Knowledge requires inquisitiveness and creativity. It cannot advance far inertially. The role of the knower is to actively ask questions and look for answers, not to sit back passively and assume all is well. Knowledge is aconstructiveactivity.

In forming one’s opinions, one has tothink things through, and not unfocus one’s stare and avoid the effort. One should not rely excessively on generally-held opinion, though of course its general acceptance is in most cases well-earned. One is duty-bound to verify, repair, and contribute, if one can.

Knowing is not mere maintenance work, ‘when something goes wrong, fix it’, but involves searching for flaws or improvements even without apparent cause. Speculation, the attitude of ‘what if things are otherwise than they now seem or are said to be?’, has considerable value in the pursuit of truth.

In forming our world-view, we all make use of some prejudicial ideas, or preconceptions.We take for granted many basic assumptions, often unconsciously, without awareness of having made them, without ever having analyzed them to any great extent, without having tried the alternative assumptions.

Some such assumptions become deeply ingrained in a sub-culture, a culture, a period of history, or all human thinking. If such a philosophical prejudice is institutionalized, it is called a dogma. But our concern here is also with unconscious dogmas. My purpose in this chapter is to show informally how such ideas can be brought out into the open and evaluated.

The first thing is always a willingness to face the issue explicitly, and confront the possibly unpleasant results. Next, try to reconcile the apparent opposites, find asynthesisof some sort. Look for the ultimate premises, and even if speculatively, consider alternative conceptions which are capable of fitting the known facts.

The synthesis of knowledge is an attempt to ‘wrap it all up’, or at least take stock of the situation as a whole thus far. You lay out the data you have, and you firmly evaluate their significance on your current opinions:

·Where are you at?

·What do you know, what don’t you know?

·What do you need to know?

·What can you know, what can’t you know?

An inventory and a summation, to the best of one’s ability.


Thus far, one’s logic may have been lenient. One perhaps wanted to get ahead, to cover ground. There was no time for scrupulous analysis of the degrees of logical probability in one’s information and inferences. Now, the whole must be reviewed, each part considered in the light of all the others. One must disengage oneself, and become a neutral referee between contending ideas.

One must challenge one’s previous viewpoints. One must look at thingsmore critically, less intent on the object than on the process which led us to our viewpoint. It is time to linger on detail, digress a little, consider the full impact of what one is saying.

This may mean taking-off in all directions, even to the point of looking into metaphysical implications. One should not limit one’s vision to one field, but range as far and wide as necessary to prove a point. One may appeal to epistemological reasons, or consider ontological outcomes.

Initially, we accept our deductions and inductions with fair-minded tolerance. But, in the final analysis, the limits of one’s certainties must be emphasized. There are different degrees of strictness of outlook; different modalities of implication. There is a ‘take it for granted’, working level; and there is a more severe, philosophical level.

Within philosophy, ‘anything goes’, and even doubts about logic, about the laws of thought or the trustworthiness of experience, have some legitimacy. At this strict level, it is healthy to give skepticism some rein, to enable us to judge with honest detachment (though total skepticism remains invalid, since paradoxical).

For instance, an adductive argument is ordinarily allowed; it is acknowledged to increase the probability of the conclusion. But viewed deductively, its inference is worthless. Synthetic logic probes into theories by considering, not only their internal consistency and continuing confirmation, but more fully and deeply:

·What are the ultimate assumptions?

·What are the implied conclusions?

·Are there alternative premises or inferences?

·How do they compare and contrast, how much do they agree or disagree?

·How reliable are the apparent consistencies and how serious are the seeming inconsistencies?

·How solid are the logical connections between postulates and predictions, and what are they based on?

·What is the data, and how empirical is it?

The enterprise of science is an open pursuit of knowledge. If it is objective, as it wants to be, then it should have no prejudice as to what the object presented to it is, or how it got there. The process of adduction, we saw, has the form:

If Theory, then Predictions:

Yes to any of these predictions,

therefore, possibly yes to the theory.

(but if No to any prediction, no to the theory.)

This may be countered by the equally valid adduction:

If Other Theory, then Same or Other Predictions:

Yes to any of those predictions,

therefore, possibly yes to the other theory.

(but if No to any prediction, no to the theory.)

Now, note the following methodological implications, according to strict logic.Here, the emphasis is more on the criteria of relevance and competitiveness. Utilitarian or esthetic criteria are not granted much weight, so that a far-fetched theory may be as respectable as a more obvious one.

(i)If the two theories make predictions which coincide exactly, or if none of their predictions logically impinge on each other, there is no way to choose between them. They are effectively undifferentiated, or irrelevant to each other.

(ii)If the two theories have some different prediction(s), but these differences are in practise or in principle untestable, again there is no ground for preferring the one to the other. But we may not regard untestable predictions as strictly logically equivalent to non-predictions.

(iii)If the two theories have been confirmed by adduction to an equal degree of logical probability — that is, as many times, by equally firmly-implied and credible phenomena, whether these phenomena be the same or different — no conclusion is permissible. The logical modality is the same.

All this applies as well to theories with mutually exclusive postulates, and to theories with postulates which are independent of each other.


Clearly, the mere fact that someone takes up a theory of his own, and keeps testing it, and finds it repeatedly confirmed, doesnotin itself make his work fully scientific, and in accord with the neutral demands of logic.

The scientific approach, under the terms set by epistemology (notontology, mind you), is to consider all other available theories, and busy oneself to an equal extent in testing and confirmingthem too. If difficulties arise, we are duty-bound to try to repairallthe known theories with equal zeal, and not just the one we hope will win, for whatever personal reasons.

The same methodological demands should be made for one’s own pet theory, as one makes for others’; and the same leniency should be granted to others’ theories, as one grants to one’s own.

Similarly, one should refrain from negative pronouncements on sectors of human inquiry about which one is not adequately informed. In other words, one may regard oneself as a specialist, advancing a limited domain of the inquiry, without laying claim to any authority beyond those limits.

To be professional in the pursuit of knowledge, completely objective and neutral, without prejudice, one must proceed in accord with the rules of argument set by logic. The scientist who merely works on one theory at a time, without regard to the inadequacy of his methodology, is kidding himself and everyone else; he has ignored the alternatives, his conclusions are strictly invalid.

Of course, one can only do one thing at a time; but one must always keep the global perspective in mind, or refrain from comment.[2]

[1]In the present (1995) edition of the book, the present chapter (49) is somewhat abridged. When I wrote it in 1990, I unfortunately included in it a number of comments which can only be classed as religious polemics. Having since then considerably evolved in my views concerning religion, I felt it wise to tone down the essay and restore it to a purely methodological function.

[2]We can use the story of Galileo (as I was taught it at school) to give an example of synthesis. Until Galileo’s time, people believed that our planet was the center of the Universe (comprising all the heavenly bodies – Sun, Moon, Planets and Stars); then, various observational and theoretical discoveries changed our picture of things, and the Sun became central (to our solar system, at least). This was initially received very harshly by a certain religious establishment; everyone knows the story. Today of course, after the Relativity theory, the issue is irrelevant to astronomy.

Now, I have no personal attachment to the pre-Galilean thesis, nor does my religion advocate it — the spiritual centrality of mankind has nothing to do with the physical position of planet Earth. However, it seems to me that the argument was in any case fallacious. For the new theories only posited that the mathematical formula describing the movement of the Planets around the Sun was muchsimplerthan the formula which placed Earth at the center of things — but that did not prove that the latter more complex equations could not be formulated.

If I am not mistaken, every trajectory can in principle be ‘turned on its head,’ and described mathematically fromanypoint of origin. Simplicity is an inductive criterion, but it is never ontologically unassailable. Thus, it is ironic Galileo was in fact not even a threat to the world-view of the Inquisitors. For me, this example illustrates the need to always clarify the precise degree of conflict between theses.