with such three-particle correlations. They are therefore not what the OP is asking about. Jun 29 at 15:50
• @GuyInchbald Yes I suppose you are right although I interpreted the question as implicitly applicable more broadly. As pointed out in the preamble to my “answer”, I initially meant this as comment to address the broader issue of reductionism in physics rather than in particle physics. Jun 29 at 16:49