.post-body { line-height:100%; } -->

Tuesday 29 March 2016

In The Beginning

What does it actually mean for something to begin?

This is a question that has troubled philosophers for millennia and, contrary to the wibblings of some apologists, it's still a problematic question if not treated with some care.

When, for example, did this post begin? Did it begin when I hit the save button for the first time? How about when I first made an account to blogspot? Or maybe the wholesale demolition of a vibrant community, the event that drove me to seek out a new platform in the first place? Or maybe it was my birth in Croydon, or my parents' births, in Nenagh and Turloughmore respectively? There are many points I could describe it as beginning, and all would be entirely arbitrary.

Now for the care, which will require the unpacking of some terms. The first of these terms is ex nihilo, which means 'from nothing'. Some religious types will tell you that for something to begin from nothing is not possible, and they even have a kinky Latin phrase to make it sound like they've given the matter more than a cursory glance; ex nihilo nihil fit - from nothing, nothing comes - attributed to Parmenides sometime around the turn of the 5th century BCE. Of course, what they can't do is to show this in any robust fashion, they merely assert it, and will often point to the thoughts of past philosophers as their only justification. This attempt at justification has problems deeper than those with the assertion itself, for reasons I hope to elucidate shortly. In brief, we can point to phenomena that arise quite literally from nothing, but in a very special sense.

The second term is ex materia which, as the astute reader will no doubt spot, means 'from material'. This seems to gel well with our understanding of the universe, and especially the laws of thermodynamics, but even this is laden with issues when we try to apply it to the universe as a whole. More on that later. All of the 'beginnings' we've ever encountered have been of this type (with some qualification, which appears below).

The third term is ex deo, or 'from god', which I hope requires no further explanation. 

Often, this discussion is couched in terms of 'cause', which relies on concepts handed down to us from Aristotle. Aristotle, student of Plato and tutor to Alexander the Great, is considered by many to be the first true scientist. As I said in my introductory post, there's a good reason we pay attention to what Aristotle said, and that reason isn't that he was right. Indeed, he was often wrong, and sometimes 'not even wrong'. To highlight this concretely, he concluded, through reason alone, that women have fewer teeth than men. He reasoned this based on nothing more than conjecture, and wouldn't be swayed from it. He could have sorted it all out, of course, by the simple expedient of having some passing bint open her gob and counting. The problem was that Aristotle was so impressed with himself that he thought he didn't need to. Besides, observation and measurement were the work of artisans, and entirely beneath him. In many ways, Aristotle was an idiot, or at least, we'd recognise him as such today. If you look at those of his ilk around today, such as William Lane Craig, we quite rightly ridicule them (of course, it's more probable that, were Aristotle alive today, he'd be in the peanut gallery along with us). We've come a long way from thinking that the umbilicus was a source of information about anything other than the colour of the lint therein (or some of us have, at least).

That's not to say that Aristotle was wrong about everything, of course, it only represents a caveat to putting stock in such authorities. Aristotle is largely talked about today because he earned his place in the history of ideas, but we shouldn't be talking about Aristotle at all when we're talking about modern physics, for example, because he has nothing remotely of interest to say on the topic, any more than we should listen to him on the topic of feminine dental hygiene. He wasn't privy to several thousand years of gathered evidence and, as a result, much of what he said, particularly about 'cause', is total bollocks. This caveat can be expressed in the term argumentum ad verecundiam, meaning 'argument from authority', a crystal clear logical fallacy. There'll be more about fallacies later (one of the coming posts will be specifically about logical fallacies), and I'll link to some helpful resources on the topic at the bottom of this post.

This brings me to another important point concerning how philosophy is taught. As Alfred North Whitehead noted, Western philosophy is a series of footnotes to Plato. This is largely correct, but there's a danger here; it's all too easy to think, especially for one who's studied philosophy in a formal setting, that doing philosophy consists of nothing more than learning by rote what others have said. In my introductory post, I implicitly warned against this in the statement concerning the collection of ISBN numbers. Far too many of those who've studied formally have missed the point of philosophy, which I shall encapsulate as follows:

1. Ideas are disposable entities.

2. Bad ideas exist only to be disposed of.

3. Philosophy is the art of disposing of bad ideas.

Now, this all looks fairly straightforward, doesn't it? Of course, it gets a bit more complicated where the pick meets the coal-face, but not intractably so. The first two statements are fairly uncontroversial, although the first of them might seem on the face to be a problem, because it is, in itself, an idea, and should thus be considered disposable. However, when we start looking at the means by which we dispose of bad ideas (and indeed of assessing whether or not an idea is a bad one), it should become clear that there is no issue.

So, how do we go about disposing of bad ideas? Well, we begin by assessing an idea for whether it's bad. And how do we do that? Simple; we learn to ask the right kind of question about it. This is what's at the heart of philosophy; learning to ask the right kind of question.

There's a famous quotation by Nobel laureate Richard Feynman dealing with how we go about discovering a new physical law. You can find it in the Feynman Lectures on Physics, and in the book The Character of Physical Law, both of which I recommend highly. Here's a clip from the former with that quotation:


Note the latter part of that quotation:

"If it disagrees with experiment, it’s WRONG. In that simple statement is the key to science. It doesn’t make any difference how beautiful your guess is, it doesn’t matter how smart you are who made the guess, or what his name is… If it disagrees with experiment, it’s wrong. That’s all there is to it."

So, what Feynman is expressing here encapsulates nicely how we go about testing ideas in science. The important part of it is, of course, formulating the right kind of question. In this instance, the question is 'what would this idea imply about what we should observe?' Feynman expressed it succinctly as 'compute the consequences'. Once we have that question in place, we can start to devise an experiment that should manifest those consequences. If those consequences do not manifest, we should modify the idea to reflect this circumstance or, where that effect is critical to the correctness of the hypothesis, we should discard it. 

If my guess is correct, I should observe this consequence in my experiment designed to look specifically for that effect. I didn't observe that effect, so my guess is not correct.

Some effects will be critical to the idea, while others are corollary effects, the non-observance of may not fatally undermine the hypothesis, but will require the modification of the hypothesis at the very least, and the devising of new experiments to test whatever consequences are implied by this new hypothesis. You can think of a hypothesis as a machine for generating predictions, because that's what it is.

I should also note that, in the sciences, there's another, related principle employed for testing ideas, namely the null hypothesis. In a nutshell, it's a reversal of what's been expressed above, and it takes the form of a prediction regarding something that will definitely not be observed if the hypothesis is correct.

If my guess is correct, I will definitely not observe X effect. I observed X effect, so my guess is not correct. 

These two principles can be expressed neatly in propositional form, and they form the basis of what Karl Popper called 'falsification'. They take the form of the modus tollens, which will be familiar to those who've studied any philosophy or logic. Formally:

P => Q, ¬Q
∴¬P

(Proposition P implies Q, not Q, therefore not P)

And the negative form, or null hypothesis:

P=>¬Q, Q
∴¬P

(P implies not Q, Q, therefore not P)

As an aside, you may hear it said that proof does not apply to science. This isn't actually true, as the above should readily demonstrate. Properly defined, proof is a formal procedure, applicable only to axiomatically grounded systems of deductive logic, in which a route is taken from true premises (axioms) via valid rules of inference (reasoning) to a conclusion that is necessarily true. Axiomatically grounded simply means that all our axioms, or premises, are definitely true. Mathematics is an axiomatically grounded system, so it's certainly true that proof applies, but a sound syllogism (such as the modus tollens above), for example, is also axiomatically grounded (I'll have more to say on syllogisms shortly). Because the majority of scientific epistemology is inductive, we can only say that we have a degree of confidence that a conclusion is probably true. You'll hear it said that the difference between deduction and induction is that deduction reasons from the general to the specific while induction reasons from the specific to the general. This is also untrue, although it does serve as a vaguely useful first approximation. The real difference is that deduction leads to conclusions that are definitely true, while induction leads to conclusions that are probably true. So, we observe an event that falls in line with our conjecture, and each new supporting observation increases the confidence (probability) that our hypothesis is correct. This expresses David Hume's famous 'problem of induction', namely that, while confirming observations increase our confidence in the truth of a conclusion, they can never actually demonstrate its truth, because that would require that all possible observations have been made and that no falsifying observation is possible. This would, of course, require omniscience, which is self-refuting (I treat the three famous omnis here).

So, we test ideas. In science, this is usually reasonably straightforward (although not always, as some areas are horrendously difficult to test properly), but it's not always possible to test ideas this categorically, not least because not all ideas have consequences that manifest physically, or in testable ways, so we have to look at other kinds of question (I'll be revisiting this shortly). We might, for example, look at the premises employed in arriving at a conclusion and asking ourselves if there's axiomatic groundedness, i.e. whether the premises are actually true. In any case where a premise is not shown to be true, we can reasonably be suspicious of any conclusion drawn therefrom. This is the core of skepticism (there are those that insist that skepticism is the rejection of all claims, but this is little more than an attempt at poisoning the well; it's not skepticism to reject claims that have good support, only to be wary of claims that do not).

We can also look at the route from premises to conclusion to see if proper rules of inference have been applied in arriving at the conclusion and, where any route is suspect or the conclusion doesn't follow from the premises, we are again enjoined to be suspicious of the conclusion. That doesn't mean that we should reject the conclusion outright, even where an argument has been shown to be entirely invalid, but we have good reason not to accept the conclusion on the basis of that argument. 

Note that any problem with the soundness of an argument is grounds for not accepting the argument. It doesn't matter whether the issue lies in the form of the argument, in the route from premises to conclusion or in the veracity of the premises, any failure of reasoning is grounds to be skeptical of the conclusion. 

I should also note that any argument that contains anything inherently untestable or unfalsifiable fails on those grounds alone, and no further attention need be given to it.

So, now the groundwork is out of the way, let's get back to talking about beginnings:

There's a famous, ancient argument for the existence of god. It stems from Islamic theology, specifically the Ilm al-Kalam, or 'science of discourse'. Today, we know it simply as the Kalam Cosmological Argument, and its main modern proponent is William Lane 'Kalamity' Craig, whom we met earlier.

Craig has gone to book length in laying out this argument and justifying his premises and has, despite having had it eviscerated from every conceivable angle, continued to give voice to it in public debate. Here, in this post about beginnings and having laid out the beginnings of what it means to be a scientific skeptic, I wish to use it as an example of how not to do philosophy. One might think that I, as a non-philosopher, attacking an argument formulated at great length by somebody with a double doctorate in the discipline, am overstepping the mark, but what I'm really demonstrating is that collecting ISBN numbers is not doing philosophy. Some professional philosophers have described Craig's formulation of the Kalam as one of the most sophisticated theological arguments of the modern era. I describe it simply as sophistic.

So, here it is:

P1. Everything that begins to exist has a cause for its existence.
P2. The universe began to exist.
C. The universe had a cause.

Seems fairly intuitive, doesn't it? So what's the problem?

To get to that, we need to unpack the argument properly. Craig does this at book length but, luckily, in public debate, being the extremely talented Gish Galloper he is, we're lucky that he's condensed it to key hit-points. Let's go one step at a time.

P1. Everything that begins to exist has a cause for its existence.

Craig even defends this premise as being 'intuitively true'. That's problematic for a start. It's intuitively true that time runs the same for every observer except, of course, that if it were actually true, your satnav system would be a pipe dream. It's intuitively true that something cannot be in two places at once except, of course, that if it were actually true, the technology I'm employing to share my thoughts with you today wouldn't even rise to the level of fantasy. It's intuitively true that I can't walk through a wall (I've never managed it yet) except, of course, that if it were actually true we couldn't exist, because fusion in stars wouldn't occur, thus elements heavier than the beryllium (in trace amounts, along with hydrogen, helium and a small amount of lithium) that theories predict was synthesised in the Big Bang would be impossible. This fallacy is the 'appeal to intuition'. Some, notably David Chalmers, have attempted to defend this as not a fallacy, but it's an irrelevant appeal and can be dismissed on that basis alone. Intuition is helpful in formulating hypotheses but, as support for an argument, it's arse-gravy.

How about the claim itself? Well, two of the examples of things that are intuitively true given above raise some problems for its factual correctness, not least because we can point to no cause with regard to quantum tunnelling, which is responsible for both fusion in stars and the device I'm holding forth on to deliver this little missive, the implication of which is that it's perfectly possible for me to walk through a wall, though I'm quite a large fella, so it's unlikely to happen in the life of the universe, but it's also purely a matter of scale. The laws that govern these processes also imply something else. The underlying principle here is Heisenberg's Uncertainty Principle. In a near-future post, I'll be doing some cosmology stuff, and I'll cover this at some length but, suffice it to say for the moment that a consequence of that principle is something known in the jargon as 'pair production' which, in a nutshell, involves pairs of particles (properly a particle and an anti-particle), which borrow a little energy from spacetime, pop into existence, move apart, and then come back together and annihilate (because their energies cancel out, or 'sum to zero'). This effect has been measured by virtue of a very clever thought experiment, the physical experiment for which was realised some years ago, and the behaviour matched the predictions with ridiculous precision (this has been compared with being asked how far you were from the moon and the response being 'from the top of my head or my chin?'). We can point to no physical cause for this behaviour (the question of whether it can be described as ex nihilo is a separate question, but it doesn't favour the argument; I'll cover this in the next post). Now, while we can't categorically state that those events have no cause, they give us some pause. In short, the premise commits, at best, the fallacy of bare assertion and, at worst, is counterfactual.

What about the second premise?

P2. The universe began to exist.

Here Craig brings in arguments from all sorts of places, not least Big Bang cosmology and mathematics. He asserts that 'an actual infinity is impossible', with no justification whatsoever. I'm no mathematician, so I won't attempt to address the mathematical claim itself, but there's a beautiful analysis of this by Wes Morriston of CU Boulder, so I'll simply link to it (pdf).

How about the cosmology? I'm going to be treating this extensively in the next post, which will deal with specifically where our current understanding is, what our best options for future progress are, and precisely why the claim simply doesn't stack up.

Craig is relying on an extremely rudimentary understanding of Big Bang cosmology. He asserts that Big Bang cosmology involves the universe having a beginning. To properly treat this assertion, we need to clarify precisely what we mean by 'universe'. More precisely, we need first to differentiate between the observable universe and the universe at large (many use the term 'multiverse' to deal with all such concepts, including those concepts that deal with only our cosmic expanse, but I'm not fond of the term). What Big Bang cosmology deals with is our local cosmic expanse only and, in fact, it doesn't even deal with the beginning of that in any robust fashion. I want to leave the majority of the meat of this topic to my next post, as it will address it far more completely and deserves to be treated separately. For the time being, we only need note that, without defining the term 'universe', and only citing broad areas of cosmology as purportedly supporting the claim that the universe at large began to exist, we can reasonably reject this claim as unsupported. 

I should also add that, while Craig's argument relies on the impossibility of an actual infinite, the singularity theorem, upon which his assertion that time began at the Big Bang is predicted, constitutes an actual infinite, which is sufficient to defeat that support on its own terms.

Craig also cites Alex Vilenkin, Arvin Borde and Alan Guth, and their paper Inflationary Spacetimes are not past-complete from 2001 as allegedly asserting that the universe must have had a past boundary, but this assertion is simply counterfactual. BVG theorem deals specifically with inflationary cosmologies, and only states that new physics is required to explain inflation. You'll note that this is a very different claim. Again, this will be treated in more detail in my next post, but the take-away here is that a) BVG theorem does NOT state that the universe must have had a beginning and b) not all cosmologies are inflationary.

Suffice it to say that, once the alleged support for this second premise is analysed in any detail, it's shown to commit yet another fallacy of blind assertion. It's reasonable to be skeptical of the conclusion on this basis.

So, now that the premises have been treated, what about the route from conclusion to premises?* Is it even valid?

Sadly, no, there isn't even any consolation for Craig here. He's identified a principle that appears to prevail within the cosmos, then applied it to the cosmos itself. This commits the informal fallacy known as the 'fallacy of composition', which is committed in any argument reliant on the idea that a property of the parts is a property of the whole. This is easily exposed by telling you that the second Interview With Matter album Remotion (produced by yours truly, incidentally; album three, as yet untitled, is running on rails at the moment, and should be released shortly, time permitting), is composed of tracks that are all less than ten minutes long, yet the album itself is just a smidge over an hour long. This is because the property 'length of time' is not expansive. Whether the apposite principles that apply within the universe are expansive is not known but, given the current state of cosmology, there's good reason not to rule out the idea that they may not be.

There's another fallacy in there between the first two premises in the term 'begin/began to exist'. Here, Craig is relying on two very distinct definitions of beginning to exist, namely ex materia, and ex nihilo. Thus he commits a fallacy of equivocation.

Recall that any error in reasoning, formal (having to do with the structure of the argument) or informal (having to do whether proper rules of inference have been applied other than in the structure, which mostly involves ascertaining whether or not the premises are true (is it really the case that P=>Q?)), is sufficient to cast doubt on the truth of the conclusion. In other words, we can consider the argument 'unproven'. As a system of deductive logic, this argument is not axiomatically grounded. In fact, it's only the fallacies holding it together.

Watch out for a reasonably comprehensive treatment of cosmology in the next few days, including a potted history of physics and brief material on relativity and quantum mechanics.

Thanks for reading.

A Taxonomy of Logical Fallacies

* Edit: This was a typo, and should have read 'the route from premises to conclusion. I thought about editing it out, but then I realised that the typo is perfectly apposite, because Craig's entire argument is one whose circumference is related to its radius by a multiple of π. Indeed, his entire career is not only circular but, given the breadth of his influence on the credulous, he might be described by Swiss astronomer Fritz Zwicky as 'spherical'.

2 comments:

  1. Excellent post but the correlation you make between discarding bad ideas in philosophy and making testable hypotheses in science is rather loose. First of all the notion of good and bad does not exist in science. There are only testable and non testable hypotheses. Whether they are good or bad is not a matter for science. Knowledge in itself is neither good or bad although how it is applied may have moral consequences. But that is a matter for philosophy not science. A bad idea and an untestable hypothesis or conversely a good idea and a testable hypothesis are not the same. I know you were trying to make a correlation but it is too loose as I said. The rest of the post is logically sound however

    ReplyDelete
  2. In which case I'm spectacularly pleased to have drawn no such correlation.

    ReplyDelete

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