reality

Transcript of Realism, Anti-Realism and Non-Realism 29/6/04





This week I want to try to make sense of what was discussed in the

previous weeks, while looking a little closer at what the underlying reality

behind our scientific descriptions might be, assuming such a thing exists.





In the first week I briefly explored the metaphysical basis of concepts

such as spacetime, matter and energy, and in the second looked more

closely at causation, law and determinism. One important thing

noticeable about these was the many current difficulties concerning their

exact definition, either ontologically or analytically, and also their

conceptual interdependency. Energy and force for instance being defined

in terms of matter undergoing change, while matter is definable in terms

of force or energy. Similarly space traditionally described as holding, or

dividing, material objects, and so defined in terms of matter, is today

almost seen as the precursor of matter, itself definable as a series of

events in spacetime. This interdefinition with no reference to a

foundational reality is taken by many (myself included) to indicate that

we are dealing with an internally defined conceptual system rather than a

reflection of an external reality. But before I look at this possibility, what

are the alternatives?





The orthodox conservative view perhaps is still Realism. That our

perceptions reveal an external reality of somekind. Though by this I do

not mean so called naive Realism, the view that the everyday reality is

what it appears to be, but rather that the reality revealed by science is real.

But if this is the case how can we account for the difficulty in defining

this world and the interdefinability of its

constituents? One way to do the latter is through cofoundationalism. The

idea that, for instance, both matter and energy are jointly real and so

definable in terms of each other. This is an attractive option, however it

raises problems of temporal evolution, what comes first matter or energy?

If we choose either we lose the ability to define it in terms of its partner.

If we choose both we enter into a difficult conceptual problem, are they

identical and if so why are they apparently different, and if they are

different what correlates their joint appearance? One way out of this

might be to define matter and energy jointly emerging from spacetime.

However relative spacetime is traditionally conceived of in terms of

matter so we lose the possibility to define it before matter

exists. We can of course describe spacetime purely mathematically and

account for matter, energy or their combination emerging from this

spontaneously, but these descriptions seem to require energy as a joint

foundation with spacetime, so we just move the problem back. Some

modern theories attempt to reduce everything to energy and particles of

space but this is not really helpful as it is still cofoundational. Although it

does perhaps help explain general causal patterns, or laws, as the arbitrary

structure adopted by these ‘bits of spacetime’ in the overall universe. All

these various theories are problematic while a single foundation remains

illusive.

One possibility open to the Realist is to unify these foundational items in

some way. Most likely through a field theory that closely relates a

universal field of force with the structure of spacetime, without reference

to material particles or waves as such.

This also has the benefit of overcoming some of the problems of

Quantum Mechanics and unifying it with Relativity theory.

Note that this also implies that our everyday perception of a pluralistic

world of individual objects is largely illusory and so goes a long way

down the conceptual scheme route. Though of course maintaining a close

correlation between this scheme and reality.

This has not yet been achieved however and poses a number of problems,

not least of which is the complexity of the system and the difficulty of

computing a whole universe in a single mathematical field description.

Complex systems of this kind can often only be represented as a

simulation rather than a normal scientific description, because the

mathematical formalism is simply too complex to compute. But in this

case how can the entire universe be simulated in a system that is itself

part of that universe? It seems only local descriptions may be possible

and so we have to return to a non-holistic solution, at least for the sake of

conceptual simplicity.

Any talk of the overall foundation of the universe would thus be

indescribable and so mere metaphor held true through faith. Or in

other words scientific Realism becomes religious belief.





A more common move by the Realist however is to argue that the

problems outlined earlier (indefinability, QM contradictions etc) are in

fact pseudo-problems based on our current lack of understanding of the

world. The sceptic is being too pessimistic, the Realist often claims. But

for many this entails an awful lot of faith on the part of the Realist, as it

assumes that all the many logical and ontological problems in the

philosophy of physics will one day be solvable, and even that quantum

mechanics will eventually

be superseded by a new more coherent and less problematic theory. But

this seems to require the kind of impossible conservative optimism, akin

to that found in the Church at the onset of the Enlightenment.

Furthermore even if current physics was superseded, who is to say that

the next, or even the ultimate, theory will not be even more problematic

for Realists. Unless physics is benign enough to be completed only in a

tidy, conceivable little theory attuned to Realist prejudice.









Another possible solution is the opposite to Realism, that of Anti-

Realism. This term has two meanings however, in the States it simply

means a scepticism towards Realism, and refers to views such as

Instrumentalism (the idea that objects like atoms do

not really exist, but are merely useful working models for the description

of reality) and related Neo-Kantianism (the idea that phenomena are

conceptually moulded percepts, that only indirectly correlate with reality,

rather than being direct representations). This view, which I shall refer to

instead as Non Realism, agrees with Realists that there is an independent

(or at least semi-independent) external reality, unresponsive to our

desires, but denies the Realist’s claims that we can know it, as it really is,

through

our conceptual schemes. Thus for instance while the Atomist theory as a

whole correlates with reality as a whole (to the extent of predictability)

the units of the theory, atoms, do not correlate to anything in reality, they

are just conceptual tools existing in theory and our imagination. It is thus

not a true Anti-Realist position. But I shall return to this shortly.





First I want to look at a completely different use of the term Anti-

Realism, its use in the UK by philosophers like Michael Dummett, who

coined the term. This view, almost the opposite of Realism, and others

like it, is basically an Idealist position which states that we can really

know of the world through our conceptual schemes, simply because the

world is really nothing but a conceptual scheme. Note Anti-Realism is a

bit of a misnomer here as the exact opposite of Realism would be a reality

dependent on our minds that we could not know, rather that a mind

dependent reality that we can know by definition, but this would be an

absurdity.

Dummett’s thesis is not purely Idealist or Phenomenalist (a Neo-

Kantianism of concept-percept without the underlying reality), but

approximates to it because, like some interpretations of Quantum Theory,

it posits situations in which a description can be neither true or false, until

a positive assertion is made. Thus it is language, according to this reading

of Dummett’s theory, that ultimately shapes the material world. Many

philosophers equate this with Idealism, the thesis that only minds and

ideas really exist. The Idealist tradition ultimately traces itself back to

Bishop Berkley who claimed existence was identical to perception. This

primitive form of Anti-Realism is easily refutable however, through the

appeal to a stable shared reality, if my perception shapes my world and

your perception shapes your world how can we both share the same

reality. Such a thesis seems to lead to solipsism. More worryingly what is

it that preserves the continuity of the world when no one is observing it.

The answer to both these questions for Berkley was God, as the ultimate

observer. It should be noted that exactly the same problem may emerge

for the Princeton Interpretation of Quantum Mechanics, which states that

the universe is an observer created one, as it is observation that is said to

collapse the QM probability function. However this whole argument

seems to be a reductio ad absurdum, and not least for its use of an

observing divinity as the ultimate source of reality (what if he nodded

off!). In contrast however Phenomenology and Dummett’s Anti-Realism

sidestep the God hypothesis by using inherent mental categories and/or

arbitrary shared language as the guarantee of a shared stable order. In a

conventional reading, this just seems to move the problem on to another

area however, in that it fails to answer the obvious question of how

language or mental categorisation achieve a stable order without any

foundation themselves. The only answer seems to be somekind of

structured disembodied consciousness communicating telepathically,

which not only seems to many as absurd as the God hypothesis, but more

importantly also flies in the face of our everyday experience of a material

reality (and rejects any notion of sensual existence). We know that an

external world exists simply by the fact that it sometimes resists our

actions in it. This table exists because it offers a resistance to any of our

attempts to exert a force on it. It is simply real in this sense. For this

reason most philosophers regard Anti-Realism as absurd, and not a

serious challenge to scientific Realism.





But things do not end there. It is now time to turn to what I’ve called Non

Realism. As previously stated Non-Realism basically encompasses all

those views that can be categorised as Neo-Kantian. The notion that the

world of phenomena we experience, and the theoretical models of the

world we generate, are not precise reflections of an external reality, but

are rooted in conceptual schemes that only indirectly match a reality,

which we can never perceive or conceive of in its pure form. This offers a

simple solution to the philosophical problems we currently have in this

area as it explains the interdefinable nature of all the elements of the

scientific model. This kind of interdefinability is exactly what we should

expect of an internally coherent conceptual scheme, the truth of which is

independent of an external reality and entirely internal. Logic itself thus

becomes a model of the structure of our minds rather than of the world.

While at the same time it does not deny the obvious existence of an

external reality. Non Realism

can thus combine the best ideas of Realism and Anti-Realism.

This view has a few problems associated with it of course, one is if logic

is merely mental structure what does this say about the way mathematics

relates to the world? Traditionally maths, like logic was seen as the ideal

model of the world and the most authentic representation of it. This also

has bearings on the mathematical models of QM. On the Non Realist

account though mathematical models are just that, more conceptual

schemes by which we understand the world, rather than the way the

world is. This view is actually supported by developments within

mathematics itself however. Goedel’s Theorem is widely regarded to

demonstrate that no mathematical model (or at least one involving

arithmetic) can ever be complete and entirely self-consistent. Something

is always left out of the equation. This can be best understood in terms of

Set Theory, which has its own classic paradox, the problem of self

reference, exemplified in the example of the library with two catalogues,

one listing all the books that refer to themselves and the other listing all

the books that don’t refer to themselves. The paradox being that while the

first catalogue can be included in itself without contradiction, the second

catalogue cannot be listed in either itself or the first catalogue without

contradiction. It can not be included in the overall system. Goedel’s

Theorem is similar in that some elements of any arithmetic system cannot

be defined within that system, it is thus incompletable. At first this was

thought to be a minor problem with a small amount of indefinability.

However

more recent studies have shown mathematics is more like a sea of

indefinables with a few large islands of coherence. Thus if maths

is a reflection of reality it must be a very strange reality. What is more

likely is that mathematics is a conceptual scheme that imperfectly

represents reality.

Some have argued that if all we have is an internally consistent

conceptual scheme that is not a reflection of reality then there is an

unbridgeable gap between this scheme and reality, in fact we have no

reason to surmise this reality exists at all. But this degeneration into

Idealism is not necessary, as the very fact of a conceptual scheme

successfully applied to a world entails the existence of that world. What’s

more the fact that our conceptual models of Causation, for example, fail

to capture what we intuit to be Causation, tends to indicate that Causation

itself is something other than a set of ideas in a conceptual scheme, and

so refutes any Idealist account of Causation. What it really is though does

raise problems. As does the status of ‘quantum logic’, incidentally, a

bizarre form of logic designed to describe the world of quantum reality.

Some philosophers argue that this is the true logic of thought (and only

approximated by classical logic), which intuitively seems wrong, given

quantum weirdness, while others regard it as the basis

of the only true description of the world. Though this flies in the face of

the view of maths and logic as conceptual schemes just outlined.

Although, given the discontinuities and ambiguities of quantum reality,

an incomplete patchwork of bizarre logical description might just be a

true representation of reality after all!





This leads us back of course to Niels Bohr’s Complementarity theory,

arguably one of the few interpretations of QM to survive a rigorous

critique. As we saw last week Complementarity describes quantum reality

in terms of various mutually incompatible but jointly necessary

superposition descriptions, that is, wave – particle descriptions,

momentum – position descriptions, and so on. Something which only

seems plausible under a Neo-Kantian philosophy, which attempts to

describe the world in terms of a logically incompletable conceptual

scheme (ala Goedel), that can only be expressed in partially complete but

logically incompatible descriptions. The only complication here being

Bohr’s claim that a wave and a particle description could not be applied

to the same experiment, or a contradiction in a single description would

result (as a wave is by definition extended in space and a particle can only

be in one place), while recent experiments seem to have demonstrated

simultaneous wave and particle nature. What this means is unclear,

though arguably it could demonstrate either an illogicality to reality, as

measured in experiment (or rather the failure of our mental categories to

match reality in atypical situations), and/or perhaps the need for a true

objective logic of a totally different kind. But either way the actual reality

of the situation could not be described coherently as either a wave or a

particle (and arguably neither as anything else in practise) using

conventional reason, which further confirms the idea that we are dealing

primarily in conceptual schemes when we describe the world. While of

course also showing that a reality underlies this conception which cannot

in itself

be conceptually represented. Note further experiments have denied a

coherent underlying reality of hidden variables and shown

that the superpositions are real and prone to objective probabilities, as

well as conceptual and prone to subjective probabilities (a controversy

even in Copenhagenism till then). This would seem to indicate there was

an odd parallel between the nature of our conceptual models and the

structure of reality, though not one compatible with conventional

Realism.





Obviously Non Realism opens up problems regarding the nature of truth.

But these are not necessarily serious ones, for instance van Fraassen

argues for the concept of Empirical Adequacy replacing classical notions

of truth (at least in scientific contexts). Here

we do not say something is ‘true’, but simply that it matches the data

adequately. Bearing in mind several interpretations of data

can match the same set of experimental results with equal adequacy. If

some of these interpretative descriptions are equally ‘true’

(in the traditional sence), or rather are sustainable and irrefutable (in the

scientific sence), but have a tendency to be incomplete,

then we return immediately to Complementarity.





Despite the attractions of this view there are some serious problems with

a Neo-Kantian Non Realism. The most serious being the actual

connection between the conceptual scheme and reality. We seem doomed

to an even deeper form of Cartesian dualism if we cannot account for the

overlap of reality and our models of it. This remains a very difficult

problem, for if as Kant claims we cannot experience the world without a

conceptual scheme then how do we discover a correspondence between

concept and reality at all? The fact that concepts are necessary for

perception also seems true from studies of physical perception in different

cultures. So how does our conceptual world come to match reality close

enough to allow us to survive in the world if it is only produced by

random Darwinistic natural selection? It would seem incredibly

improbable that such a conceptual scheme would emerge this way.





One bold solution is through the philosophy of a philosopher who

followed in Kant’s footsteps and claimed to have superseded him, G W

Hegel. Hegel argued that all progress was dialectical, in that two

opposing notions could be closely related though their mutual opposition

but interdependence. It follows from this that our conceptual scheme of

the world may be in dialectical relation with the reality beyond this

scheme. But what could this mean in practise, how can a detached

conceptual scheme get to grips with reality in the first place to even begin

a dialectical relation? I don’t intend to go into the complexities of

Hegelianism here but a few more basic observations might be useful.





For one Hegel maintained that the our concepts and reality while not in a

relation of mirror and object are none the less closely entwined, that is

they are mutually dependent on each other. Ideally most neo-Hegelians

would like this interdependency to be a

material one with a conventional epistemology, though it is not

impossible that a form of participatory epistemology is involved, that is

that our knowledge of the world is not passive but active, and modifies it

in someway. What this means is difficult to grasp, it is not Idealism, as it

is not saying that only minds and ideas exist, so it is not an Anti-Realist

stance, although it is obviously not a Realist one either. It seems to

suggest that reality is only independent of our perception up to a point

and remains in some sense incomplete until we have knowledge of it.

This is very close to some interpretations of QM, though it should be

emphasised this is not the same as the observer effect put forward by

Wheeler (whether this is interpreted as a form of Idealism or not) as we

are not dealing here with individual observation but with shared

knowledge. This may sound odd, but there are empirical indications that

it could be true, and not all of these are from QM.





At the turn of the last century it was conventionally thought that glycerine

could not crystallise, however when a piece of crystalline glycerine was

discovered (there are various accounts of how this happened) and the

news spread, it is claimed glycerine began to be crystallised by applying

deterministic methods that had previously failed. Sceptics argued that

microscopic fragments from sample glycerine crystals sent to the labs had

contaminated the results, seeding crystal formation, and where no

samples were sent crystals were said to have been carried in the beards of

visiting scientists. However when glycerine allegedly crystallised

spontaneously in ‘sealed jars’ something odd seemed to be happening.

The records of these events are to be in the scientific literature of the

period, though unfortunately the details are a little too vague to reach

definite conclusions. One more interesting sceptical account argued that

the crystallisation of glycerine was always possible but had a very low

probability, and all that had happened was that an improbability occurred

at this time, akin to a long run of heads in a series of coin flips. There

being no mystery. But this only works if we see the initial discovery of

crystalline glycerine as a separate coincidence (unless this itself

spontaneously appeared). Oddly this explanation may support a highly

controversial claim which I shall now explore.









It should be remembered that there are outstanding problems with QM

even if we take a Neo-Kantian view. There seems to be a real

experimentally demonstratable reality beyond our conceptual schemes

that indicates that superposition is real. That the duality of quantum states

is not just a matter of description of a complex reality, in which our

normal conceptual categories don’t apply, it is a real state of an

ambivalent quantum reality. That is our contradictory descriptions do to a

certain extent match a ambiguous reality in some sence. But as we saw

last week objective superposition is hard to explain. One solution

however is Decoherence theory.

To recount from last week, one interpretation of Decoherence theory

suggests a simple quantum system (ie a coupled electron pair) requires

little information to represent it, while a large quantum system (such as

Schrödinger’s cat) requires an enormous amount of information to

represent it (three times the number of its particles at least). The latter is

thus inaccessible to a limited representational information system (the

human mind). Thus to human observers the world appears classical, but is

really quantum mechanical. Measurement connects a simple system to the

larger world making it appear classical. This is sometimes used in

Copenhagen interpretations to explain probability wave collapse, even

though the wave function doesn’t actually change, it just appears to.

Similarly it is used to explain why we only perceive one history in a State

Relative, ‘Many Histories’ interpretation. But if this is so, and all states

remain real they can still interact, so why isn’t reality weirder!? A

speculative answer may be that the information we have about the world

actually changes how it behaves. Or more precisely modifies the

probabilities about how it will behave. Thus we have a perfect

explanation for the glycerine effect which involves no coincidences. The

information that glycerine could crystallise, when shared by a critical

mass of scientists, actually changed the probabilities of its spontaneous

crystallisation. This might sound bizarre, but experiments do seem to

indicate the incredible fact that random systems can be influenced to a

small extent by the human mind. Though these experiments are ongoing.

If this turns out to be true then we have a real mechanism for a

participatory epistemology akin to that proposed by Hegel. Although the

conceptual details would have to be fleshed out to say the least.

Furthermore given the theories of David Chalmers in Philosophy of

Mind, that all systems that hold information (even mechanical ones) are

by his definition ‘conscious’, and so are bearers of knowledge in some

sense, it might also be possible that a participatory epistemology is

possible even without a human mind being present, thus avoiding obvious

problems with an account of the physical evolution of the universe. While

this remains a highly speculative and controversial thesis, given the

absurdity of Anti-Realism and the almost ridiculous optimism of Realism,

this form of Non Realism remains a serious option. Though it is also quite

possible that some yet unknown less radical or hybrid position is

applicable.





Finally I’d like to apply this last thesis to the topic of free will I raised last

week. Then I stated my belief that moral choice

and personal responsibility depended on the existence of free will to be

part of a coherent ethical theory. I further observed that deterministic

accounts of science and deterministic interpretations of QM excluded the

possibility of the kind of free will required

by a theory of this sort. And so declared my preference for indeterminate

interpretations of QM, such as the Copenhagen Interpretation, with QM

corresponding to the true physics of the world (as it does with

Decoherence theory). However while this does not exclude free will it

does not make it possible either. The orthodox account of the

Copenhagen Interpretation merely states that the results of measurement

are determined probabilistically, with the actual result being effectively

random. Not the kind of background theory to make free choice a

possibility. However as we have seen there is a case to be made for the

modulation of probability through knowledge or information, and

arguably this process is part of the Decoherence effect that creates the

appearance of classical physics. If the human mind was one of the

information systems that could modulate probability in this way the

possibility that the results of measurement are not random after all, but

based on the conscious phase shifting of the wave function by the

modulation of its probability content. Thus a conscious decision process

could be taking place within the brain, making free will a component in a

wider deterministic system. While this is a speculative idea, it has to be

taken seriously due to its arguably currently closer correspondence with

empirical reality. It remains however only a sketch of a solution and

would have to be developed much further to make it a viable thesis. In

particular a mode of interaction between knowledge or information and

material reality would have to be devised. One way might be to combine

Ernst Mach’s version of a neutral monism, or Chalmer’s aspect dualism

(a broadly equivalent thesis of the mutually irreducible ‘mental’ /

‘physical’ properties of matter), with the latter’s theory that

consciousness is basically a physical system storing information, where

the information has a conditioning effect on the conscious aspect of the

dualism, and further conditions the material aspect of the dualism.

Something also understandable in terms of probability, which currently

has two definitions, subjective or epistemic probability (the degree of

certainty in our knowledge) and

objective or ontic probability (the likelihood of an event occurring).

Conventionally viewed as two separate phenomena, they could also be

viewed as two aspects of one phenomena under aspect dualism. A

definition of will within this context would also have to be found, perhaps

as an assertive proposition ( ‘I will …..’), within the context of a

propositional theory of knowledge. This might form a domain of local

knowledge able to modulate probabilities in the immediate environment

of the subject (the human brain for instance) and thus make the

indeterminate wave function collapse a subject determinate collapse,

constituting a form of agent causation within a wider deterministic

framework. For such a concept of local knowledge to make sense

however it would not only have to be more closely related to an act of

volition, but also the neural processes associated with that act of volition.

There would also have to be way of describing the interaction of global

knowledge (shared beliefs about the world) with local knowledge, while

at the same time maintaining the semi-autonomy of the ‘laws of nature’.

Such a task may be daunting but necessary.