§50. Hypotheses Non Fingo

Aristotle's idea of science assumes a world of essences arranged in a hierar­chical structure that can be exhaustively explicated by syllogisms—a com­pletely intelligible teleological order.

By 1350, this idea of science, received enthusiastically in Europe a cen­tury earlier, is increasingly viewed with skepticism. Nominalist epistemology after Ockham put severe but realistic limits on where and how this model of science applied. But demonstrable necessary truth still seemed essential to anything that merited the name scientia. We forget how resilient this tra­dition was, now so thoroughly discredited. Certainty, necessity, and uni­versality were still connotations of scientia in Bacon, Descartes, Hobbes, Locke, even Hume. Gassendi, Boyle, and Locke pioneer a new concept of experimental natural philosophy, though it is but consolation for a scientia that is beyond us. Ockham unknowingly nudged natural philosophy in an experimental direction, being implacably skeptical of logical links among the concepts of natural philosophy. If everything in nature could have been otherwise, as Augustine said and the authorities recently confirmed (1277), then no high road to a science of nature passes through concepts and logic without descent into contingent experience.

Ideas had long been urged about method in natural philosophy not from proven success, which was scarce, but because they seemed to align with what was believed about mind and nature.

Descartes was a student of the Jesuits at La Fleche, 1607-14. His studies in­cluded natural philosophy, for which he read the then up-to-date Eustachius of St. Paul, Summa Philosophiae Quadripartita (1609). Later, Descartes remembered this work as “the best book ever written on this matter,” that is, natural philosophy, a risible judgment considering Suarez is Eustachius’s contemporary. But Eustachius suited Descartes’s impatience with detail and scholastic subtlety, all of which he abbreviates or suppresses. The young Descartes was not badly served by the philosophy he learned at school, which was probably better instruction than in the universities of the day. Philosophy in France was no longer reserved to the university arts faculty, and students could follow the subject in the many colleges founded since the sixteenth century in the nation's administrative and ecclesiastical centers. These courses studied philosophy in greater detail than the old arts-faculty courses and for a longer period, and the subject was taught without regard to preparing students for the higher faculties.²⁰¹

The choicest colleges, including Descartes's at La Fleche, were controlled by the Jesuits. Practically all the professors of philosophy followed the same pattern developed at the University of Paris in the sixteenth century.

The format of these textbooks had been establish by Philip Melanchthon in the previous century. Instead of following Aristotle's order of argument, he followed a list of loci selected and ordered for didactic clarity. The purpose was to demonstrate divine providence in the physical world. Such books drilled students into getting the right answer to given questions without in­viting them to consider the principles or goals of natural philosophy, which was the comprehensive study of God's providence in the physical world. The most “modern” quality of these textbooks is their preoccupation with method. They promise methodical order, a methodical presentation, though the method is pedagogical, to facilitate retention, not a method of discovery or proof.²⁰³

There is nothing to discover, and the expectation of discovery plays no part of their philosophy, nor does experience, despite avowed commitment to its value. A scholar of this material observes that these authors “vigorously re­pudiate uncritical reliance on written authority—that of Aristotle or anyone else. They affirm over and over again that the primary and proper philosoph­ical criteria of truth are reason and experience” In practice, these professions mean little. Experience “seems to mean simply the fund of information which an ordinary, educated man picked up by living in the world.” Putative observations are typically lifted from earlier books rather than from the author’s experience.

Descartes wants experiments in natural philosophy as much as Bacon does. It is only a question of what requires experimenting and what is avail­able a priori. One need of Cartesian science is information about the phe­nomena that actually exist for explanation. No geometrical cogitation will conclude that thunder is heard or meteors seen. Like Galileo before and Newton later, Descartes expects experiments to decide which among pos­sible mechanisms actually produce the phenomena. That is the role Bacon assigns to crucial instances, and Descartes concurs in their value. “In order to be able to tell us which of these two causes is the true one, it is necessary to consider other experiments which cannot be brought into accord with both.” Not just the decision among alternatives but even the single hypothesis has to be refined and controlled by experience. “I use that kind of philosophizing in which there is no principle that is not mathematical and evident, and its conclusions confirmed by true experiments.”²⁰⁵

Descartes’s official position on hypothesis is to permit it when it is explained by non-hypothetical self-evident first principles. However, such explanations proved rare. It seemed enough that the hypotheses of science are morally certain. Moral certainty did not mean high probability; it was genuine certainty in its sphere, which is conduct. With moral certainty no reason exists for doubt apart from God’s absolute power. “Some things are considered as morally certain, that is, as having sufficient certainty for appli­cation to ordinary life, even though they may be uncertain in relation to the absolute power of God.”²⁰⁶

Later Cartesians tended to interpret scientific certainty in these terms.

The stimulus to the idea comes from experience, but the content is the mind's creation. “Our mind is so constituted by nature that general propos­itions are formed out of knowledge of particulars” Descartes wrote, recapping Aristotle's epagoge without making it less dogmatic. We have an inexplicable power to apprehend the universal and necessary as they flash through our experience. “This mental vision extends both to all those simple natures, and to the knowledge of the necessary connections between them.” Exactly what Ockham forbade. Cartesian experiments are useful, even necessary, for judging scientific propositions, but are no good for discovering them. Descartes explains to Mersenne, “As for the more particular experiments, it is impossible not to make many that are superfluous, and even false, if one does not know the truth of things before making them.” That is what Leibniz will criticize in Boyle. Experiments confirm conclusions independently de­rived. “There are no paths to certain knowledge of the truth accessible to men save manifest intuition and necessary deduction.” Descartes gives an ex­ample, using a favorite problem in natural philosophy:

He who reflects that there can be nothing to know in the magnet which does not consist of certain simple natures will have no doubt how to pro­ceed.

Newton makes an entirely different use of experiments. Their evidence is the primary warrant for scientific claims, and nothing that experiments teach is self-evident. That is why we go to experiments to learn, but it also means that experiment affords no insight into essences. In his Mathematical Principles of Natural Philosophy (1687), Newton lays down an austere em­piricism, affirming a principle from Democritus and recently revived by Gassendi. “What the real substance of anything is we know not. In bodies we see only their figures and colors, we hear only their sounds, we touch only their outward surfaces, we smell only their smells, and taste the savor; but their inward substances are not to be known either by one sense, or by any reflex act of our minds.” From this he derives a rule of reasoning in nat­ural philosophy. “We no other way know the extension of bodies than by our own senses.... That abundance of bodies are hard, we learn by expe­rience.... That all bodies are impenetrable, we gather not from reason, but from sensation.”²⁰⁹

Newton’s experimental method is a synthesis of Hippocrates and Galileo, usually referred to as the hypothetico-deductive method. First, use experiments to simplify the phenomena, then discern and define character­istics that experiments show to vary quantitatively, and seek a mathematical expression for their variance. These are the “hypotheses” of “hypothetico- deductive,” though Newton would take exception to that term. Finally, re­turn to experiments to verify the deductions, reduce them to a common form, and identify additional causes or new phenomena suggesting further experiments. Experiments are the beginning and end of every advance on a scientific problem. It is always the sensible facts of experiment that are explained, and the explanation is only as comprehensive and exact as it is expressed in mathematical language.

Newton’s use of experience was more fruitful than anything ever had been in natural philosophy. He combines Galileo’s emphasis on mathematically expressed data with Bacon’s (and Galen’s) emphasis on methodical testing. What is new in Newton is the formulation of hypotheses in response to em­pirical data now stated exactly enough to allow stringent tests. It is rare be­fore Newton that data are formulated with sufficient exactitude to make this procedure fruitful, although it also makes natural philosophy fallible in a way that would be unacceptable to Aristotle, Bacon, Descartes, or Hobbes. “No deduction from an accepted principle, no matter how general or clearly de­rived from past phenomena, can therefore pass for absolutely or physically certain, without careful and continued experimental verification.”²¹⁰

Newton is more disposed than Descartes or even Galileo to seek experi­mental confirmation for his deductions, and was annoyed by the hostility this scruple drew. Descartes’s first principles are clear and distinct ideas, known by the light of nature. In Newton, such principles are one and all empirical in origin, inductions from experimental data, with the ultimate criterion in natural philosophy being empirical rather than mathematical. Newton shares Galen's impatience with ideas that no experiment can illuminate. “The proper method for inquiring after the properties of things is to deduce them from experiments.” His term for anything not so deduced is hypothesis, and “Hypotheses, whether metaphysical or physical, whether of occult qualities of mechanical, have no place in experimental philosophy”²¹¹

This emphatic disavowal of hypothesis is something of a puzzle in light of Newton's seeming obvious reliance on what he claims to disregard; for instance, his hypothesis of particulate light and an ether to propagate it. It seems that an acceptable function for hypothesis is to direct inquiry and sug­gest experiments, as Bacon also held. An unpublished draft for a planned third edition of Principles gives examples of objectionable hypotheses: that the quintessence is different from the four elements and subject to none of the senses; that prime matter possesses no qualities; that planets inhere in solid orbits or float in subtle matter. Compare those seventeenth-century examples with what Aristotle calls his protai hypotheseis, which are the assumption of simple movements (e.g., simply up or down); that such movements belong only to elements; and that they are limited in number. Elsewhere he says, “We hypothesize, and hypothesize on the basis of what we see, that nature never fails nor makes anything in vain.” In a letter to Cotes, then preparing the second edition of the Principles, Newton explains his objection to all of these “hypotheses” “The word Hypothesis is here used by me to signify only such a proposition as is not a Phaenomenon nor deduced from any phaenomenon but assumed or supposed without any experimental proof”²¹²

He allows no starting point other than the analysis of observations. Empirically undecidable ultimate causes fuel futile dispute. Newton disavows the intention to determine anything with regard to what force essentially is. What is light? Something propagated according to a mathematical law is all he will say (on the record, as it were; he is more suggestive in his alchemical notebooks). Criticized for not explaining what color is, Newton replies that it is not his intention and is beside his purpose; he wanted only to show the experimental observations, “and leave to others to explicate by Mechanical Hypotheses the Nature and Difference of those qualities.”²¹³

In the century between Galileo and Newton, the Aristotelian assump­tion that something meriting the name scientia must be demonstrative and necessarily true is displaced by a new natural philosophy that elevates the “medical” cognition of sagacious conjecture controlled by method­ical experiments to an independent source of natural scientific knowledge. Galileo knows his demonstrations are not Aristotelian scientia, but looks to the science of Archimedes for a no less rigorous paradigm, while Boyle and Locke, disdaining demonstrative certainty, look to the medical tradi­tion. Both tendencies culminate in Newton. “Although the arguing from experiments and observations by induction be no demonstration of general conclusions,” he writes, “yet it is the best way of arguing which the nature of things admits of, and may be looked upon as so much the stronger by how much the induction is more general.”²¹⁴

We approach the situation of today, where we consider it normal that sci­ence is constantly revising and disdains finality. Writing in the Philosophical Transactions (1715), Newton wondered why critics (Leibniz) regarded it as almost a crime for a natural philosopher to be content with what he can establish experimentally and acknowledge residual uncertainty. “Must Experimental Philosophy be exploded as miraculous and absurd, because it asserts nothing more than can be proved by Experiments, and we cannot yet prove by Experiment that all the Phaenomena of Nature can be solved by meer Mechanical Causes?”²¹⁵