IB Philosophy 6.1.3 Methodologies of Science Notes

The Nature of Theories and Hypotheses

1. The previous topics in this sub-theme examined the nature of science (6.1.1) and the demarcation problem (6.1.2). This final topic shifts from what science is and how it is distinguished from non-science to the more specific question of how science is conducted: what methods do scientists use, and can these methods produce unified knowledge?
2. A hypothesis is a provisional, testable explanation for a specific phenomenon. In scientific practice, hypotheses are formulated in response to observations or theoretical puzzles and then subjected to empirical testing through experiment or systematic observation.
3. A scientific theory is more comprehensive than a hypothesis. Theories are broad frameworks that organize and explain a wide range of phenomena, integrating multiple hypotheses and established laws into a coherent explanatory structure.
   1. Examples include the theory of evolution by natural selection and Einstein's general theory of relativity.
4. Theories provide the conceptual framework within which hypotheses are formulated and tested.
   1. A hypothesis that proves successful may be incorporated into an existing theory, while a theory that repeatedly generates failed hypotheses may need to be revised or replaced.
5. Scientific laws describe regularities in nature (such as Newton's laws of motion) but do not by themselves explain why those regularities hold. Theories go beyond description to provide explanatory mechanisms: they tell us not just what happens, but why it happens.
6. As discussed in 6.1.1, Popper, Kuhn, and Feyerabend offered different accounts of how science develops and what counts as a legitimate scientific method. This topic builds on their insights by examining the specific methodological tools that scientists employ: induction, deduction, unification, and reduction.
7. Understanding the nature of theories and hypotheses also connects to broader questions in epistemology (the branch of philosophy concerned with knowledge and how we acquire it), particularly the sources of knowledge (3.1.3).
8. Scientific methodology is one approach to generating knowledge, and its reliability depends on whether the reasoning patterns underlying it, inductive and deductive, can be trusted.

Induction and the Deductive-Nomological Model

1. Induction is the process of reasoning from particular observations to general conclusions. When scientists observe that every sample of copper they have tested conducts electricity, they inductively conclude that all copper conducts electricity.
2. Induction has traditionally been regarded as the foundation of empirical science.
3. However, as the philosopher David Hume famously argued, inductive reasoning can never provide certainty. No matter how many observations confirm a generalization, it remains logically possible that the next observation will contradict it.
4. This is the problem of induction, one of the most significant challenges in the philosophy of science.
5. Karl Popper, as discussed in 6.1.2, attempted to sidestep the problem of induction by arguing that science proceeds through conjecture and refutation rather than through inductive generalization. On Popper's account, scientists propose bold hypotheses and then attempt to falsify them, rather than accumulating confirming instances.
6. Despite Popper's critique, induction remains central to scientific practice. Scientists regularly use statistical methods, probabilistic reasoning, and pattern recognition to draw general conclusions from limited data.
7. The question is not whether scientists use induction, but whether inductive reasoning can be justified philosophically.

The Deductive-Nomological Model

1. The deductive-nomological (D-N) model of explanation, developed by Carl Hempel and Paul Oppenheim, says that a genuine scientific explanation must show the event was predictable, given what we already knew.
   1. It works like a logical argument: start with one or more general laws (rules that hold across all cases) and specific initial conditions (the particular facts of this situation), and the event to be explained follows as the logical conclusion.
   2. The event to be explained follows logically from these premises.
   3. For example, the D-N model would explain why a particular ice cube melted by citing the general law that ice melts above zero degrees Celsius and the initial condition that the room temperature was 25 degrees. The melting follows deductively from the law and the conditions.
2. The D-N model has several strengths.
   1. It provides a clear, formal structure for scientific explanation, making explicit what counts as an adequate explanation.
   2. It also captures the predictive power of science: if we know the relevant laws and conditions, we can predict what will happen.