Philosophy Discussion Forums

Can you clarify what you mean by "realism"?

I see two different basic definitions for scientific realism, and I consider both definitions meaningful even though they are unrelated to each other, in my opinion:.

1. The contents of empirically validated theories correspond to objective objects in Reality.

2. The contents of empirically validated theories can illuminate the ontological nature of Reality.

Since QM fails to determine values corresponding to objective (empirical) measurement, Realism demands we consider it an incomplete theory (at best) and to seek a more fundamental theory capable of yielding correspondence to empirically objective measurements. Many Worlds and Bohmian Mechanics are examples of theorizing driven by a need to uphold scientific realism.

(As an aside, I personally feel there is a practical distinction between scientific realism and a more general philosophical realism. Philosophical realism requires the ontological status of Reality to be independent of the Mind. Scientific realism doesn't seem to operate from any such explicit commitment, even when scientists (or philosophers of science) extrapolate ontological statements from empirical results).

Are you equating realism with determinism?

I think determinism implies realism, but realism need not require determinism. For example, Feynman's sum-over-histories approach is driven by a desire to uphold realism but settles for non-deterministic outcomes. Since the 1990s there have been experiments capable of displaying wave/particle duality simultaneously, as well as maintaining quanta in controlled superposed states. Such empirical outcomes suggest to me that anyone who accepts Realism should learn to dispense with determinism.

A Poster He or I wrote: Such empirical outcomes suggest to me that anyone who accepts Realism should learn to dispense with determinism.

I don't think we can know if nature is ultimately deterministic/indeterministic because we can't a priori exclude the possibility that QM is not the ultimate/final truth as argued here:

This fact is often used to claim that QM implies that nature is fundamentally random. Of course, if the usual form of QM is really the ultimate truth, then it is true that nature is fundamentally random. But who says that the usual form of QM really is the ultimate truth? (A serious scientist will never claim that for any current theory.) A priori, one cannot exclude the existence of some hidden variables (not described by the usual form of QM) that provide a deterministic cause for all seemingly random quantum phenomena. Indeed, from the experience with classical pseudorandom phenomena, the existence of such deterministic hidden variables seems a very natural hypothesis. Nevertheless, QM is not that cheap; in QM there exist rigorous no-hidden-variable theorems. These theorems are often used to claim that hidden variables cannot exist and, consequently , that nature is fundamentally random. However, each theorem has assumptions. The main assumption is that hidden variables must reproduce the statistical predictions of QM. Since these statistical predictions are verified experimentally, one is not allowed to relax this assumption. However, this assumption alone is not sufficient to provide a theorem. In the actual constructions of these theorems, there are also some additional“auxiliary” assumptions, which, however, turn out to be physically crucial! Thus, what these theorems actually prove, is that hidden variables, if exist, cannot have these additional assumed properties. Since there is no independent proof that these additional assumed properties are necessary ingredients of nature, the assumptions of these theorems may not be valid. (I shall discuss one version of these theorems in more detail in Sec. 5.) Therefore, the claim that QM implies fundamental randomness is a myth.

Quantum mechanics: Myths and facts
http://xxx.lanl.gov/pdf/quant-ph/0609163.pdf

I don't think we can know if nature is ultimately deterministic/indeterministic because we can't a priori exclude the possibility that QM is not the ultimate/final truth...

It is precisely because we can't know if nature is ultimately deterministic/indetermistic that we must not assume a priori that it is deterministic. It is because our best empirical methodologies currently do not indicate a determistic universe that we should pursue science with the goal of cataloguing relations between phenomena and extrapolating new relations therefrom; not the goal of expanding our hubris to undergird new discoveries in order to force them to "make sense" according to our epistemological biases.

A Poster He or I wrote:It is precisely because we can't know if nature is ultimately deterministic/indetermistic that we must not assume a priori that it is deterministic.

It's not so much the issue of determinism/indeterminism that is the major issue for many physicists/scientists. It's the realism versus non-realism (instrumentalism) issue. Consider the wave function. Is it an entity (e.g. some type of new field) that exists irrespective of whether we are here or not or is it just information? If information, what is it information about?

I agree with you in one regard: the various Reactionary Interpretations of QM that I'm most familiar all wish to uphold scientific realism. However, the way they go about suggests that it is determinism they are actually bent on achieving. To this end they are willing to introduce elements into reality that have no correspondence to any phenomena that humans can objectify (typically corresponding to some sort of "hidden variables").

At first glance, it hardly seems befitting of a Realist to invent physical structures or processes for reality that have no empirical substantiation other than consistency with abstract mathematical formalisms. But in fact it isn't so simplistic once you notice that the purpose of these hidden variables is to force the wave functions' values to represent deterministic outcomes at all times in all instances (rather than remaining superpositional outcomes) either by 1) physically removing superpositional outcomes from being accessible, or 2) mathematically cancelling out all superposition.

Example 1: Many Worlds dispenses with any collapse to unity, declaring by fiat that the values of the wave function are real values at all times. Superposition is eliminated by erecting physical barriers (whole discreet universes) for each value to inhabit. Nevermind that the mechanisms for this endless creation of discreet physical realities is conveniently outside the purview of MW.

Example 2: De Broglie/Bohm's Pilot Waves/Holomovement essentially introduces a subquantum layer of reality in the form of an infinite non-degrading field effect that proactively "directs" quantum-scale outcomes. In short, the probabilities of the wave function are merely mathematical artifacts: there is always only one possibility to be determined and its determination is independent of observation. Nevermind that the agency doing the determination lies outside of normal space-time, conveniently outside the purview of empirical validation.

Example 3: Cramer's Transactional Universe introduces an inverse wave function for every wave function, propagating from the future into the past, cancelling out the wave function at the exact Present (the moment of observation), eliminating all superposition. Nevermind that the mechanisms for this "handshake" between future and past require ubiquitous (and unverifiable) backward time travel.

So it seems that Realists are willing to invent hidden structures and processes for reality so long as those inventions allow the wave function to represent potentially objectifiable (i.e., deterministic) results at all times, in all instances, even though such objectification exists merely in principle (that is, there is no empirical methodology for achieving such objectification).

Perhaps you can begin to appreciate why I consider Realism to be anti-scientific.

A Poster He or I wrote:Perhaps you can begin to appreciate why I consider Realism to be anti-scientific.

Not really. I view both realism and anti-realism/instrumentalism as philosophy. Anti-realism or the view that physics is the science of meter readings (instrumentalism) is also just philosophy. And there are realists who are not determinists or at least are agnostic on the determinism/indeterminism issue. The physicist Gisin, comes to mind:

A possible definition of a Realistic Physics Theory
http://arxiv.org/abs/1401.0419

I wouldn't be so quick to equate non-realism with instrumentalism. Realism and non-realism stand on their own as epistemological platforms. In my opinion, instrumentalism per se, need not. There are plenty of researchers bearing realist sensibilities whose first order of business each day is to put the need for realism on the shelf and confine their work to interpreting the relationships between their data. I'm confident that the men and women who manufacture industrial lasers, accurate enough for microsurgery, are not bothered that the theoretical foundations for laser light can only produce probabilities.

Regarding the Gisin article, Gisin's proposed redefinition of what constitutes a realistic theory made me realize that the debate isn't limited to realism vs. non-realism and determinism vs. indeterminism. It is also a debate about reductionism vs. holism. Gisin's analysis of what is necessary to devise a better definition of "realistic" theories is so steeped in his unspoken dedication to reductionism, that the irony (to me) of applying it to something so fundamentally holistic as reality is almost embarrassing.

Gisin writes that for a theory to be realistic, "The state of physical system must determine the probabilities of all possible measurement outcomes...and at least some physical properties must be written in the physical system...(if not how would one even recognize the system?)."

By this statement, it seems clear that Gisin very correctly sees the need to be able to designate what actually constitutes the system being investigated, yet he also seems incapable of understanding that any such designation is always going to be merely a line drawn in the sand: a purely epistemic creation whose parameters are circumscribed by the limitations of the devices used to investigate. With such artificial circumscription, is there any wonder why realist interpretations of QM end up "groping in the dark" to fill in the holes created by their presuppositions?

Gisin concedes that "mixed states [where all states of the system do not correspond to a unique probability distribution] do not satisfy [his] proposed definition." But he cavalierly dismisses this glaring problem by then stating, "This should be no surprise, as mixed states describe situations involving some ignorance, either about the precise state preparation or about a part of the global system which as been ignored..."

It's hard for me to imagine how such a brilliant researcher can so evidently miss the significance of his own words. Because, to a holist, ALL systems ignore the global system merely by being defined (circumscribed) by the experimental arrangement! I can only assume that Gisin's reductionistic paradigm of analysis allows him to ignore whatever he needs to, so long as he can say something consistent about whatever he has decided constitutes the system under study. After all, reductionism always worked for mechanics before, so why shouldn't it work even after the advent of quantum mechanics?

But QM is a game-changer, and realists just don't seem to want to accept that. And reductionism will not work as an analytical paradigm if the phenomenon under study is fundamentally holistic, feeding back the Whole into the operation of the parts. One of the things I do admire about David Bohm is how he eventually realized that non-locality is ultimately analyzable only from a holistic paradigm.

A Poster He or I wrote:I wouldn't be so quick to equate non-realism with instrumentalism. Realism and non-realism stand on their own as epistemological platforms. In my opinion, instrumentalism per se, need not. There are plenty of researchers bearing realist sensibilities whose first order of business each day is to put the need for realism on the shelf and confine their work to interpreting the relationships between their data. I'm confident that the men and women who manufacture industrial lasers, accurate enough for microsurgery, are not bothered that the theoretical foundations for laser light can only produce probabilities.

I agree that some instrumentalists try to be a bit non-committal on the realism/non-realism issue. But I consider them to be quite similar. The anti-realist subscribes to the non-reality of "unobservable" entities, while the instrumentalist might remain neutral on the issue. Let me ask you this question. Do you believe that electrons and other particles have properties independent of measurement; that is, that they have objective properties at all times, whether or not they are measured or not and whether they are measurable or not? And note I'm not talking about observables, here as there's a difference:

According to realists, an observable is an outcome of a measurement and can either be a direct revealing of the underlying property, or a result of multiple influences such as all the objective properties of all the entities and variables involved at the moment of measurement. All of the above is consistent with realism. Realists make no claim whether direct observation of particle properties in quantum scale particles are possible...

No, I don't believe quanta have properties independent of measurement. I believe electrons, photons, neutrons, muons, pions, on and on, are only labels for mathematical objects. Mathematics exhibits a high degree of correlation with human experience, and humans tend to ontologize their experience rather than admit that experience is explicable only to the extent that it can be correlated to objects and relations within pre-existing epistemological frameworks. The need to justify our need to ontologize drives us to seek evermore fundamental relations, but that effort ultimately exposes the limitations of our epistemological toolsets. The fact that science is able (eventually) to proactively evolve epistemologically is what makes science humans' greatest intellectual achievement. I suspect that the effort to make non-locality scientifically explicable will eventually drive physics to its next paradigm shift.

Because symbolic association (namely language and mathematics) has most utility when it is highly correlated with human experience, I accept that highly successful theories can provide "windows" to even more fundamental relations between phenomena, but these new relations are still only explicable within existing epistemology, and the evidence to date suggests that these newly discovered relations are precipitated by evolving epistemology, not reflective of underlying ontology.

For example, we went looking for non-locality once QM suggested it could happen (we actually went looking to disprove its possibility!). Now that we've empirically verified that non-locality happens, can we really say anything ontological? We can't even integrate non-locality very effectively into our own epistemologies, so how are we in any position to speak of its ontology? The statement "The universe is non-local" is not a statement about the ontological nature of the universe; it is a statement of the current status of scientific epistemology. I acknowledge that non-locality is now a feature of human experience, but the status of human experience hasn't been illuminated because of it. Ontologically, I may still just be a brain in a laboratory vat.

I see no evidence whatever that anything ontological has ever been discovered or recognized by humans, at least not since the advent of language. Furthermore, I consider ontology per se utterly irrelevant to human experience. Ontology, like all abstract concepts, is an invention of epistemology.

A Poster He or I wrote:No, I don't believe quanta have properties independent of measurement. I believe electrons, photons, neutrons, muons, pions, on and on, are only labels for mathematical objects.

But surely there's a difference between mathematical objects and the "stuff" they represent. Consider "gravity". It affects rocks, humans, squirrels, planets, stars, galaxies, etc. I understand that our present mathematical model/theory of gravity is likely not final as we cannot foresee what a future theory of quantum gravity, etc. will be like, but our model is talking about some aspect of an observer independent reality; that is, the mathematical structure of the theory refers to or represents something that there is in the world independently of our mathematical theories. There are physicists like Tegmark who do subscribe to that Platonic, 'mathematical universe' picture but I think they are confusing the map for the territory.

And one surely can be a realist and have no difficulty with the idea of spacetime as emerging from some non-spatiotemporal structure (e.g. non-locality), that we may not be able to presently or perhaps ever conceptualize. As Gisin points out:

It might be interesting to remember that no physicist before the advent of relativity interpreted the instantaneous action at a distance of Newton’s gravity as a sign of non-realism (although Newton’s nonlocality is even more radical than quantum nonlocality, as it allowed instantaneous signaling).

Is realism compatible with true randomness?
http://arxiv.org/pdf/1012.2536v1.pdf

Our model is talking about some aspect of an observer independent reality; that is, the mathematical structure of the theory refers to or represents something that there is in the world independently of our mathematical theories.

The "stuff" that exists in reality, independent of our models, is EXPERIENCE. The representational models we make are NOT models of an observer-independent reality; they are models that parse, consolidate, and systematize human experience. It is, of course, very important for me to concede that the productive modeling of experience, which is the basis of Reason and is exemplified by science, only works because experience demonstrates tremendous CONSISTENCY.

But I simply and categorically refuse to assign any a priori fundament to explain the existence of this consistency. I consider the presupposition of observer-independent reality to be -- at best -- irrelevant (and at worst, inimical) to the exercise of reason and science. It is sufficient and responsible to acknowledge that consistency of experience happens and is the basis for all epistemological construction.

Naturally I wonder why experience is consistent, but since I am a holist, I consider the idea of a reality independent from humans to be a contradiction in terms. Nothing is independent of anything except in the cognitive schemata of sentience that has reached a state of complexity sufficient to yield self-reflection and a sense of spatio-temporal extension, allowing for delineation, disambiguation, and identification. Since cognitive schemata circumscribe all attempts to explore the basis for consistency, I accept that humans are not equipped to "step outside" of reality in order to objectify it (mystical experience notwithstanding; but that's not for the science forum).

Even if I did feel disposed to grant an independent existence to reality, with not a sentient mind in the universe to consider it, such a reality offers no insight or solution to the issue of reality's ontological status. It just adds another layer to the issue.

There are physicists like Tegmark who do subscribe to that Platonic, 'mathematical universe' picture but I think they are confusing the map for the territory.

I have no regard for Platonism, so if Tegmark believes in the Platonic existence of numbers and numeric operators then I feel he is wrong, same as you.

And one surely can be a realist and have no difficulty with the idea of spacetime as emerging from some non-spatiotemporal structure (e.g. non-locality), that we may not be able to presently or perhaps ever conceptualize. As Gisin points out:

It might be interesting to remember that no physicist before the advent of relativity interpreted the instantaneous action at a distance of Newton’s gravity as a sign of non-realism (although Newton’s nonlocality is even more radical than quantum nonlocality, as it allowed instantaneous signaling).

It sounds like Gisin makes the same mistake as Maudlin (whom you cited earlier) in thinking that non-realism implies non-existence. It doesn't. Non-realism implies that existence (i.e., ontological status) is outside of science's purview, on science's own terms.

Personally, I feel science should always continue forever to look for cause-and-effect, reductionistic modeling to explicate new phenomena because these are paradigms with excellent track records for utility. But it is a mistake to give up testability just so we can continue to confine ourselves to such models. If Bohm models a non-spatiotemporal quantum potential or full-blown "Implicate Order" we must have testability, otherwise such models just stand in the way of scientific progress by insulating science from that which science is not yet prepared to explain. I hate to say it, but that is exactly what dogma does.

A Poster He or I wrote:The "stuff" that exists in reality, independent of our models, is EXPERIENCE. The representational models we make are NOT models of an observer-independent reality; they are models that parse, consolidate, and systematize human experience.

I don't understand this. I agree that everything is filtered through our cognitive structures/minds and those are limited like that of all organisms and will never give us some perfect match to mind-independent reality, but doesn't something "outside" our minds nudge us in one direction versus another in terms of theory formation. Otherwise, it seems, if I'm understanding your argument, your position would lead to solipsism.

I

agree that everything is filtered through our cognitive structures/minds and those are limited like that of all organisms and will never give us some perfect match to mind-independent reality, but doesn't something "outside" our minds nudge us in one direction versus another in terms of theory formation. Otherwise, it seems, if I'm understanding your argument, your position would lead to solipsism.

In my opinion, all reason-based philosophy leads to either tautology or solipsism if analyzed far enough. Solipsism is logically irrefutable, which frankly says more about the limitations of logic than about the legitimacy of solipsism (which is a model of very low utility). The only reason I can imagine that you would target my position specifically as solipsistic is that you think I am trying to ontologize subjective experience when I claim experience as the subject of our scientific (and philosophical) models.

But I don't ontologize anything. I have stated consistently that the ontological status of reality is irrelevant. I am only interested in establishing what makes models useful, how we can develop even more useful models, and what can we do when our models meet their limits.

A model is most useful when it integrates with a very broad spectrum of human knowledge. Under scientific epistemology, knowledge is that which has been empirically verified. There is no scientific possibility of verifying the existence of anything independent of our minds simply because science makes critical utility of the mind (cognition) as part of empirical verification. Don't get me wrong: I'm not saying mind is the only thing that exists (I'm no Idealist, no sir!). I'm saying that anything actually independent of mind would be unrecognizable and therefore not within the purview of scientific inquiry. The exercise of reason (and therefore science) begins only once the subject under scrutiny has become a cognitive gestalt: an object of the mind.

Empirical verification works mostly because human experience exhibits consistency. Each of us experiences such consistency as subjectively true, and there is broad consensus among the culture of science about what is consistent (though not always why). We then project hypotheses to explain this consistency by positing interrelationships across as many phenomena as needed to ground the hypothesis in the consensus of experience.

That's why I said models model experience.

A Poster He or I wrote:Empirical verification works mostly because human experience exhibits consistency. Each of us experiences such consistency as subjectively true, and there is broad consensus among the culture of science about what is consistent (though not always why). We then project hypotheses to explain this consistency by positing interrelationships across as many phenomena as needed to ground the hypothesis in the consensus of experience.That's why I said models model experience.

I think I understand some of your points but I'm not convinced by most of your arguments. You say models model experience. But experience of what? Sure, any theoretical term will necessarily fail to capture the world correctly or truthfully (as it is in itself, etc.) since we are "prisoners" of our cognitive structures, so to speak. The world as it is, in itself will necessarily escape our characteristics of it. But our phenomenal realm and theoretical models aren't purely arbitrary merely "spinning in void". They are causally driven by something external to us; that is, there is a reality that underlies our observations for surely something affects our senses and measurement devices. As Timpson writes:

However, it is important to recognise that there is a very obvious difficulty with the thought that what can be said provides a constitutive contribution to what can be real and that physics correspondingly concerns what we can say about nature. Simply reflect that some explanation needs to be given of where the relevant constraints on what can be said come from. Surely there could be no other source for these constraints than the way the world actually is-it can't merely be a matter of language. It is because of the unbending nature of the world that we find the need to move, for example, from classical to quantum physics; that we find the need to revise our theories in the face of recalcitrant experience. Zeilinger and Bohr (in the quotation above) would thus seem to be putting the cart before the horse, to at least some degree. Schematically, it's the way the world is (independently of our attempted description or systematisation of it) that determines what can usefully be said about it, and that ultimately determines what sets of concepts will prove most appropriate in our scientific theorising. It is failure to recognise this simple truth that accounts, I suggest, for the otherwise glaring nonsequitur in the proposed answer to `Why the quantum?'...Of course, what statements can be made depends on what concepts we possess; and, trivially, in order to succeed in making a statement, one needs to obey the appropriate linguistic rules. But the point at issue is what can make one set of concepts more fit for our scientific theorising than another? For example, why do we have to replace commuting classical physical quantities with non-commuting quantum observables?

Information, Immaterialism, Instrumentalism: Old and New in Quantum Information
http://users.ox.ac.uk/~bras2317/iii_2.pdf

Maybe Louis de Broglie was right after all: http://www.simonsfoundation.org/quanta/ ... m-reality/

http://www.nature.com/news/quanundrum-1.15415