Responding to an article on reductionism

 Big Think wrote an article on how the universe really is 100% reductionist in nature, this is my response to it.  

The article as a whole has 6 sections, namely: 

• The Introduction  
• Section 1: "The Fundamental" 
• Section 2: "How A Reductionist Sees The Universe" 
• Section 3: "How 'Apparent Emergence' Is Readily Explained By Reductionism" 
• Section 4: "The 'God-of-the-Gaps' Nature of Non-Reductionism" 
• "Final Thoughts" 

I will address each of them in turn. 

The introduction

The article is arguing for reductionism, which it defines in this manner: 

"the fundamental laws that govern the smallest constituents of matter and energy, when applied to the Universe over long enough cosmic timescales, can explain everything that will ever emerge. [...] This simple idea — that all phenomena in the Universe are fundamentally physical phenomena — is known as reductionism." (ellipses mine) 

The article provides a general statement of what it is setting out to argue when it says: 

"As far as we can tell, the Universe is truly 100% reductionist in nature. Our ignorance about why certain emergent phenomena exist and how they behave is no excuse for magical thinking."

and the introduction likewise in turn concludes with a similar summary of what it will argue, namely that:

"While it’s true that many phenomena are not obviously emergent from the behavior of their constituent parts, reductionism should be the default position, with anything else being the equivalent of the God-of-the-gaps argument" 

Thus in turn, I can summarize what my proceeding critique will be in this way; (i) we cannot tell that the universe is 100% reductionist in nature, (ii) our ignorance about why certain phenomena emerge cannot all inherently be explained by reductionism, and so (iii) it is a question begging straw man of the non-reductionist position to say that we engage in magical thinking and God-of-the-gaps style reasoning.  

Analysis and Critique of the sections

Section 1: "The Fundamental" 

This section explains what the author understands 'The Fundamental' to mean, it does not seem to contribute anything to the overall argument they are making, except to introduce those who are unfamiliar to the concept of fundamentality to it. Philosophers have also spoken about the topic, so I'll link to an article on the topic, if one wishes to look more deeply into it than what the article gives. 

Section 2: "How a Reductionist views the Universe" 

The issue with this part of the article is that it makes this claim: 

"You either understand your system well-enough to understand what should and shouldn’t emerge from it, in which case you can put reductionism to the test, or you don’t, in which case, you have to go back down to the null hypothesis: that there’s no evidence for anything novel"

The second clause is clearly question begging. The idea that reductionism is the null hypothesis is identical in meaning to the claim that reductionism is the default position, but this is the very claim that the article was meant to argue for; but at this point in the article there has been no argument made for the claim, so that this is evidently question begging. 

On the contrary, it's not just question begging, but evidently false.  

In any argument between a variety of positions, the default position is 'no' position i.e. to be 'neutral' between the views, to not take a side one way or the other, to suspend judgement. The 'null hypothesis' then, is the hypothesis takes the form of a probability evaluation of all views under consideration, namely the view that no side has better arguments in favor of it than the other, that they are equiprobable. 

In this case, there are exactly two views under consideration, the reductionist and non-reductionist view, so the null hypothesis is that each have a probability of exactly 50% of being true, so that nothing favors one or the other either way. (so that we should prefer to suspend judgement either way, than to affirm the assertion or negation of reductionism.) From here, a neutral investigator should investigate the arguments one way or the other, to see if they think the arguments for and against each side ultimately weight more in favor of the one view or the other, or if they ultimately balance out. 

This is the 'null' hypothesis because (to use wikipedia's definition) it is the hypothesis which says that "no difference or relationship exists between two sets of data or variables being analyzed" in this case, the sets of data are the arguments for and against reductionism and non-reductionism respectively; we start off on the assumption that there is no difference or relation between these two, and so, nothing to effect the probability of either position to be greater than the other, and so, no reason to prefer one position over the other, and so rather, every reason to remain neutral. Neutrality is 'always' the null in cases of dispute, to believe otherwise is already to introduce bias into the situation, and so already to begin begging the question. 

Thus reductionism clearly is not the null hypothesis in the dispute between reductionism and non-reductionism; the idea that any one side of a dispute can be the null 'in that dispute' is itself incoherent.  

Prior to this question begging claim, the article argues this: 

"It’s very important, when we discuss the idea of reductionism, that we don’t “strawman” the reductionist’s position. The reductionist doesn’t say — nor does the reductionist need to assert — that they have an explanation for each and every complex phenomenon that arises in every imaginable complex structure. Some composite structures and some properties of complex structures will be easily explicable from the underlying rules, sure, but the more complex your system becomes, the more difficult you can expect it will be to explain all of the various phenomena and properties that emerge."

The issue here is that this is also false. As noted in the introduction, the article itself defines reductionism in this way: 

"the fundamental laws that govern the smallest constituents of matter and energy, when applied to the Universe over long enough cosmic timescales, can explain everything that will ever emerge. [...] This simple idea — that all phenomena in the Universe are fundamentally physical phenomena — is known as reductionism."

But of course,  if by the articles own admission, reductionism says that the 'fundamental laws' can 'explain everything that will ever emerge' then it is no straw man to say that the reductionist should 'have an explanation for each and every complex phenomenon that arises in every imaginable complex structure' for that's literally just a restatement of the very view in question. 

If the reductionist wishes to claim to know that reductionism is true, to be justified in supposing it's true, and in accusing non-reductionists of 'magical thinking' then it's perfectly rational for both non-reductionists and those in affirming neither the assertion nor negation of reductionism should ask for the reductionist to be able to back up their claim in 'all cases' since it is precisely all cases that the reductionist is saying reduction applies to; for even the non-reductionist can grant that reduction applies in 'some' cases, but it would be a hasty generalization fallacy to presume that it therefore applied to 'all' cases. 

As such, this complaint reeks of a reductionist attempt to shirk the burden of proof, and added with their presumption of being the null, it is also a clear attempt of them to 'shift' that burden onto the non-reductionist. Thing is 'both' have a burden, and the reductionist pretending they have no burden is just more question begging. 

Section 3: "How 'Apparent Emergence' Is Readily Explained By Reductionism"

The article begins this section by listing some cases where emergence is explained, and then outright says: 

"And then there are properties that we cannot fully explain, but that we also are incapable of making robust predictions for as far as what we expect to see under those conditions. These “hard problems” often include systems that are far too complex to model with current technology, such as human consciousness."

Now aside from this statement being question begging (it assumes it's merely a matter of 'complexity' that prevents the reduction form occurring, rather than the thing in question being well and truly irreducible, so again, the article is providing no argument for the reductionist view, just assuming) it's also an admission that the very premise of the section isn't attained. The sections title promises some general method by which anything might be reduced, and not only never provides, but outright admits that it won't be doing so.   

Instead, it just goes on begging the question against the non-reductionist view, saying: 

"In other words, what appears to be emergent to us today, with our present limitations of what its within our power to compute, may someday in the future be describable in purely reductionist terms."

They do try to provide something of an argument here, namely: 

"Many such systems that were once incapable of being described via reductionism have, with superior models (as far as what we choose to pay attention to) and the advent of improved computing power, now been successfully described in precisely a reductionist fashion. Many seemingly chaotic systems can, in fact, be predicted to whatever accuracy we arbitrarily choose, so long as enough computational resources are available."

This issue with this induction is that, on reductionism, it's essentially self-refuting; for if we reduce all things to the physical, then in what sense is there a past to root our induction of the value of reductionism in? The past 'of what?' of human persons and their scientific enterprise?

If reductionism is true, these things don't exist (not even as hallucinations) since they are nothing over and above their constituent parts (for, as the article itself says: "The whole isn't greater than the sum of its parts; that's a flaw in our thinking"); human beings and the scientific enterprise (and so the history of human beings and the scientific enterprise) reduce too their most fundamental constituent parts, as the article says in the section on the fundamental they reduce to "the most indivisible, elementary entities of all and the laws that govern them". 

Thus, on reductionism, there simply is no past in which reductionism was ever useful, there was only so many elementary entities operating in accord with the laws that govern them, there is 'nothing more' to reality than these things, not even the illusory appearance of them (for then the appearance would be something distinct from the constituents, and so wouldn't reduce to them, but would themselves have to be among the fundamental entities, which is contrary to the whole physicalist reductionist enterprise); the claim that there is such an appearance is not true or even meaningful (for then the meaning would be distinct from the constituents, and so wouldn't reduce to them, but would again, have to itself be among the fundamental entities, which is again, contrary to the whole enterprise of physical reductionism) instead, both the claimed meaning and appearances are not distinct realities unto themselves, but nothing more than (again, to quote the article itself) "the most indivisible, elementary entities of all and the laws that govern them". 

But you see how this undercuts the induction; the induction wishes to appeal to meaningful descriptions of past appearances in the realm of humanity and human scientific endevour, to show that the reductionist enterprise should be assumed to apply to all cases; but then that would include those self-same meaningful descriptions and past appearances of humanity and human scientific endevour; so that on the assumption that reductive physicalism is true, we must also assume that there is no such past to root the induction in, in the first place and so must admit that it's a terrible induction, since it's inductive from precisely no data whatsoever. 

Heck, on physicalist reductionism, the 'induction itself' cannot be anything over and above the fundemental entities operating in accord with their laws, so that the induction itself cannot have a unique and distinct truth, meaning, strength, or what have you; all of those traits by which we would consider induction as good 'cannot exist' on physical reductionism, but must rather reduce again, to 'the most indivisible, elementary entities of all and the laws that govern them'  The reductionist wants to have their cake and eat it too, it doesn't work. 

The section concludes by saying this: 

"The existence of systems that are too complex to predict with current technology is not an argument against reductionism." 

We find that not only has this conclusion not been established, but what's must worse, something of the inverse has shown itself; namely, the existence of systems that are sufficiently 'simple' to predict with current technology cannot possibly be an argument for physicalist reductionism; indeed, 'nothing even in principle' can be an argument 'for' this view; for the more successful such an argument is, the more it proves that such arguments do not and cannot exist, and so are not and can never be sound. In the end, physicalist reductionism undercuts any argument for itself.

 Indeed,  more strongly, reductionism is self-defeating, for if it is true, then abstract realities like 'truth' and the very meaning of the term 'reductionism' and all the propositions defining the view, all must not exist. Of course, if truth doesn't exist, then reductionism isn't true, and if the meaning of the term or propositions describing the view don't exist, then there is no 'reductionism' to be or not to be true, in the first place. Thus if physicalist reductionism as a view so much as exists, we therefore know it to be false; simply in virtue of the implications of it's own self-describing claims. 

In other words; if everything is nothing more than 'the most indivisible, elementary entities of all and the laws that govern them' Then so too must reductionism itself be, and as elementary entities and the laws governing them aren't the kinds of things that can be true or false, nor are they themselves inherently the meaning of reduction itself (for it is not just reductionism that reduces to them, but all other things as well), then since that's the case if reductionism is true, then it's not true, indeed, if reductionism is true, then reductionism doesn't even exist, so as to 'be or not be true' in the first place. 

Thus reductionism is refuted thus: 

• Premise 1: If reductionism exists, reductionism is not true. 
• Premise 2: reductionism exists
• Conclusion: reductionism is not true.  

Section 4: "The 'God-of-the-Gaps' Nature of Non-Reductionism"

This is where the article begins to formulate criticisms of the alternative non-reductionist view, specfically by attempting to argue it is a kind of God-of-the-Gaps argument. The article says that the non-reductionist view: 

"basically says, “Well, we know how things behave on a certain scale or at a certain time, and we know how they behaved on a smaller scale or at an earlier time, but we can’t fill in all the steps to get from that small scale/early time to understand how the large scale/later time behavior comes about, and therefore, I’m going to insert the possibility that something magical, divine, or otherwise non-physical comes into play"

Now before addressing their arguments against this particular view, it has to be noted that this is already problematic. Namely, this is a straw man (which, to be clear, I presume to be unintentional, we should always give the benefit of the doubt), and that it's a straw man is evident simply in virtue of the fact that it's not actually engaging with an actual non-redutionist interlocutor; the article does not quote non-reductionist philosophers, but sets up a generalization of the non-reductionist view for itself and attacks that view, as though it exhausts all possible non-reductionist positions; but while there may be some who hold to this view, that does not mean all forms of non-reductionist are committed to it; or at least, if the author wishes to claim that we are, then they are bound to provide an argument for it, and they nowhere do this. 

Thus consider my above arguments; my issue with reductionism is not merely an empirical issue, not merely that there is this or that thing that reductionism doesn't presently account for, my issue rather was always also a conceptual one i.e. there are somethings that reductionism cannot in principle account for it; my issue is with the internal logic of physicalist reductionism itself, that reductionism, taken on it's own terms, is not just self-defeating, (i.e. if it's true, we can't know it) it's outright self-refuting. (i.e. if it's true, then it's logically false i.e. incoherent, neither true nor not true, but non-existent, and sine it clearly exists, then it's surely false) so that whenever a reductionist talks about having the potential to account for certain things (truth, meaning, qualia, persons, the history of science, the existence of 'reductionism itself', etc.) I find such claims not just unsbstantiated, but incoherent. 

If the reductionist wishes to escape this critique, they need to show a 'conceptual problem' in the critique, they need to show that I am somehow straw manning them (for I am using their own words to form my critique, they have to show how this use was a misuse, though to note, if I am, this is accidental, not intentional) or perhaps they have to admit that they personally have been unclear, and so have to clarify themselves to show how their view escapes the conceptual problem I'm putting forth i.e. admit that my critique does succeed at refuting a certain fair reading of what they said, but they meant something subtly different that the critique does not succeed at addressing. 

Until such a time as this has been done, we non-reductionists are perfectly rational in holding physicalist reductionism (at least as presented in this particular Big Think article) to be an inherently incoherent position. 

Consider, in any case, the critiques the article brings to bear of the straw man they've erected:  

"Although this is an assertion that is difficult to disprove, it’s one that has not only zero, but negative scientific value. The whole process of science involves investigating the Universe with the tools we have at our disposal for investigating reality, and determining the best physical model, description, and set of conditions that describes that reality. What a fool’s errand it is to assert, “Maybe we need more than our current best model to describe reality” when:we don’t even have the computational or modeling power necessary to put our current model to the test, and where these are the regimes most likely — if you insert something magical, divine, or non-physical — where science is very likely, in the very near future, to show that such an intervention is wholly unnecessary."

The issue I would take with this argument is that it seems to me just wrongheaded. If we had two theories that could equally explain the same data, but one theory required more computational resources to explain the phenomena than the other theory, it seems to me that, all else equal, we should favor the less resource intensive theory. 

In this case it's worse than that, because we 'don't actually know' if our problem is 'simply' a matter of having not enough computational resources; for all we know, even 'infinite' computational resources couldn't compute these things. The author of this article assumes this, but does not 'prove' that it is simply a matter of limited resources, if this wasn't evident from the preceding, I'll say this explicitly: this begs the question. A neutral person not favoring either of our views would not be obliged to accept this as a premise, any more than they would have to accept the non-reductionist view that they can't be computed as a premise; if either of us wish to make our respective claim about whether this is or is not a matter of computational resources, we both have a burden to prove our claim, we can't just assume it without argument. 

More to this, in the case of non-physicalism, we're not suggesting that we get rid of the present theory (as though it is in competition with modern physics) but merely that there is more to reality than what physics studies; there is data that physics is not and can not pay attention to; namely, non-empirical data i.e. purely conceptual data; and I have worked to provide an indirect argument for this via my critique. 

(for of course, you cannot argue for data, data is just data, it is given; and of someone doesn't accept it, it's not a matter of not accepting an argument, but more a matter of being blind to the data; the only way to argue for the data then, is to argue against the arguments against it; so as to, hopefully, direct the attention of the non-inferential faculties of one's interlocutor to the data as it presents itself to us.)

From the above it should be evident that this critique really fails; when a view can only make sense of a thing by appealing to more and more processing power (indeed, 'can't yet' make sense of it, 'until' it gets yet more power, and so is only 'holding out' on an explanation rather than actually giving one), then that in fact is a good sign that we might want to find another view that is not so reductive (and so, has more fundamental entities), and so which doesn't require so many computational resources. 

To be sure, this is no final nail in the coffin, reality is allowed to be difficult to compute, it has no obligation to be easy for us to understand; and likewise we should avoid multiplying entities beyond what are necessary to explain the data; but the issue here is that we should also be making sure that we're actually explaining the data in the first place, rather than just explaining it away. 

Thus, just as reality is allowed to be difficult to compute, so it's allowed to be typologically complex i.e. allowed to have very many fundamental (i.e. irreducible) entities than we presently suppose; and just as we should not multiply entities beyond what is necessary to explain the data; so likewise we should be multiplying entities as much as is necessary to explain it as well, and so should not 'eliminate' entities when they are necessary. 

Rather, we should go by a Goldilocks principle of explanation:

• We should have neither too few nor too many entities to explain the data, but having just as many as are necessary, and no more. 

Physicalist reductionism is problematic then because it pushes us to eliminate proposed fundamental entities which are necessary to make sense of the data we have. Indeed, such reductive physicalism ignores that data, or at least, ignores the logical implications of it's own view were it to be applied to said data.

The author goes on to say a few other things, such as this:  

"If you either believe or simply want to believe that there’s more to the Universe than the sum of its physical parts, that’s a statement that science is completely agnostic about. However, if you want to believe that a description of the physical phenomena that exist in this Universe require either: something more than the physical laws that govern the Universe, and/or something other than the physical objects that exist in in the Universe, perhaps the least successful decision you can make is to put those “metaphysical” entities in a place where science, once it advances just a little bit further, can disprove the need for them entirely.

I have never understood why one would be so willing to assert the existence of the divine or supernatural in a place where it would be so easy to falsify the need for it. Why would you believe, in a Universe that’s so vast, that something beyond the capability of our physical laws to describe would primarily appear in such an extraneous, unnecessary place? If the Universe, as we observe and measure it, isn’t able to be described by what’s physically present within it under the known laws of reality, shouldn’t we determine that to actually be the case before resorting for non-scientific, supernatural explanations?"

The author admits of not understanding the purpose here; but I find it rather odd that the author is taking issue with us rooting our religious belief in something falsifiable; for isn't it rather a common critique of the anti-religion crowd that religion is wrong precisely because it's unfalsifiable? Isn't falsifiability a virtue of a theory, rather than a flaw? How then could this be grounds of objection?  

The reason falsifiability is rational is precisely because it aids in the search for truth; for if one's view is falsifiable and false, then it will eventually show itself false, whereas if one's view is true, then everything becomes continual confirmation of that truth; whereas in an unfalsifiable view, if one's view is false, it will never show itself false, and so one will never escape it. 

To be fair, some unfalsifiaiblity is unavoidable; say, the unfalsifiability of the laws of logic, or of any truth which simply follows from a conceptual analysis of 'falsifiability' itself (though even there, there is room for falsifiability insofar as you are speaking of this or that person's usage of a term.) but the idea then is to try and 'reduce' unfalsifiability as much as possible i.e. all else equal, more falsifiable views are to be preferred.  

Still, the problem here is that the author is complaining that metaphysical and religious views have a explanatory 'virtue', but then of course, that is no complaint, but a compliment. That they don't see this seems only to show how dogmatically committed they are to reductionism.

Indeed, much the issue of reductionism is that, from a purely empirical lens, it most certainly is not falsifiable; no matter how many examples of things that have yet to be reduced to more fundamental realities I point out, the reductionist can always say that we will figure out a way to reduce it eventually, find out how to compute it eventually, they can always hold out on that hope; but of course, that something cannot be falsified on the empirical level does not mean it can't be falsified on the conceptual level, and I have already done that precisely for this view above; or at least, for the specific interpretation of the article that I am putting forth. 

Section 5: "Final Thoughts" 

The article goes on more or less to summarize what it said before, which as the preceding would argue, is just so much question begging straw men; so there's nothing strictly new here to address, so I won't. Instead, I will give my own final thoughts, in a similarly summary fashion: 

Provided I have properly understood the article, I have shown that the the view of the reductionistic view of the author of this article is, on the whole, quite wanting as a view. Perhaps this is a problem simply with their presentation, perhaps not; I won't comment on reductionism as such, for there may be more capable defenders of it; but as this view stands, I have shown this reductionism to be a question begging, straw-manning, empirically unfalsifiable, and conceptually self-defeating view; and for that reason should be rejected in favor of a non-reductionist view.