2. The scientist and scientific method

2.1 The arch of knowledge and its "empiriometric" core

The chapters to follow will argue that the results of modern science not only in no way conflict with the central claims of Aristotelian philosophy of nature, but in some respects even vindicate those claims. The present chapter argues that the very methods of modern science vindicate those claims - and in an even more decisive way. For while the results of science might change (as currently accepted theories are abandoned and replaced by new ones), at least the core elements of scientific method will not.

Naturally, this raises the question of exactly what the scientific method is - something which has, of course, been a matter of great controversy in modern philosophy of science. I am not suggesting that that controversy is susceptible of easy resolution, nor will I try to resolve it here. My point is that although various details of scientific method are matters of dispute, there are some basic assumptions that all sides to the debate tend to agree on, and it is these which presuppose an essentially Aristotelian conception of nature (whether most philosophers of science realize this or not).

To see what these core assumptions are it will be useful to consider the history of what philosopher of science David Oldroyd (1989) has called the notion of "the arch of knowledge," and in particular the ways that notion was developed by early modern thinkers like Bacon, Galileo, Descartes, Newton, Boyle and others, and modified by more recent philosophers of science. As Oldroyd notes, the basic idea of the "arch" in fact goes back at least to Plato. But it is the construal of the "arch" associated with the fathers of the scientific revolution that has in modern times come to define what constitutes "science."

Bacon famously put heavy emphasis on observation as the evidential foundation of science. That was not by itself in any way novel. The thesis that nothing is in the intellect that was not first in the senses had been a commonplace of medieval Aristotelianism. Where Bacon took himself to be departing from his Aristotelian predecessors was in his application of this principle. For Bacon, the Aristotelians were too uncritical in their appeal to empirical evidence, in two respects. First, they were in his view too quick to draw general conclusions from that evidence. What was needed was patience, and in particular the slow and painstaking assembly of as many observations as possible of the phenomenon under investigation, under as wide a variety of circumstances as possible. Only after this was done could one be confident of the general conclusions one might draw about the nature of that phenomenon.

Second, in Bacon's view the Aristotelians had an insufficient appreciation of the biases that can infect individual observations. These biases were enshrined in what Bacon's Novum Organum characterizes as the "Idols of the Mind," of which there are four. The first are the Idols of the Tribe, by which Bacon means the biases inherent in human nature, such as our tendency to take it for granted that things really are as they appear to the senses. The second are the Idols of the Cave, or the biases that derive from a person's individual temperament, education, experiences, social setting, and so forth. The third are the Idols of the Marketplace, our tendency uncritically to suppose that the way language carves the world up corresponds to the way things really are objectively. The fourth are the Idols of the Theatre, our tendency to suppose that reality must conform to some philosophical or scientific theory to which we are especially attached. This can lead us to think that we are reading the truth of the theory off from what we observe, when in fact we are reading it into what we observe.

Once we have made a sufficient number of careful observations and corrected as far as we can for biases of these sorts, then in Bacon's view we can begin to reason inductively to general conclusions. The canons of inductive reasoning emphasized by Bacon were the sort refined and expanded upon by John Stuart Mill in the nineteenth century, in the famous "Mill's Methods" of establishing causal relationships between phenomena. In Bacon's view this kind of reasoning had been insufficiently appreciated by the Aristotelians, who, as he saw it, were impatient to reason deductively, from the general conclusions to which they had too hastily arrived to conclusions about what the empirical world must be like. Of course, Bacon was not opposed to such deductive reasoning, but his emphasis was on what he took to be the long neglected inductive aspect of science. With this we have one of classic expressions of the idea of the "arch of knowledge." At the foot of the left side of the arch is the body of empirical evidence painstakingly assembled. The left leg of the arch is constituted by the inductive reasoning that takes us from this evidence up to general conclusions, which form the apex of the arch. The right leg of the arch is constituted by deductive reasoning from these general conclusions, down to the specific empirical predictions that lie at the foot of the right leg.

Bacon represents, in effect, the empiricist side of the scientific revolution (to construe "empiricism" very broadly - a Baconian need not be committed to the desiccated conception of experience associated with the modern British empiricists). The mathematization of nature championed by Galileo, Descartes, and Newton represents the rationalist side (to construe "rationalism" very broadly too - naturally I am not attributing to Galileo or Newton all the epistemological commitments of Descartes and other continental rationalists). Galileo's The Assayer famously declared that mathematics is the language in which the book of nature is written, and attributed to matter only quantifiable primary qualities, relegating irreducibly qualitative secondary qualities to the mind. Galileo modeled the physical world in terms of mathematical abstractions like frictionless planes. Descartes reduced matter to the geometrical attribute of extension. Newton brought to fruition Galileo's and Descartes' project of tracing the behavior of all bodies down to the operation of a set of fundamental mathematically formulated laws.

Descartes qua rationalist tended to emphasize deductive reasoning down from an abstract mathematical model of nature, in contrast to Bacon's emphasis on inductive reasoning up from painstaking empirical observation. However, the greatest impact on modern scientific method of the mathematization of nature had to do neither with the legs nor the feet of the "arch of knowledge," but rather with the character of the general description of the world that would form its apex. The idea was that, at least ideally (if not always in practice, especially in special sciences far removed from physics), the general theories toward which scientific inquiry worked ought to be formulated in terms of quantifiable properties and mathematically expressible laws.

A further component of the early modern conception of the apex of the arch was the mechanical philosophy's program of analyzing observable bodies and their behavior in terms of unobservable particles in motion. There was, initially, some disagreement on how this program ought to be fleshed out. Descartes advocated a plenum theory on which matter is infinitely divisible - so that there is no fundamental level of particles - and on which there is no void or empty space between particles. Gassendi and Hobbes advocated the ancient atomist view that there is a fundamental level of particles which are indivisible in principle, and empty space through which the particles pass. What won out, eventually, was the corpuscularian position of Boyle and Locke, which also affirmed a level of fundamental particles passing through void space, but held that they are merely undivided in fact rather than indivisible in principle.

What all of these variations on the mechanical philosophy agreed on, however, was a commitment to an essentially quantitative and mathematical conception of the particles and their properties and behavior. Secondary qualities like color, odor, sound, taste, heat, and cold were relegated to the mind. Final causes, the actualization of potentialities, substantial forms and the like were denied or at least ignored. For purposes of scientific description, quantifiable primary qualities alone were affirmed of the particles and of the objects they composed, and their changes were analyzed in terms of mathematically describable movements through space.

As physics progressed, however, commitment even to the corpuscles of Boyle and Locke disappeared. In place of discrete particles changing their positions in space over time, relativity theory speaks of four-dimensional space-time worms extending through a static block universe, and quantum mechanics speaks of wave functions. (More on these notions in later chapters.) What has remained, however, is commitment to an essentially quantitative and mathematical mode of describing nature.

In the Opticks, Newton gave classical expression to the settled early modern conception of the "arch of knowledge" with his method of analysis and composition. Analysis constitutes the left leg of the arch. Through observation and experiment, we work from compounds to the ingredients...