Beyond Kuhn

Scientific Explanation, Theory Structure, Incommensurability and Physical Necessity

Edwin H.-C. Hung

• Editore: Routledge
• Anno: 2017
• ISBN:
  9781351955638
• Edizione: 1
• Argomento: Philosophy
• Categoria: Philosophy History & Theory

Thomas Kuhn's celebrated work, 'The Structure of Scientific Revolutions' revolutionized thinking in the philosophy of science and to a large extent his 'paradigm shift' view has replaced logical positivism and the philosophy of Karl Popper. This book goes beyond Kuhn by explicating the non-deductive notion of 'paradigm shift' in terms of the new concept of representational space. In doing so, Edwin H.-C. Hung is able to produce the first-ever unitary theory that solves the five central problems in the philosophy of science: scientific explanation, the structure of scientific theories, incommensurability, scientific change and physical necessity. The book identifies the main task of science as representing reality. This involves the construction of a representational space and the subsequent modeling of reality with configurations of 'objects' in that space. Newton's mechanics, Einstein's relativity and quantum mechanics, then, all serve as representational spaces. 'Beyond Kuhn' is a significant progression in scientific methodology. Other than serving as a sequel to Kuhn's 'Scientific Revolutions', it will be of great use in the fields of artificial intelligence, cognitive psychology and education.

Chapter 1
Introduction: The Road to Kuhn and Beyond

What is [philosophy]? Philosophy is many things and there is no formula to cover them all. But if I were asked to express in one single word what is its most essential feature I would unhesitatingly say: vision. (Friedrich Waismann 1956: 32)

1.1 Truth

In search of truth, science has progressed through two stages: the empirical stage and the theoretic stage. At the empirical stage, typically induction by simple enumeration and Mill's five methods of induction are used to arrive at empirical generalizations such as ‘Ice floats on water’, ‘Metal expands when heated’, Hooke's law and Boyle's law.¹ The theoretic stage came later. In the beginning, the invention of theories was sporadic. Theories came in dribs and drabs. For instance, the ancient Greeks Leucippus and Democritus (c. 420 BC) speculated that matter is made of atoms, moving in a void. Ptolemy (c. 140 AD) proposed his epicycle theory of the planetary system. Theodoric of Freibourg (14th century) suggested that rainbows are the products of rain droplets. Then, in the 17th century, the great Newton launched his theory of gravitation. Henceforth, theories blossomed as wild flowers in spring. Modern history of science is the story of great theories.²

Philosophically, empirical generalizations do not pose any obvious problems. Theories, however, seem puzzling. What are they? What is their structure? How do theoretic terms acquire meaning since what they denote are usually unobservable? Traditionally, theories are conceived of as sets of statements. This we cannot agree with. In Chapters 3, 4 and 5, respectively, we shall show that theories have the structure of category systems, representational spaces and languages.

1.2 Explanation

In search of explanations, science has also progressed through two stages, namely, the causal-nomological stage and the theoretic stage. In the causal-nomological stage, phenomena are explained in terms of cause and effect, and causes are said to bring about their effects through laws of nature.³ Laws are characterized as being universal in that they apply throughout space and time.⁴ And, from them, all observed regularities follow.

Theoretic explanations are more sophisticated. For instance, Thomas Young (1773–1829) and Augustin Fresnel (1788–1827) independently proposed a wave theory to explain the phenomena of light. That theory is not merely a set of laws. It presents a totally new ontology to explain something familiar, namely, light phenomena that we daily encounter. Very often this new ontology (ontology of the explanans) is in blatant contradiction with that of the explanandum. A well-worn example is that of the special theory of relativity proposed by Einstein to explain the ‘strange’ results of the Michelson-Morley experiment. The Minkowskian space-time of the former is incompatible with the Newtonian space and time, which the latter presupposes.

It is said that both causal-nomological explanations and theoretic explanations are deductive in nature.⁵ Hence, it is generally taken that theoretic explanations are similar.⁶ But, prima facie, this cannot be. How can the explanandum follow logically from its explanans if they are incompatible with each other?⁷ It is not an isolated case that relativity is incompatible with Newtonian space and time (in terms of which the result of the Michelson-Morley experiment is described). For instance, quantum mechanics purports to explain phenomena such as black body radiation, the photoelectric effect and Rutherford's experimental results on the scattering of alpha-particles by metal foils, which are all expressed in classical terms.⁸

How do theories explain, if not deductively? This is the problem. The answer lies with the notion of conceptual shift. Theoretic explanations explain by replacing the current conceptual framework with a new one. The logic is non-deductive. We shall deal with this issue in detail in Chapter 2.

1.3 The Classical View of Science

The classical (empiricist) view of science had its roots in Aristotle. It was successively developed by Francis Bacon, John Locke, J.S. Mill, and Karl Popper, among others. Here are its basic tenets.

1. There is an objective reality, independent of thinking and thought. The highest aim of science is to capture that mind-independent reality, to give it as accurate a description as possible.
2. Contact with that objective reality is through the senses, which can be employed to collect data about that reality. The observer is a passive receiver of information.
3. These data, known as empirical data, form the basis of acquisition of truth about that objective reality. Inductivists (e.g., Bacon and Mill) see these data as the basis for inductive inferences, through which further truths about reality can be obtained. Thus science starts with the collection of empirical data. On the other hand, deductivists, led by Karl Popper, claim that science starts with problems and hypotheses. Empirical consequences of these hypotheses are then obtained through deduction for comparison with reality.
4. Empirical data (obtained through observations) may not be veridical. Nevertheless, they are objective in the sense that trained scientists in similar situations should largely perceive the same ‘thing’, and hence would arrive at the same empirical data when performing the same experiments.
5. Conflicting hypotheses often compete for acceptance. Various methods have been proposed for hypothesis choice.

In the first half of the twentieth century, the classical view was given an anti-realist, anti-metaphysical twist by the logical positivists, the proclaimed heirs to the empirical philosophy of Hume, Comte and Mach. Their main concern is with the structure of scientific theories, the meaning of theoretical terms, and the logic of scientific explanation.

1.4 The Logical Positivist View on Theories

Theories typically postulate theoretical entities, which are usually unobservable. How can theoretical terms be meaningful if they denote entities which are unobservable? How does the postulation of unobservable entities manage to explain observable phenomena? To solve these problems, logical positivists led by Rudolf Carnap, Ernest Nagel and Carl Hempel proposed the following two-tier view of science.⁹

According to them, the vocabulary of science can be classified into observational and theoretical terms. Observational terms denote the observables. They come with our ability to observe and are thus commonly shared by all scientists, irrespective of their theoretical mentality and beliefs. In contrast, theoretical terms come with theories. Each theory carries its own terms. What they denote, commonly known as theoretical entities, are usually unobservable.

A theory is said to consist of two parts, the internal principles and the bridge principles.¹⁰ The internal principles, containing only theoretical terms, specify the properties of those postulated theoretical entities. The bridge principles, on the other hand, have both theoretical and observational terms. They relate those theoretical entities to observable phenomena. Being so related, theoretical terms obtain their empirical meaning from the observational terms. Explanation of empirical generalizations is done through deduction with those internal and bridge principles as premises.

One distinctive feature of this philosophy is that there is a universal language for science, namely, the observational language. This language performs two functions: (i) it supplies theoretical terms with meaning via the bridge principles, and (ii) it is the medium in which results of observation are recorded. It can be seen that this is a two-tier view of science. Empirical data occupy the lower tier. Encoded in the observation language, they form the base, or foundation, of science.¹¹ The upper tier is occupied by theories encoded in theoretical terms. In contrast to empirical data, these theories usually compete. The successful ones are often later replaced. Thus the upper tier is always in a state of flux.

This logical positivist view of science is unfortunately full of difficulties. Here are some of the well-known problems.

1. The problem of observational/theoretical distinction: There seems to be no sound theoretical basis for the distinction between observational and theoretical terms.¹²
2. The ‘bridge principle’ problem: What logical forms should these bridge principles take? Carnap famously failed to capture these forms with his correspondence rules.¹³
3. The ‘deduction’ problem: It is claimed that the logical relationship between the explanans (internal principles) and the explananda (empirical generalizations) is that of deduction (via the bridge principles). This, however, is problematic, as has been pointed out in Section 1.2 above.

1.5 Kuhn's Paradigm View of Science

According to Thomas Kuhn (1922–96) and Paul Feyerabend (1924–94), the observational/theoretical distinction is untenable. There is no such thing as a stable observational vocabulary, independent of our theoretical commitments. The way we perceive d...