Parsimony: The Simple Truth
If asked to explain why a simple explanation of some phenomenon or event is preferable to a complex explanation, many people would likely respond by saying that it is just natural to think that the simpler explanation is better. After all, the simpler explanation can do everything the more complex explanation is capable of; therefore, why not choose the simpler explanation as the better choice? However, this answer may prove to be too simplistic to count as an adequate response for some. Historically, there have been two ways to attempt to justify preferences for the simpler explanation: first, a theological or natural law grounding of justification for the principle of parsimony; and second, a methodological grounding of justification. Here, we will explore these two types of justificatory accounts of the principle of parsimony; as well, I will offer my own opinion on the principle of parsimony.
It is thought that the more parsimonious of two explanations is more likely to be the best explanation. That is, the explanation of a particular phenomenon that involves fewer steps is considered the more favourable or acceptable account of the cause of the phenomenon. In the past, the grounds for reasoning in such a way were often given by making reference to one of two metaphysical positions: natural law, or theology. The natural law argument can essentially be outlined as follows: there is no waste in nature; hence, everything in nature has a purpose. If it is true that everything in nature has a purpose, then nature would not remain the nature it is if something were removed. Therefore, nature is as simple as it can be. Given that nature is as simple as it can be (if we accept this argument), and furthermore, that there is nothing better than nature, the best possible type of thing must be the simplest thing. Hence, the best possible type of explanation must be the simplest explanation.
This argument can be criticized on any number of grounds. Even if we accept that there is nothing better than nature, what connection exists between nature and explanation, such that, we ought to be convinced that the simplest explanation is the best explanation? In other words, just because the simplest nature is the best nature, why think the simplest explanation is the best explanation? And if we are somehow able to reconcile this problem, are we justified in thinking that nature is in fact simple?
The second type of metaphysical account of the principle of parsimony is the theistic account. The argument from theology is as follows: 1) of all the possible worlds that God could have created, He chose to create thisworld; 2) this world is a simple world; therefore, 3) God prefers simplicity (p. 6). Hence, if God prefers simplicity, He must prefer the simpler of two competing explanations. This argument in support of the principle of parsimony can be challenged in a number of ways: one need only bring into doubt the existence of God to demonstrate how such an argument would fail; and the existence of God is a matter of great philosophical debate. Without the certainty of the existence of God, we lose our justification to use the theistic account of the principle of parsimony to think the simpler thing is the best thing. Moreover, even if you believe God does exist, perhaps He thinks the best universe is the simplest one, but the most complex explanation, the best.
Clearly, the above two accounts of the principle of parsimony are somewhat problematic for justifying the preference of the simplest explanation, at least for most people. This is why I have credited them with so little attention here. One might say that many people in today’s secular societies (whether rightly or wrongly) have moved past these natural or theistic accounts of justification. Yet, for whatever particular reason we dismiss each of these metaphysical positions, there still seems to exist some underlying natural desire to choose the simplest explanation between two competing accounts.
The methodological account of the principle of parsimony essentially involves observation of some event or phenomena, and then finding the simplest way of accounting for its occurrence. The methodological account simply describes the way in which we do things. That is, it seems most agreeable with human reason to choose the simplest explanation when two competing theories both offer an account of some event or phenomena, both of which seem highly plausible. The principle of parsimony seems intuitively correct.
I think a useful way to attempt to justify the methodological account of why the simplest explanation is the best explanation is to compare explanations to arguments. In an argument, the fewer premises used to establish a conclusion, the less likely you are to make an error in your reasoning. If you can reach the same conclusion with two premises rather than three – and your argument remains as good as it was previously with more premises – you have given a critic less of an opportunity to find fault with your argument. Thus, the simplest argument is the best argument because the simplest argument does not expose itself to as many potential problems as a lengthier argument attempting to reach the same conclusion.
A critic might charge this attempt to equate arguments with explanations as no better than the attempt to equate God’s preference for simplicity of worlds with God’s preference for simplicity of explanations. But I would respond to this criticism by pointing out that in order to accept my comparison between arguments and explanations, we need not assume the existence of a supernatural being (which is a daunting task, philosophically speaking) . In fact, all we need for the comparison to be credible, is to show how explanations are comparable to arguments. An example will help make the point.
Let us suppose we have a steel bar that is 10 millimetres in length. We heat the bar to 1000 degrees F, and the bar expands to 11 millimetres in length. We can easily explain the expansion of this metal bar through the form of an argument, as follows: 1) all metals expand when heated; 2) I am now heating this metal bar; therefore, 3) this metal bar will expand. This is both an argument and an explanation: it explains why this particular piece of steel expanded when it was heated, and it does so in the form of an argument. Therefore, explanations are, at least sometimes, comparable to arguments.
While one example does not prove our case beyond doubt, it does at least give us some reason to think arguments are comparable to explanations. And if we admit that it is a credible comparison, we then have reason to believe we are justified in our preference for the simpler of two explanations. For, if explanations attempt to use more reasons to explain an event or phenomenon than is necessary, the explanation has only subjected itself to the increased likelihood of refutation. Therefore, the best explanation is the simplest explanation, because the simplest explanation offers fewer possibilities for error.
Leibniz, G.W., Discourse on Metaphysics, eds., Peter G. Lucas & Leslie Grint (Oxford: Manchester University Press, 1961).