At the heart of modern day realism lies a huge contradiction. And the most incredible fact about this, is that the majority of those who have thought about it are aware of it, but are still prepared to brush it under the carpet as an inconvenient truth, best ignored. However, the failure to recognise this comes with real world implications.
Background
There are many different strands to realism, but at it its heart, it may be thought of as the rather intuitive claim that the objects we experience such as tables, chairs, spoons, aeroplanes etc exist, and that this is so, independently of anyone’s beliefs or perceptions. If we close our eyes for a few minutes or cease to exist, these objects do not vanish; they are still in existence, despite the fact that we are no longer directly perceiving them. This contrasts with the idealist position which claims that the objects of our experience are fundamentally mental, mentally constructed or otherwise immaterial. It raises the possibly, highlighted in films such as The Matrix, that we could be living in a dreamworld, blissfully unaware of the objects that ‘really’ exist, or even that there could be no such ‘real world’ at all – that you and the figments of your imagination may be all that there is.
The purpose of this blog is to highlight the inconsistency in the realist position but then go on to show that despite this, it remains possible to achieve the realist goal of obtaining good grounds for accepting the view that the objects of our experience are ontologically primary, albeit that some concessions need to be made.
While the view that the objects of our experience exist independently of our perception of them seems reasonable on the surface, a little probing suggests we need to think a bit more carefully about what is being claimed here. It is now well established, through an examination of their sensory apparatus, that not all creatures perceive the world in the same way. So for example, when a human looks at an aeroplane, they are having a different experience to when, say, a dog looks at an aeroplane.
If our experience of the aeroplane is affected by the way it is perceived, then what can be claimed about the aeroplane as it really is – as it exists independently of our perception of it? The 17th century philosopher John Locke discussed objects in terms of their primary and secondary qualities, the secondary qualities being their subjective features and the primary qualities, such as mass and size, being intrinsic to the object. Of course, traditional empiricism or idealism argues that there is nothing that can be claimed about the aeroplane that is not an aspect of our perception of it, even its size and mass, but the realist disagrees. The realist might say that the aeroplane, as it exists independently of its subjective features (as a ‘thing-in-itself’), is the collection of nature’s building blocks, or atoms, that make it up. If this claim is to have substance, the question is then raised as to what these building blocks are.
Before c. 1909, the atom was understood as a uniform sphere of positive charge with electrons embedded within it (known as the ‘plum pudding model’). That was until the Rutherford scattering experiments showed that the positive charge was concentrated at the centre. This demonstrated conclusively that the plum pudding atom does not exist and highlights a problem without end. Today, the nature of the most fundamental particles of physics is inferred from theories that give predictions that are best in line with experimental results. But the jury will always be out as to whether nature is actually structured as the theory implies or whether it just appears to be structured this way. The famous ‘No Miracles’ argument put forward by Philosophers of Science such as Hilary Putnam, argues that science gets closer to the ‘truth’ of the way nature is, as it progresses, but the strength of this argument has been seriously undermined e.g. by Larry Laudan in his ‘A Confutation of Convergent Realism’ (1981) when it is pointed out that new theories that result in slightly better predictions often involve a complete overhaul of the ontological picture.
So when pressed as to what it is that actually exists, the realist finds themselves on the defensive. The objects of our experience have subjective features, so the object, as it exists as a thing-in-itself must be different to this. But the nature of these ‘things-in-themselves’ – natures building blocks – is unknown and unknowable. The realist is therefore charged with not knowing what it is that they are claiming exists. However, the realist is not ready to throw in the towel just yet. The realist might still assert that fundamental particles exist in the sense that there is a something that is the ground of why our experiments, such as those conducted using CERN’s large hadron collider, produce the results that they do. And while the exact nature of these particles may be unclear, it is not unreasonable to claim that these particles exist – that there is a one-to-one correspondence between such a particle as we understand it and the particle as it is in itself.
This diluted version of realism remains problematic. If what distinguishes one object from another is either its different structural make-up or its different position in time and space, then it may be that what we understand to be two distinct objects is actually two different aspects of a single underlying object. Or that we haven’t probed deeply enough and that what we understand to be a single object is actually a collection of smaller objects. Or a combination of the two – perhaps the structure that we see is our representation, based on what we can detect, of a completely different underlying structure, much like when the entangled states of quantum mechanics revolutionised our thinking about what it means to say that objects are distinct. There are many different ways the universe could be structured that all result in us perceiving it in the way we do. It appears therefore that the realist claim needs to be diluted even further, to one that says ‘the objects of our experience exist independently of our perception of them, in the sense that they are grounded on something external (structure unknown), rather than being figments of our imagination’.
However, we have now moved a long way from our original definition of realism, and indeed Laudan dubbed this position ‘Realism without the Real’. The plum pudding atom is a model, but doesn’t actually exist. The object we know as an aeroplane may indeed be grounded on something external, but all its features – it’s colour, its size and our understanding of the building blocks that make it up – are aspects of the way we perceive it, not qualities of any underlying aeroplane as-it-is-in-itself. And so the aeroplane, as it is in the absence of our perception of it can’t really be said to be an aeroplane at all. This is because all the things that make an aeroplane an aeroplane – the fact that it is a man-made flying vehicle with wings – are features of how it appears. In the absence of having anyone to do any perceiving, all we have is a something that exists (structure unknown). It requires someone to perceive this structure to turn it into the reality that we are familiar with, complete with the objects of everyday life.
Is this the end of the road for realism? It could be argued that this depends on what the realist wants to establish. Obviously, the idea that the objects of our experience exist exactly as perceived is immediately undermined when it is noted that different observers perceive things differently. But this is not really the claim of realism. It might be said that the primary realist goal is to address the sceptical challenge raised above - the concern that the objects of our experience might be no more than figments of our imagination, rather than being grounded on something external.
Popularised by Descartes in the 17th Century, the sceptic asks ‘how can you know that the objects of your experience are not all just figments of your imagination?’ This raises the worry that the real world may be very different to world you are experiencing. Popular films like the Matrix highlight the possibility that there may be no correspondence at all between the objects of your experience and those objects that ‘really exist’. However, for the realist, this is unhelpful speculation. The sceptic cannot show that the realist is wrong when asserting that there are ’real things’ that form the ground of the objects we experience, and the realist cannot show that the sceptic is wrong in questioning this. We reach an impasse.
Recognising this much, it could be said that most of us take what can be termed a ‘functional realist’ approach. Essentially, the functional realist says ‘well I accept that the sceptical challenge has weight and I can’t rule out the idea that all these things that I interact with every day might not really exist. But for practical purposes, I'm going to assume that our latest and best understanding of the world around us is right. I’m going to assume that the world is structured as described by our latest physical theories and that tables and chairs and electrons and quarks and muons and all these things are real – there is no Matrix and they don't just depend on my perception for their existence’.
All the World is Models
Let's take a look at whether that stands up to scrutiny. First of all you say ‘I accept that things might be different to how they seem’ and then in the next breath you say ‘but actually I assert that things are as they seem’; a direct contradiction. If you accept that things might not be as they seem, then it is logically inconsistent to claim at the same time that things are as they seem. Instead you are forced to move from the realist claim that things are as they seem, to the weaker, non-realist claim that things might be as they seem. This implies that the idea that they are as they seem is a model – a conjecture. Other people might have different models as to which objects ‘really exist’ and these models cannot be rejected as wrong if the sceptical challenge is accepted. If you accept that your model might be wrong, then refusing to tolerate other people using a different model is just chauvinism.
Now your functional realist at this point might argue that this is just an argument over semantics. If you want to be picky, they might say, you can argue that the idea that the objects of our experience really exist is a model, but everyone else will be happy to assert that these things are real and go from there. If everyone is on board with these things actually existing, then it doesn't really make a difference whether you characterise it as a model or not. To this point I would say that this is reasonable when you're dealing with the Cartesian ‘is it all a dream?’ scenario, because it's a very black and white picture – either everything exists or nothing exists. However in real life, there are many more shades of grey. There are lots of objects where the jury is out on whether they exist and science doesn't help us. For example, whether God exist, whether other points of view on the world exist, whether there are other dimensions that we can’t interact with, whether subatomic particles are real or just mathematical constructions, to name but a few. None of these questions can be answered by scientific investigation, because they're all part of this ‘many-to-one’ problem – there are many ways the universe could be that result in the universe that we perceive.
For realism to be credible, we would need to make the case for a one-to-one relationship between the way the world is and the way it behaves. Only then could we infer how the world is from how it behaves. This idea was famously captured by W.V.O. Quine, when he said
But what if… we have achieved a theory that is conformable to every possible observation, past and future? In what sense could the world then be said to deviate from what the theory claims? Clearly in none
In fact, clearly the opposite. For example, there are many versions of string theory that all describe the universe slightly differently, but result in the same predictions at the level we can test. And if you are nervous that Quine may have a point in principle, you need only consider the favourite get-out of Creationists - that the world was created 6,000 years ago in such a way that it appears as if it was created 13.8bn years ago. There are an infinite number of ways that reality could be structured that all result in it appearing to us in the way it does. But if this is the case, how can we ever know which ontological picture of reality is right?
The Map is the Territory
Once it is accepted that any way of understanding our experiences is a model – that it might be wrong – this is easily resolved. You start with that ‘functional realist’ idea of saying "well I know the true nature of reality is unknowable, so I'm going to pick the model that that makes the most sense to me and say that this is my reality". The difference is that we don’t try to pretend that this isn’t a model. As we've acknowledged explicitly that it's a model, the door is still open for other people to pick a different model.
I might say "The idea of God sounds absurd to me, so I'm going to pick an atheist model as my reality". Someone else can equally legitimately say "I prefer this other ontology or this other picture of the way things are" which still gives the same experience at the nuts and bolts level, but provides it with a different context. Contrast that with the kind of chauvinistic functional realist who says that there's only one right answer, despite the fact that they’ve already acknowledged that it's many-to-one when they accepted that the sceptical challenge has weight!
Now there are many who baulk at the idea that you can somehow ‘say that this is my reality’. “We don’t get to choose how the world is” they cry. They want a clear demarcation between ontology (how the world is) and epistemology (what we can know about it). However, I suggest that things are not so simple. In the many-to-one situation that we are in, it is a mistake to focus on how the world actually is, rather than how we model it. This is because if there is no way of inferring which of the many options about the way the world is might be correct, then which one is actually correct is irrelevant! If there are many explanations for the same behaviour, then each one serves equally well. If the true nature of reality is forever unknowable, then it cannot act as a standard of correctness to assess your model against. The only standard of correctness that remains is whether your model produces sensible results. This means that when discussing ontology, we need to shift the focus away from the objects of ‘the universe as it really is’ – this is unknowable and hence no more than unproductive speculation. This is central. To move forward productively, we need to discuss the objects of our models*.
Such a discussion can involve debates about which models are best in terms of their explanatory power or simplicity or the scope of the experiences they describe. But note, these are not arguments about the way things are; they're arguments about what makes a good model. So ultimately that comes down to what you personally are looking to get from your model – what work you want it to do for you. I have dubbed this the ‘Anything Goes’ (‘AG’) approach for brevity – given there is no knowable ‘right answer’ for how the universe is structured, then ‘Anything Goes’ as long as it does what you need it to do as a model.
How does this help us address the sceptical worry that the real world may be very different to world you are experiencing? What the realist want is grounds for an ontology that says the world is structured in line with our intuitive belief that the objects around us exist independently of our perception of them. This would allow us to rule out the Descartes evil demon / Matrix scenario, where there is the world of illusion and the real world. In this scenario, these two worlds are independent – knowledge of the world of illusion doesn’t help you understand the behaviour of the real world and therefore you are exposed to the risk that the real world will ‘get you’ – the robots of the Matrix might tear you limb from limb while you blissfully believe you are cuddling puppies.
Anything Goes thinking allows us to deal with this concern, because unlike with the incoherent functional realist approach, there are no disembodied facts. Facts require statements, which require concepts, which require a context – a framework or a model – to give them meaning. So any statement, such as ‘Angles in a triangle add up to 180 degrees’ or ‘God exists’ or ‘This robot is sentient’ are made within the context of a model, and its truth or falsity can only be assessed within the context of this model. So if under your model, the objects of your experience are real, the statement that they might be given to you by the Matrix is false.
‘That doesn’t help me at all’, I hear you cry. What if my model is wrong and the correct model is that there is an evil Matrix? AG responds by reminding you that there are no disembodied facts. Your model cannot be wrong when the only standard of correctness is whether the model produces sensible results. When it’s many-to-one, there is no single right answer in any meaningful sense. But sadly, although Anything Goes is a powerful approach, it does not allow you to predict the future with certainty(!) and the idea that the bogeyman will unexpectedly jump out and get you is still a possibility. If tomorrow, you wake up and there is nothing but blackness, you may want to switch to a model that says the Matrix has pulled the plug. But until then, most of us would probably agree that the model that helps us best understand our experiences is the one that says the objects of everyday life are real. This is the model we use every day and is the world as we live it.
Now inevitably, some will insist on maintaining the realist distinction between ontology and epistemology and insist that it is only the objects as 'things-in-themselves' that can really be said to exist. To this I would say that while there is a logic to this way of thinking, it is unhelpful and not actually consistent with the use of the term ‘exists’ as it is commonly applied. As explored above, the unknowable objects as ‘things-in-themselves’ cannot be perceived and we cannot infer anything about them. This is pretty much the definition of non-existence! It is more common to consider the things that exist as being the things that impact us; these being the objects of our models.
So to sum up, what I am saying is that firstly, the sceptical challenge forces us to accept that we cannot know the way the world ‘really is’ and so undermines the common sense realist claim that the objects of our experience exist, in a recognisable sense, independently of our perception of them. This implies that any conception of the world is in some sense a model; a way of giving context to our thoughts about our experiences and making sense of them. But while the sceptical challenge delivers a death blow with one hand, it resurrects with the other - it also forces us to accept that there is no meaningful single right answer of how the universe is, and so the realist goal of being able to trust a model that says the objects around us are ontologically primary can still be achieved**.
Implications
Now this is all very well, but this shift in thinking does raise further questions. For example, need there be a reality, distinct from us, that is the source of our experiences? After all, as Wittgenstein pointed out, a nothing will do as well as a something about which nothing can be said. This is the age-old solipsist question.
And furthermore, if your model is your reality, how does Anything Goes deal with disagreements over which model is best? Who is right? What if someone insists that they use a model that is clearly absurd or ‘wrong’ (in the sense that it results in false predictions)? Should you believe them?
Most interesting for me, is how this approach helps us deal with the ‘Other minds’ problem – one that is receiving increasing attention following the recent AI explosion. If the difference between a human with a point of view on the world and a zombie is not physical, then it comes down to how you see the world around you – the model you are using. But can that be right? Surely if you look at some creature, there either is something in there looking out or there isn’t?
Sadly, there is not enough space to continue exploring these issues in this blog, but they are all discussed in my book Anything Goes – A Philosophical Approach to Answering the God Question, out now on Amazon.
*It has been argued that the self-referencing within the idea that the map or model is the territory is problematic. If the territory is the sense impressions that we experience, and the map is the model we use to make sense of these impressions, then these are two different things, so it makes no sense to claim that they are the same thing. However, the point I am emphasising is that as thinking beings, we cannot avoid thinking about and giving context to our sense impressions, and so the ‘territory’ that we actually live in is a combination of both sense impressions and model.
**Clearly, there are parallels with Kant here.
Comments