15 May 2014

The death of metaphysics

I've had a difficult time understanding the point of metaphysics over the years, despite quite a few armchair philosophers (mostly theists) telling me metaphysics are really important. It doesn't help that the very term "metaphysics" is ambiguously defined, and its usage varies depending on which dead philosopher you read and which living philosopher (armchair or otherwise) you talk to. Reading through the Stanford Encyclopedia of Philosophy's article on metaphysics just adds to the confusion; the definition of metaphysics has changed over time (and in my experience, predictably, plenty of people still operate on the more archaic usages of the term), and the 'metaphysical problems' highlighted in the article struck me as either being scientific problems or semantic conceptual problems – I couldn't see any particular reason to think that the word 'metaphysics' is useful.

What is 'real', anyway?

I've mentioned recently that I'm re-reading Philosophy of the Flesh, a fantastic book by cognitive linguist George Lakoff in which he advocates a philosophy he calls "embodied realism". In the book, he describes metaphysics as follows:
Metaphysics, for example, is a fancy name for our concern with what is real. Traditional metaphysics asks questions that sound esoteric: What is essence? What is causation? What is time? What is the self?
I've mentioned many times before that I subscribe to what Stephen Hawking and Leonard Mlodinov describe in The Grand Design as 'model-dependent realism'. This essentially says that we do not have unfettered access to 'absolute truth' or complete knowledge of what is literally real. All we can do, then, is construct models with varying degrees of predictive reliability. Fussing over what is 'real' is meaningless – all that's relevant to our epistemological horizon is how well our models comport with observation.

And that, essentially, is precisely what science is: the attempt to create reliable models of reality. We do this by constructing predictive hypotheses. For example, the recent discovery of gravitational waves in the Cosmic Microwave Background confirmed a long-standing prediction of Inflationary Cosmology, and the CMB itself – as well as certain asymmetries, called anisotropy – was a prediction of the Big Bang Theory. The phylogenetic tree in evolutionary biology allows us to predict precisely where we will find certain fossils, and how old we will find them to be. And in every case, the predictions could, in principle, be wrong. This is known as falsifiability. The gravitational waves might not have been there, or we could find fossils that do not conform to the predictions of the phylogenetic tree. When a prediction fails, we have two options: revise the model, or discard it completely.

Indeed, falsification is precisely how we are able to weed out unreliable models of reality. Prior to Einstein's formulation of Special Relativity, physicists had proposed the existence of an invisible substance permeating all of space through which light travel – they called this the 'luminiferous aether'. But predictions based on the theory did not hold up, and when predictions based on Special Relativity proved accurate, the luminferous aether (along with any conception of 'fixed' time) was relegated to the dustbin of failed sciences.

Science has proved itself a remarkable means by which to model reality. It has vastly increased our understanding of the universe in a remarkably short span of time. Just a few hundred years ago, this post would have been written on parchment with a quill and ink; now, it's written and read on devices that make use of quantum mechanics (not to be confused with quantum computing). If you're reading it on your phone, your GPS app is only made possible because of General Relativity. We've eradicated countless diseases and doubled our lifespans. That speaks volumes to the success of science.

What about the 'limits' of science?

Whenever I've brought these points up in discussions with (armchair) philosophers – not coincidentally, I think, almost always theists – the response is something like this:
Science is great, but it can only tell us about the natural world; metaphysics gets to the very core of what is real – the nature of change, causation, time, existence, consciousness, and much more.
However, this type of response immediately strikes me as a tautology. I'm not the first to point out that defining science as 'the study of the natural world' and then the natural world as 'that which can be studied by science' is circular. I think it's more accurate to point out that we have no reason to believe that this dichotomy between 'natural' and 'supernatural' exists in the first place – indeed the very presumption of a 'supernatural' reality distinct from 'natural' reality needs to be defended and established before we're obligated to take it seriously. If our concern is with what is 'real', then we ought to be able to construct some reliable model of supernatural things. Not surprisingly, in my experience the discussions inevitably return to the same tautology: constructing reliable models of things (indeed the very concept of reliability) is only relevant to science, which studies the natural world... which is that which can be studied by science.

I think there's a much simpler way of approaching these kinds of issues. I'm not convinced that the term 'metaphysics' is useful or even meaningful due to its ambiguity, and because the very meaning of the words used to construct metaphysical propositions are, in fact, derived from our empirical experience.

The semantic structure of metaphysics is derived from empirical models

It's been argued to me many times that empiricism (the idea that knowledge can only or primarily be derived from our sensory experience), with which science is undoubtedly deeply intertwined, is just one of many possible philosophical assumptions about what is 'real'. But again, according to model-dependent realism, asking what is 'real' is meaningless; our epistemological boundary is limited by models that comport, with varying degrees of reliability, to observation.

The reasoning behind the assumption of model-dependent realism should be easy to comprehend. Often, what we think is real turns out to be something else. Predictions often do not hold true; we're subject to false memories and cognitive biases; and our own cognitive models of reality often fail us. This demonstrates that we do not have access to Absolute Truth, or complete knowledge of what is literally real. It's inescapable that constructing models of reality with varying degrees of reliability is what we as humans do.

But those who advocate metaphysics seem convinced that there is another way to answer questions about what is real – not through modeling, prediction, and observation, but through our human capacity for reasoning. Through the principles of logic, we can construct semantic frameworks that lead us to inescapable conclusions about the fundamental nature of reality and being.

The problem with this type of thought is that the very semantic frameworks that are used to construct metaphysical propositions and derive their conclusions are fully dependent on both our physical embodiment and our empirical experience.

In discussing the metaphysics of logic, for example, I was presented with this argument:
There is no reason to think that the language promulgated in the laws of logic, or explanation for that matter, are incoherent beyond the bounds of this universe. The laws of logic really only rely on the concept of identity, and anything that exists has an identity--no matter whether it's physical, non-physical, located in or out of space-time etc. If it exists, it has an identity, and therefore the laws of logic are surely coherent no matter where they are applied.
This is a metaphysical proposition, and an important one for theists. If the laws of logic do not hold beyond the universe, then God, assuming one even exists, is incomprehensible – surely a damning proposition for theism.

But the very language here only has meaning because of its structure in our empirical, embodied experience. Take the term 'beyond the universe'. It might sound profound, but that is only so because its precise meaning is ambiguous. We tend to visualize the universe like a container, with stars and galaxies and everything else contained within it, and some mysterious realm existing 'outside' or 'beyond' it – possibly the realm of the supernatural.

This abstraction is intuitive because it comports with our cognitive models of reality; we see containers with things in them and space around them. There is a clearly delineated 'within' and 'beyond', or 'inside' and 'outside'. But these terms derive their meaning by their description of points in space-time, and space-time is a property of the universe. Asking what is 'beyond the universe' is meaningless, like asking what came 'before time' or what is 'South of the South Pole'. Notice that the proposition similarly uses the phrase 'in or out of space-time' – but it is meaningless to use words that describe spatiotemporal relationships without the context of spatiotemporality. The same can be shown to be true of words like 'change' and 'cause'.

This demonstrates that these sorts of metaphysical propositions are nonsensical. What metaphysicians attempt to do is construct a semantic framework and abstract conclusions from it; that is, if we assume a certain meaning to these words, the outcome that follows must be x. Or, as Lakoff explains,
... philosophers engaged in making metaphysical claims are choosing from the cognitive unconscious a set of existing metaphors that have a consistent ontology. That is, using unconscious everyday metaphors, philosophers seek to make a noncontradictory choice of conceptual entities defined by those metaphors; they then take those entities to be real and systematically draw out the implications of that choice in an attempt to account for our experience using that metaphysics.
Metaphysics in philosophy is, of course, supposed to characterize what is real – literally real. The irony is that such a conception of the real depends upon unconscious metaphors.
We can thus avoid the confusion of these metaphysical propositions simply by pointing out that there is no reason to think that the words used to construct such propositions are meaningful outside of the empirical context from which their meaning is abstracted in the first place.

Clarifying 'empiricism'

When I use the term 'empirical', I'm using it in an everyday context that describes our sensory experience and our ability to construct testable models of reality. Whether or not empiricism in the broad philosophical sense of 'only empirical statements can be true' is valid is not really a concern of mine. (It's often pointed out that such a statement can't be tested against itself.) I don't need to assume that only empirical statements can be true in order to reject metaphysical propositions or acknowledge the integral nature of our sensory, embodied experience not only in shaping models of reality but also in constructing the very metaphors we use to describe those models.

Years of discussions with others about metaphysics had only left me confused, and I've begun to understand why. While in discussions with theists I would be on the receiving end of comments suggesting that I simply did not understand these sophisticated arguments, a more critical investigation has led me to conclude that my difficulty deciphering the meaning and relevance of metaphysical propositions is not a failing of my own, but a problem deeply intertwined with metaphysics as an intellectual endeavor:
  • Metaphysicians attempt to describe reality not by constructing testable models, but by abstracting the outcomes of semantic frameworks.
  • Those frameworks are themselves composed of words and metaphors derived from our empirical experience, and there's no reason to think that they hold true when stripped of that context.
  • Because of this, there is no reason to think that a meaningful distinction exists between 'natural' and 'supernatural', nor that there are extant realities which transcend our ability to model reality with falsifiable predictions.
  • The precise meaning and usage of the term 'metaphysics', as well as what constitutes a 'metaphysical problem', is ambiguously defined. 
  • No metaphysician has ever been able to demonstrate that their semantic frameworks do, in fact, correspond with what is literally real – precisely because of the fact that the semantic structures of their metaphysical propositions are derived from our model-dependent, empirical experience.

These factors combine to show, in my view, that metaphysics is dead. It cannot meaningfully contribute to our understanding of... well, anything, and it exists only as an obsolete tool usurped by the immense success of modern science.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.