Sometimes the attitudes of philosophers towards science baffle me. A good example is the article Defending Humanistic Reasoning by Paul Giladi, Alexis Papazoglou and Giuseppina D’Oro, recently in Philosophy Now.
Why did Caesar cross the Rubicon? Because of his leg movements? Or because he wanted to assert his authority in Rome over his rivals? When we seek to interpret the actions of Caesar and Socrates, and ask what reasons they had for acting so, we do not usually want their actions to be explained as we might explain the rise of the tides or the motion of the planets; that is, as physical events dictated by natural laws. […]
The two varieties of explanation appear to compete, because both give rival explanations of the same action. But there is a way in which scientific explanations such as bodily movements and humanistic explanations such as motives and goals need not compete.
This treats “science” as though it stops where humans start. Science can deal with the world as it was before humans evolved, but at some point humans came along and — for unstated reasons — humans are outside the scope of science. This might be how some philosophers see things but the notion is totally alien to science. Humans are natural products of a natural world, and are just as much a part of what science can study as anything else.
Yes of course we want explanations of Caesar’s acts in terms of “motivations and goals” rather than physiology alone — is there even one person anywhere who would deny that? But nothing about human motivations and goals is outside the proper domain of science.
If we ask about a gazelle running from a lion, or about social interactions within a chimpanzee troop, such as two beta males forming an alliance to depose the alpha male, we again want explanations in terms of goals, thoughts and motivations, and yet no-one denies that studying animal behaviour is “scientific”.
Explanations in such terms are not “unscientific”, they are simply from one of the explanatory frameworks that is most appropriate to adopt when dealing with biology and with self-aware animals. Even Giladi et al accept that biology is a science, and yet when it comes to humans they imply that science suddenly stops and can no longer function.
Western philosophy has adopted increasingly naturalistic views. In current Anglo-American philosophy, the norm is to assume a reductive form of this naturalism which claims that everything can be explained just in physical terms. This position is usually called physicalism or materialism. According to this version of scientific naturalism, the image of the world provided by the physical sciences (basically, physics, chemistry, and biology) is all the world there is.
There are several elements here to clarify. First, “naturalistic” is not to be contrasted with “humanistic”. Humans are entirely natural and biology is a science (as is anthropology). Naturalism is instead contrasted with supernaturalism, such as the notion that humans possess a dualistic soul. The idea that the “world provided by the physical sciences is all the world there is” denies supernaturalism, but does not deny that humans exist and have thought, goals and motivations.
Second, let’s unpack the idea that: “everything can be explained just in physical terms”. The underlying doctrine of “supervenience physicalism” says that higher-level phenomena such as mammals supervene on lower-level phenomena such as atoms and molecules. And one can completely describe a system in low-level terms. That means that, if one exactly replicated all low-level aspects of a system (a complete and exhaustive listing of every atom and molecule and its behaviour), then the higher-level phenomena would also be entailed and replicated.
But — crucially — a complete description is not the only explanation, and it is likely not the most useful nor the most interesting explanation. Complete explanations (ones that are sufficient to exactly replicate the entire system) are simply too unwieldy for nearly all purposes. Higher-level explanations are reduced-information explanations that focus on salient aspects and so are usually more useful and interesting.
The different explanations do not compete, they are complementary. Nothing prevents there being more than one explanation or description of the same thing! Thus a “complete” explanation does not push out all other explanations, it sits alongside them.
That is the case even within physics. For example, when considering a simple system of a gas inside a container, one could (in principle) list every individual gas molecule and its trajectory — but one can’t in practice, it’s too hard — or one can instead provide explanations in terms of higher-level concepts such as “pressure”, “temperature”, “entropy” and the laws of thermodynamics, which apply only to the system as a whole, and cannot be applied to individual particles.
The doctrine of supervenience physicalism requires that all the explanations at the different levels must be consistent (they cannot conflict with each other), but in no sense does naturalism or physicalism disqualify or push out the higher-level and humanistic explanations.
Giladi et al continue:
One reason physicalist forms of scientific naturalism have become so widely accepted is that many philosophers tend to find it difficult to make room for complex phenomena such as consciousness within the world presented to us by the natural sciences. Because of this, some physicalist philosophers reduce complex psychological phenomena down to their component material parts –- things such neural mechanisms –- or even to the very components of matter itself.
One has to be careful with that word “reduce”. Many philosophers use it to mean a fairly strong notion of eliminative reductionism. In this thesis, if one understood the lower-level aspects of a system at the level of physics, then one could, by a few lines of algebra, calculate the higher-level properties such as desires and motivations. The higher-level properties could then be eliminated, and the proper level of discourse would be only at the lower-level of physics.
It is unclear to me whether any philosophers actually hold to this idea, as opposed to it being a straw-man attributed to others, but I’m fairly sure that no physicist or scientist holds to it.
Yes, one could, in principle, replicate a system by replicating all low-level aspects, but this is nearly always impossible in practice, even within physics. Further, the thought experiment of replicating, say, a jaguar — such as with an atom-by-atom, Star-Trek-transporter device — is not at all the same as calculating the behaviour of a jaguar from the low-level description plus a set of simple laws and some algebra. The dream of finding laws to enable us to do the latter is sometimes attributed to physicists by philosophers, but it is not how the world works and not how physicists think.
So after having, to my mind, falsely created a division between physicalism and humanism, how do Giladi et al then attempt to reconcile them? They invoke a philosophical tradition that seeks to “defend the independence of the human sciences from the natural sciences” (again, what about humans is unnatural?) by starting with the claim that “all knowledge rests on presuppositions”. The investigation of the “background assumptions of science”, they say, “cannot itself be carried out using the methods of natural science”.
This claim that science rests of assumptions that cannot be tested by science is, I think, wrong. Science is a web of ideas that is aimed at modelling empirical reality. The scientific method tests the set of ideas against reality, and, where it is found wanting, then adjusts the model to do better, before testing once again.
In Neurath’s analogy of science’s web of ideas as the set of planks of a raft afloat on the sea, any idea, any plank, can be examined and replaced if that means that the raft floats better. In this view there are no “presuppositions” of science; every idea can be tested.
The way to test a “presupposition” is the same as the way to test any other idea: simply replace the idea with its negation, make consequent adjustments elsewhere in the web, and then again test the web of ideas against reality. If the web containing the negation does worse (however much you try to make compensatory adjustments) then the negation is rejected. Thus the idea has now been tested, not merely assumed, and is thence adopted because it works.
If, alternatively, replacing the idea with its negation makes no difference in how well the overall web matches reality, then that means that it is not a presupposition of science, since (ipso facto) one needn’t suppose it. Either way, the idea that there are ideas that science must assume but cannot test is erroneous.
What actual ideas do Giladi et al then point to?
By carefully unpacking the background assumptions of the different forms of inquiry we can lay bare the most important difference between the natural sciences and the human sciences. For example, natural science presupposes a uniform universe governed by universal laws; on the other hand, a historian does not approach history in such a way. Another difference between the natural sciences and the human sciences amounts to the natural sciences aspiring to grasp the universal, the general, whereas a human science such as history aspires to make sense of the particular, of unique events.
Well sorry, but this is just wrong. Yes, physical scientists are indeed interested in uncovering the universal laws of a uniform nature where that is appropriate, but they are just as interested in the chance, contingent events that are just as important in leading to actual outcomes.
As just two examples from the natural sciences, first, our moon is believed to have formed when, early in the history of our Solar System, a Mars-sized proto-planet collided with the proto-Earth, throwing up a cloud of debris that coalesced into our moon. This was a unique and particular chance event of great importance to the history of the Earth. A second example is the collision with Earth of the Chicxulub meteor that is believed to have wiped out the dinosaurs and radically changed the history of subsequent life on Earth. Again, it was a particular and chance event.
To say that the natural sciences are uninterested in such historical contingency, or that they cannot deal with such singular events, or that the natural sciences are only interested in uniformities in nature and universal laws, is just wrong.
Giladi et al want to divide different areas of knowledge up (natural sciences versus humanities) according to the different presuppositions that each area must adopt. I think this is wrong-headed. The different areas of knowledge are all part of the same whole. But there is no conflict, no competition between the different areas. The different types of enquiry and types of explanation must all be mutually consistent and sit alongside each other. They are different simply because they focus on different features of interest, different aspects of the whole.