The Two Fundamental Problems of the Theory of Knowledge

Karl Popper, Troels Eggers Hansen, Andreas Pickel, John Kinory

• Publisher: Routledge
• Year: 2014
• Print ISBN: 9780415394314
• eBook ISBN: 9781135626839
• Edition: 1
• Topic: Philosophy
• Subtopic: Philosophy History & Theory

In a letter of 1932, Karl Popper described Die beiden Grundprobleme der Erkenntnistheorie – The Two Fundamental Problems of the Theory of Knowledge – as '…a child of crises, above all of …the crisis of physics.'

Finally available in English, it is a major contribution to the philosophy of science, epistemology and twentieth century philosophy generally.
The two fundamental problems of knowledge that lie at the centre of the book are the problem of induction, that although we are able to observe only a limited number of particular events, science nevertheless advances unrestricted universal statements; and the problem of demarcation, which asks for a separating line between empirical science and non-science.

Popper seeks to solve these two basic problems with his celebrated theory of falsifiability, arguing that the inferences made in science are not inductive but deductive; science does not start with observations and proceed to generalise them but with problems, which it attacks with bold conjectures.

The Two Fundamental Problems of the Theory of Knowledge is essential reading for anyone interested in Karl Popper, in the history and philosophy of science, and in the methods and theories of science itself.

Book I

The Problem of Induction

Experience and Hypothesis

The Two Fundamental Problems of the Theory of Knowledge Volume I

Chapter I

Formulation of the Problem

1. The problem of induction and the problem of demarcation. This analysis focuses on two questions: the problem of induction and the problem of demarcation.

The problem of induction:

We are only able to observe particular events, and always only a limited number of them. Nonetheless, the empirical sciences advance universal propositions, such as the natural laws; that is, propositions which should hold true for an unlimited number of events. What is the justification for advancing such propositions? What is actually meant by these propositions? These questions indicate in outline the problem of induction. The “problem of induction” will denote the question concerning the validity or justification of universal propositions of the empirical sciences. Or, put another way, can empirical propositions that are based on experience be universally valid? (Or more simply, can we know more than we know?)

The problem of demarcation:

Most of the empirical sciences, as their history shows, have sprung from the womb of metaphysics. Their last pre-scientific form was speculative-philosophical. Even physics, the most highly developed among them, has perhaps to this day not completely freed itself from the last remnants of its metaphysical past. Especially in recent times, it has been subjected to a revolutionary cleansing process. Metaphysical reasoning (for example, Newton’s absolute space and absolute time, Lorentz’s ether at rest) has been ruthlessly eliminated. The less highly developed sciences (for example, biology, psychology, sociology) have always been much more strongly laced with metaphysical elements than has physics, and the same is still true today. Indeed, even the view that metaphysics must be eliminated as “unscientific” is explicitly rejected by some proponents of these sciences.

Is metaphysics rightly rejected or not? What is actually meant by the terms “metaphysics” and “empirical science”? Is it possible at all to establish strict distinctions, to determine certain limits? These questions, which indicate in outline the problem of demarcation, are of general and decisive importance. Any form of empiricism must, above all, demand from the theory of knowledge that it secures empirical science against the claims of metaphysics. The theory of knowledge must establish a strict and universally applicable criterion that allows us to distinguish between the statements of the empirical sciences and metaphysical assertions (“criterion of demarcation”). The question concerning the criterion of demarcation is what I call the “problem of demarcation”. Or put another way: in case of doubt, how can one decide whether one is dealing with a scientific statement or “merely” with a metaphysical assertion? (Or more simply, when is science not science?)

This investigation will have to demonstrate that these two questions, the (Humean) problem of induction and the problem of demarcation (Kant’s question concerning the limits of scientific knowledge), can rightly be called the two fundamental problems of the theory of knowledge. The problem of demarcation deserves our primary interest. It is by no means of only theoretical-philosophical significance. Rather, it is of the greatest relevance for the separate sciences, particularly for the research practices of the less highly developed ones. But even from a philosophical-epistemological point of view, it proves to be the central problem to which probably all other questions of the theory of knowledge, including the problem of induction, can be reduced.

These epistemological questions are of an entirely different nature from the psychological question of how our knowledge actually comes into being. The question is not about the way in which scientific statements are discovered, or how they develop, but about their justification, and about their validity. The epistemological questions, as questions of justification or validity (Kant: “quid juris?”), must be strictly distinguished from psychological (and historical-genetic) questions of fact (“quid facti?”), that is, from questions concerning the discovery of knowledge.

(In the present work, factual psychological and historical-genetic questions of cognition will be discussed only to the extent necessary in order to separate these questions from the epistemological problem formulation and to eliminate them from the analysis.)

The view that the theory of knowledge should deal exclusively with questions of validity but not with questions of fact, makes it, so to speak, into a general methodology for empirical science. For method in science is not the way in which something is discovered,*¹ but a procedure by means of which something is justified.

*¹ Thus, methodology is distinguished here from heuristics. This does not mean, however, that heuristics has nothing to learn from methodology.

Chapter II

Deductivism and Inductivism

2. Comments on how the solutions are reached and preliminary presentation of the solutions. Are we justified in calling the problem of induction, but even more so the problem of demarcation, the fundamental problems of the theory of knowledge?

Are we justified in regarding the theory of knowledge as the methodology of the empirical sciences?

Evidently, these questions can only be answered through an analysis that takes into account the historical circumstances. However, such analysis need not, as a consequence, have a historical interest. It will only have to show that the typical problems that have, time and again, been dealt with by the theory of knowledge, are reducible to the problem of induction and then to the problem of demarcation; and it will also have to show that these problems may be viewed as methodological problems, and that such a view is a productive one.

For these reasons if not for others, much attention will be devoted to the presentation and criticism of the most important epistemological approaches; at all times, however, an attempt will be made to make this criticism productive; that is, to penetrate to the positive questions, to the methodological questions underlying the positions criticised.

According to the view advocated here, the “epistemological problems” can be divided into two groups. The first group contains methodological questions; the second contains speculative-philosophical questions, which in most cases may be described as misinterpretations of methodological problems. For the most part, typical epistemological prejudices (for instance, the psychologistic, the inductivist, the logicistic or the language-critical prejudice) can be held responsible for these misinterpretations. If this view is justified, then the productivity of the epistemological method and of a successful formulation of the epistemological problem will prove itself by allowing replacement of the questions from the second group by those of the first; in other words, not simply by dismissing the epistemological misinterpretations as pseudo-problems, but by identifying and solving the genuine and concrete methodological problems that underlie them.

For the following critical and positive analyses to be understood and assessed from a unified viewpoint, the most important points of the epistemological position advocated in this work will now be briefly highlighted. They will not be further explained at this point; this will be the task of the analysis itself (cf. Section 47).

a) On the method of the theory of knowledge:

The term “transcendentalism” will denote the view that epistemological assertions and concepts can and must be critically examined – exclusively – in terms of the actual justification procedure of the empirical sciences. This “methodological method” may (for reasons suggested in Section 9) be called the “transcendental method”. The theory of knowledge is a science of science. It relates to the individual empirical sciences in the same way as the latter relate to empirical reality; the transcendental method is an analogue of the empirical method. The theory of knowledge would, accordingly, be a theoretical science. It also contains free stipulations (such as definitions); yet it consists not only of arbitrary conventions but also of statements that are refutable by comparison with the actual and successful methods of the individual empirical sciences. All other epistemological methods (psychological, language-critical, etc.) are altogether rejected by transcendentalism – of course with the exception of logical criticism, the exposing of internal contradictions in the opponent’s position.

b) Fundamental ideas of the epistemological solution:

The view advanced here may be called radical “deductivism”. It holds that all scientific methods of justification are, without exception, based on strictly logical deduction, and that there is no induction of any sort qua scientific method.

Theories of knowledge may have either a deductivist or an inductivist orientation, depending on how they assess the significance of deduction (logical derivation) and of induction (generalisation). Thus, classical rationalism (Descartes, Spinoza), for example, has a strictly deductivist orientation (its model is geometrical deduction [Euclid]), whereas classical empiricism is inductivist. Radical inductivist positions (such as Mill’s) deny that deduction has any significance at all; for, it is argued, what can be deduced is only that which induction has originally placed in the major premises. But even intermediate positions (such as that of Jevons), which seek to characterise the empirical-scientific method as a synthesis of induction and deduction, will be rejected here as “inductivist”. The deductivist view advocated here denies that induction has any significance.

The only admissible inferences in an inductive direction – that is, proceeding from a theory’s minor premises to its major premises – are the deductive inferences of the modus tollens, the falsification of major premises by way of falsifying the conclusions deduced from them.

(The idea of a strictly deductivist theory of knowledge, if consistently applied, leads to simple solutions of epistemological problems. All the following considerations are based on this idea.)

A further consequence of deductivism and the rejection of induction may be denoted by the term “hypotheticalism”; that is, the view that empirical-scientific theories (universal empirical statements) can never be more than tentative assumptions, or unfounded anticipations,*¹ because an empirical verification of theories – a reduction of universal empirical statements to singular empirical statements (induction) – is logically inadmissible.

The position advanced here is empiricist by virtue of its fundamental principle (the fundamental thesis of empiricism) that only experience can decide the truth or falsity of an empirical statement.

According to the deductivist-empiricist view advocated here, there is only one relationship between natural laws, theories and universal empirical statements, on the one hand, and singular empirical statements (the “empirical basis”: cf. Section 11) on the other, namely tha...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Dedication
• Table of Contents
• Editor’s foreword
• Publisher’s acknowledgements
• Preface to the first German edition, 1978
• Introduction, 1978
• Exposition [1933]
• Book I: The Problem of Induction (Experience and Hypothesis)
• Book II: The Problem of Demarcation: Experience and Metaphysics
• Appendix: Summary Excerpt (1932) from The Two Fundamental Problems of the Theory of Knowledge
• Index of Names
• Index of Subjects