I read in Wikipedia:

Antireductionism is a philosophical and scientific position that ...

It seems to me that Antireductionism cannot be a scientific position as it is in direct conflict with physics which is a part of science(?). Physics makes predictions about things as if they were made of quantum particles. So, if everything is made of quantum particles(?), then the laws of Quantum Field Theory or String Theory would predict something (probabilities) about the future state of the universe given its current state. If this prediction is true(?), then everything is explained by one frame of thinking (Quantum), and there is no need for other "planes" of explanations.

Since so many great philosophers have held/are holding this position, I must have made a mistake in reasoning. So, can somebody please explain this aspect of Antireductionism to me?

NOTE: The parts of the reasoning that seem a possible suspect to me are marked by (?).

  • Some thinking about Godel' theorems might be relevant. 'Reducing' something to probabilities is perhaps a too easy way to criticize antireductionism.
    – sand1
    Commented Apr 6, 2018 at 10:20
  • Well, no. Physics itself doesn't really provide the backdrop of philosophy of science. In other words: you can be a physiscist and against scientific realism. From there on it doesn't seem like a far leap to perhaps take a stance against reductionism. Anyway, the SEP has an article about it. plato.stanford.edu/entries/scientific-reduction
    – Marc H.
    Commented Apr 6, 2018 at 11:38
  • Also note that "There is no need for other planes of explanations" is NOT strictly a reductionists thesis, but already eliminativism. You could think that theories can ultimately reduce down, yet that right now other planes of theories serve their purpose.
    – Marc H.
    Commented Apr 6, 2018 at 11:40
  • 1
    I think your incredulity is due to the loose use of "antireductionism" for widely diverse claims. If all it means is that methodologically "not all properties of a system can be explained in terms of its constituent parts" then this is the situation we currently have in biology, or even chemistry. It is not feasible at this time to derive chemical properties of elements from quantum mechanics, so we have to approach them "holistically", i.e. study the chemistry of elements directly. When one moves from practiced methodology to what is possible "in principle" that is metaphysics, not science.
    – Conifold
    Commented Apr 6, 2018 at 18:36
  • 1
    Everything following QFT wouldn't be inherently reductionist. To predict anything real with certainty, you will still need to predict the entire universe. Things like universal gravitation, nonzero quantum probabilities throughout the field, etc combined with chaos theory directly implies any reduction in the total consideration of the entire universe undermines predictive power totally. Antireductionism.
    – otakucode
    Commented Apr 7, 2018 at 0:16

5 Answers 5


Antireductionism is a scientific position, it fits the mathematical results encountered in places like chaotic dynamics. However much you simplify an explanation, if there are feedback loops and other interactions involved, you cannot actually foresee the real outcome of the simplified system at certain levels of detail. So even if you have given a most basic set of principles, you are going to fail to predict complex behavior.

In that sense, reducing the problem does not solve it. You still need other methods and measures that capture the behavior at other levels, and they may be of far more use than an underlying explanatory mechanism. So science cannot proceed entirely through reduction, it needs to be multiple-layered and the more complex multiple realizations need to be maintained as a part of the science even if they are explained by more basic behavior.

There is no contradiction here even within physics. We make use of both statistical mechanics and the flow theory of heat. We make use of both models of proton bonding and the pH notion of acidity, we admit that turbulence requires an entire extra layer of principles in addition to the basic theory of fluids to allow any kind of useful prediction, etc. etc. etc. The higher-level observations both verify the underlying theory and make it more usable. And they demonstrate interesting subtleties we would miss if we focused entirely on root causes. Downward is not always the best direction to look.

And there is often a good reason to look askance at a reductivist agenda. In fields like biology and geology, more reductivist approaches routinely misled exploration until the relevant reducing elements were introduced from outside (DNA came from chemistry), or came together naturally (Subduction as the source of plate tectonics came out of 30 years of attempts to put together data from three different warring subdisciplines).

Worse yet, there are sciences where reductivism has, to date, done only harm: dwelling too closely on the methods of physics has always done harm to more complex sciences like psychology. It draws them to solidify underlying principles before they have an adequate survey of the actual field. It makes them pursue narrow, closed experiments that fail to capture meaningful data. We end up with Freud before Rogers, and Reich before Linehan, and we spend all of our time discussing nonsense.

  • I agree that reductionism can be really harmful at times. But is it possible to reconcile the idea that "multiple planes of explanation distinctly exist and are fundamentally irreconcilable" with the spirit of science? You point out that such planes are necessary and powerful in terms of predictive power, but that is only because, you know, we don't have better computation devices, or something, ryt?
    – BlowMaMind
    Commented Apr 8, 2018 at 2:12
  • Wouldn't the invention of better and better computation devices, make it more and more powerful to try from bottom up also?
    – BlowMaMind
    Commented Apr 8, 2018 at 2:21
  • @BlowMaMind. No, better computational devices cannot remove chaotic dynamics. There will always be derived structure that cannot be determined from reduced structure. Consider a standard example of chaos: The structure of the Mandelbrot set can be 'explained' in terms of a simple iteration, but it is not really an explanation, or even an adequate description. So to claim to have reduced that structure to its cause helps no one.
    – user9166
    Commented Apr 8, 2018 at 18:11
  • @BlowMaMind . Wherever your quote comes from it is irrelevant. Reductionism and anti-reductionism can be reconciled. They almost always happen together. I can have an underlying cause but still not have a useful explanation, and still need other principles to make a science mean something.
    – user9166
    Commented Apr 8, 2018 at 18:14

The mistake is to see explantory layers as incompatible or competing.

In evolution theory, it has been understood for a long time that the gene is the fundamental level of selection, and kin-selection theories and of group selection coming to dominate over individual selection have been dismissed. But what is now increasingly accepted, is that evolution occurs on the level of the gene and other levels https://en.m.wikipedia.org/wiki/Group_selection#Multilevel_selection_theory (increasing numbers of studies are demonstrating this happens in practice, whatever Dawkins says).

Emergence in physics https://en.m.wikipedia.org/wiki/Emergence can be seen as analogous. A level of ordering and complexity occurs in addition to the 'fundamental' one, that is not independantly or in contradiction to it, but requiring an additional explanatory layer for clear understanding. It is like computer code that runs independentally to the particular computer, because it involves structure/s which self-reinforce or interact as assemblies.

Entropy doesn't exist at the scale of individual particles. Yet for assemblies of particles it gives arguably (by Einstein) the strongest laws we have. Hawking was able to use the process of analogy to 'reverse engineer' the temperature of black holes (which showed they produce Hawking Radiation) and so, that they have entropy solving the apparent violation of the https://en.m.wikipedia.org/wiki/Black_hole_information_paradox This is even though we think a black hole is a 'particle', lacking internal structure.

Antireductionist, or more conventionally holistic or systems thinking is proving increasingly important in many sciences. Substance dualism (like Descartes') is contrasted with materialism, and has been apparently neccessary to explain qualia. But property dualism, specifically in the form of emergent materialism, reconciles qualia as an explanatory layer. Yes in a very complex way it is still just particles. But just as it is simpler, more elegant, more computable, to deal with modelling a 'program' instead of calculating every transistor (and that ignores transistors can be replaced with all kinds of other switches inc quantum superpositions), so we have 'supervening' organisational states, like cells, minds, and subjective experiences, far more effectively dealt with in their own explanatory layers, even though fundamentally 'reducible'. Like entropy can be pictured as a 'cause', even though it is only statistically valid, so minds and their states be causes, even though they are also 'just' fluctuations of atoms.

  • 1
    Multi-levels including group selection seems like a good example of antireductionism. Commented Apr 8, 2018 at 0:49
  • So are you agreeing that Ontological Anti-reductionism is not a scientific position, or not?
    – BlowMaMind
    Commented Apr 8, 2018 at 2:04
  • 1
    @elliotsvensson So, if we left a particle alone, you would expect it to eventually change, totally on its own, in some irreversible fashion without other particles having to be invented to complicate its state? You would eventually get virtual particles, but those would not increase the entropy of the single particle, only of the system. Whether or not entropy can be defined in some arbitrary way for a single particle, thermodynamics exists only in composites. The interaction of things relies on there being things, plural.
    – user9166
    Commented Apr 9, 2018 at 12:22
  • 1
    @elliot svensson "mind would be an emergent thing, not an emergent property" Why? I don't see a clear answer in you comment or link. I just disagree. Open chat, or dedicate a question by all means. I see consciousness as like 'pile' in the sorites paradox - or statue in your comment. It is conventional use, but not causal, not explanatory, and most of all fuzzy. A causal explanatory framework for consciousness must explain it in principle at least in atoms or fields, or else justify new new constituents, 'magic juice'. Information and it's transmission are fundamental to atoms, but, in them..
    – CriglCragl
    Commented Apr 9, 2018 at 21:39
  • 1
    @CriglCragl, I am only asserting a common meaning to "free will", which I feel is well-enough received to simply state. I think that what you are proposing here is logical enough, but needs to compete against the position I stated. Commented Apr 10, 2018 at 22:53

A couple points:

  1. It's not clear what a scientific position is. So let's replace 'position' with 'theory' so that the discussion is clearer.
  2. Requiring that a scientific theory not contradict any other scientific theory is definitely not a criterion of science. If it were, we would be in really bad shape. New theories always contradict those they replace. This is how science progresses.
  3. Furthermore, I would suggest the following principle: If X is a scientific theory or hypothesis, then it's antithesis (~X) is also a scientific theory or hypothesis. Note that I'm not saying that they are both equally good. I'm just saying it doesn't make sense to say that claim $X$ is scientific, but ~X is not scientific. This principle is especially important in consideration of the percentage of scientific theories that have turned out to be wrong.
  • The theory that gravity acts in proportion to mass and to push things apart is 'scientific'?
    – CriglCragl
    Commented Apr 7, 2018 at 23:35
  • @CriglCragl yes, of course. Why wouldn’t it be?
    – njs
    Commented Apr 7, 2018 at 23:47
  • It is a flying spaghetti monster, an invisible pink unicorn, or a Russel's teapot theory. You mistake logically coherent for scientific.
    – CriglCragl
    Commented Apr 7, 2018 at 23:59
  • @CriglCragl it's an intelligible claim about the universe that's easily testable. Therefore, it is a scientific hypothesis. Note that it's not necessarily a good scientific hypothesis, but it is a scientific hypothesis nonetheless. Again, by what criterion do you claim it is not scientific? Also, note that the examples you provide above are meant to illustrate non-testable (and scientifically questionable) hypotheses. You're U = +Gm1m2/r hypothesis, however, is quite easily tested. Thus your examples of spaghetti monster and invisible pink unicorns don't apply to this argument.
    – njs
    Commented Apr 8, 2018 at 0:45

I think that some people would prefer if antireductionism was a scientific position. Then they could account for irreducible things without denying materialism.

Others (including Stephen Jay Gould and his "non-overlapping magesteria" and Karl Popper) are perfectly happy to think that science is reductionistic, because they acknowledge the irreducible things to be true but not known to science.

People who have problems with science and reduction are those who idealize science as reductionistic and also assert that science is the only reliable source of truth claims.

BTW: I changed the claim in Wikipedia to be less broad... now it reads "Antireductionism is the position in science and metaphysics that stands in contrast to reductionism (anti-holism) by advocating that..."

Clearly, antireductionism is a position in science one can take on interpreting a dataset or experimental result. But now Wikipedia is not asserting that science is antireductionistic.

  • This is not what I meant in my question. But you are right, maybe less confusion this way.
    – BlowMaMind
    Commented Apr 9, 2018 at 15:23
  • I think that some people would prefer if antireductionism was a scientific position. Then they could account for irreducible things without denying materialism. Others (including Stephen Jay Gould and his "non-overlapping magesteria" and Karl Popper) are fine thinking science is reductionistic, because they acknowledge the irreducible things to be true but not known to science. People who have problems with it are those who idealize science as reductionistic and also assert that science is the only reliable source of truth claims. Commented Apr 9, 2018 at 16:31

To see where antireductionism has value consider the natural numbers (that is, positive integers) where one does not need antireductionism because of the presence of unique factorization. Given a natural number there exists, although it may be impossible to find within the finite age of the universe, a unique analysis of that number into its prime factors. That analysis reduces the natural number to those factors. The synthesis occurs through multiplication which unambiguously brings us right back to the original number.

If reality were so easy to analyze and synthesize as factoring a natural number, that is, if reductionism were true, there would be little doubt that the reality that we are part of would be just as deterministic as mathematics.

The antireductionist position has epistemological and ontological components. The epistemological component is like the problem of not being able, within the lifetime of the universe, of actually factoring most natural numbers. The ontological component is totally unlike the situation with the natural numbers. That claims that we cannot reduce everything to something else, say a quantum reality. It is like saying that there exist some natural numbers that we cannot, even in principle, factor.

Is there anything in reality that might suggest ontological antireductionism is true? The article suggests one can find examples supporting antireductionism in areas such as cybernetics, systems theory, history and economics.

Now to answer paraphrases of the questions:

Can antireductionism be a scientific position?

To say antireductionism is not a scientific position would be to say there exists a demarcation between what is science and what isn’t. Perhaps antireductionism is a kind of pseudoscience? One of the proponents of finding a demarcation to separate science from pseudoscience is Karl Popper. According to the antireductionism wikipedia article, Popper “was a famous proponent of antireductionism” in his book Of Clouds and Clocks: An Approach to the Problem of Rationality and the Freedom of Man. So, not all ways to demarcate the boundary between science and pseudoscience reject antireductionism.

If everything is made of quantum particles can’t we use this one framework to explain everything?

A reductionist might answer yes. An antireductionist would more likely answer no. If the reductionist is correct, reality may be like the natural numbers. We are part of reality, so there would be a reductionist explanation for everything we do. That would suggest superdeterminism were true and everything is determined. That would include, according to John Bell, “the 'decision' by the experimenter to carry out one set of measurements rather than another”. So if that reductionist view is correct the reductionist needs to find a way to avoid superdeterminism or give up science itself.

There is a comment asking "What is eliminativism?"

A reductionist position attempts to analyze something into its components and hopes that such analysis can replace what one started with. For example, 6 can be analyzed into the factors 2 and 3. An eliminativist position would say that there is no need to bother reducing something. It is illusory. Succeeding at such an argument is to say that something does not exist. It would be like saying we do not have to bother factoring the imaginary number i in the natural numbers because i does not exist in the natural numbers. Believing it does is an illusion.

  • If there is no demarcation whatsoever between science and pseudoscience, one could claim anything is science, ryt? Is that what you are saying? I thought there was fuzziness in what exactly is science, but everybody has an idea what it more-or-less is ryt?
    – BlowMaMind
    Commented Apr 6, 2018 at 14:09
  • How could one reconcile the following two positions: $\mathbf{1}$. Physics is a science. $\mathbf{2}$. Antireductionism is a scientific position.
    – BlowMaMind
    Commented Apr 6, 2018 at 14:10
  • @BlowMaMind Although I don't have a quote, I think Feyerabend would say there is no demarcation and we should question our motives for why we look for one. Others would disagree. Popper thought there was. He thought psychoanalysis was not science because it didn't allow falsification. Since Popper believed in the possibility of demarcating between science and non-science and he supported antireductionism, it will not be easy to reject antireductionism as a scientific method. Commented Apr 6, 2018 at 14:17
  • 1
    Is antireductionism "not all properties of a system can be explained in terms of its constituent parts and their interactions" Or "not all properties of a system are easy to be explained in terms of its constituent parts and their interactions" or "not all properties of a system need be explained in terms of its constituent parts and their interactions"?
    – BlowMaMind
    Commented Apr 6, 2018 at 14:24
  • 1
    @CriglCragl The article is the wikipedia article on "antireductionism". en.wikipedia.org/wiki/Antireductionism Commented Apr 8, 2018 at 0:44

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .