Thus I see the problem of knowledge in a different way from that of my predecessors. Security and justification of claims to knowledge are not my problem. Instead, my problem is the growth of knowledge: in which sense can we speak of the growth or the progress of knowledge, and how can we achieve it? 
Category: Objective Knowledge
Oxford University Press: 1979.
What is the characteristic difference between a scientific theory and a work of fiction? It is not, I hold, that the theory is possibly true while the descriptions in the story are not true, although truth and falsity have something to do with it. The difference is, I suggest, is that the theory and the story are embedded in different critical traditions. They are meant to be judged by quite different traditional standards (even though these standards might have something in common.)
What characterizes the theory is that it is offered as a solution to a scientific problem; that is, either a problem that has arisen before, in the critical discussion of earlier tentative theories, or (perhaps) a problem that discovered by the author of the theory now being offered, but being discovered within the realm of the problems and solutions belonging to the scientific tradition. 
Among the inmates of my ‘third world’ [World 3] are, more especially, theoretical systems; but inmates just as important are problems and problem situations. And I will argue that the most important inmates of this world are critical arguments, and what may be called—in analogy to a physical state or to a state of consciousness—the state of a discussion or the state of a critical argument; and, of course, the contents of journals, books and libraries. …
Let me repeat one of my standard arguments for the (more or less) independent existence of the third world [World 3].
I consider two thought experiments:
Experiment (1). All our machines and tools are destroyed, and all our subjective learning, including our subjective knowledge of machines and tools, and how to use them. But libraries and our capacity to learn from them survive. Clearly, after much suffering, our world may get going again.
Experiment (2). As before, machines and tools are destroyed, and our subjective learning, including our subjective knowledge of machines and tools, and how to use them. But this time, all libraries are destroyed also, so that our capacity to learn from books becomes useless.
If you think about these two experiments, the reality, significance, and degree of autonomy of the third world [World 3] (as well as its effects on the second and first worlds [Worlds 2 and 1]) may perhaps become a little clearer to you. For in the second case, there will be no re-emergence of our civilization for many millenia. [107-8]
If mental entities or, better, mental states should exist—and I myself do not doubt that the do exist—then positing mental states is necessary for any true explanation of them; and shoud they one day be reduced to physical states, then this will be a tremendous success. 
In upholding an objective third world [World 3] I hope to provoke those whom I call ‘belief philosophers’: those who, like Descartes, Locke, Berkeley, Hume, Kant, or Russell, are interested in our subjective beliefs, and their basis or origin. Against these belief philosophers I urge that our problem is to find better and bolder theories; and that critical preference counts, but not belief. 
The commonsense theory of knowledge (which I have also dubbed ‘the bucket theory of the mind’) is the theory most famous in the form of the assertion that ‘there is nothing in our intellect which has not entered it through the senses’. (I have tried to show that this view was first formulated by Parmenides—in a satirical vein: Most mortals have nothing in their erring intellect unless it got there through their erring senses.) 
The phenomenon of human knowledge is no doubt the greatest miracle in our universe. It constitutes a problem that will not soon be solved, and I am far from thinking that the preset volume makes even a small contribution to its solution. But I hope that I have helped to restart a discussion which for three centuries has been bogged down in preliminaries.
Since Descartes, Hobbes, Locke, and their school, which includes not only David Hume but also Thomas Reid, the theory of human knowledge has been largely subjectivist: knowledge has been regarded as a specially secure kind of human belief, and scientific knowledge as a specially secure kind of human knowledge.
The essays in this book break with a tradition that can be traced back to Aristotle—the tradition of this commonsense theory of knowledge. I am a great admirer of common sense which, I assert, is essentially selfcritical. But while I am prepared to uphold to the last the essential truth of commonsense realism, I regard the commonsense theory of knowledge as a subjectivist blunder. This blunder has dominated Western philosophy. I have made an attempt to eradicate it, and to replace it by an objective theory of essentially conjectural knowledge. This may be a bold claim but I do not apologize for it. [Preface]
But I should go even further and accuse at least some professional historians of ‘scientism’: of trying to copy the method of natural science, not as it actually is, but as it is wrongly alleged to be. This alleged but non-existent method is that of collecting observations and then ‘drawing conclusions’ from them. It is slavishly aped by some historians who believe that they can collect documentary evidence which, corresponding to the observations of natural science, forms the ’empirical basis’ for their conclusions. …
Worse even than the attempt to apply an inapplicable method is the worship of the idol of certain or infallible or authoritative knowledge which these historians mistake for the ideal of science. Admittedly, we all try hard to avoid error; and we ought to be sad if we have made a mistake. Yet to avoid error is a poor ideal: if we do not dare to tackle problems which are so difficult that error is almost unavoidable, then there will be no growth of knowledge. In fact, it is from our boldest theories, including those which are erroneous, that we learn most. Nobody is exempt from making mistakes; the great thing is to learn from them. 
Thus I oppose the attempt to proclaim the method of understanding as the characterisitic of the humanities, the mark by which we may distinguish them from the natural sciences. And when its supporters denounce a view like mine as ‘positivistic’ or ‘scientistic’,* then I may perhaps answer that they themselves seem to accept, implicitly and uncritically, that positivism or scientism is the only philosophy appropriate to the natural sciences. 
* The term ‘scientism’ meant originally ‘the slavish imitation of the method and language of [natural] science’, especially by social scientists; it was introduced in this sense by Hayek in his ‘Scientism in the Study of Society’, now in his The Counter-Revolution of Science, 1962. In The Poverty of Historicism, p. 105, I suggested its use as a name for the aping of what is widely mistaken for the method of science; and Hayek now agrees (in his Preface to his Studies in Philosophy, Politics and Economics, which contains a very generous acknowledgement) that the methods actually practised by natural scientists are different from ‘what most of them told us … and urged the representatives of other disciplines to imitate’.
[A]ll the important things we can say about an act of knowledge consist of pointing out the third-world objects of the act—a theory or proposition—and its relation to other third-world objects, such as the arguments bearing on the problem as well as the objects known.