TRUTH-RELEVANT SPECULATIONS

whose antecedent is h. But then it is the entire conditional that is the speculation, not h by itself.^[20]

If P introduces h in a way that satisfies the “theorizing” conditions (a) and (b), we might say that

(Spec): h is a (truth-relevant) speculation for P if and only if P does not know that there is evidence that h.

One way to know that there is evidence that h is to know what that evidence is and to know that it is evidence that h—i.e., to know of some fact e that it is evidence that h. But it is also possible to know that there is evidence that h without knowing what that evidence is. If authoritative textbooks all tell me that that there is evidence for the existence of the top quark without telling me what that evidence is, and I intro­duce the assumption that the top quark exists and do so in a way that satisfies the “theorizing” conditions, then my as­sumption is not a speculation for me, since I know there is evidence for its existence.

If you introduce a hypothesis h under the “theorizing” conditions, and you don't know that there is evidence that

h, then if h is true you don't know that it is. Speculating in this way entails lack of knowledge of the truth of the specu­lation. However, the converse is not true. If you introduce h under the “theorizing” conditions without knowing that h is true, it doesn't follow that you are speculating, at least on the concept of speculation I am proposing, since you may know that there is evidence that h is true. On this concept, lack of knowledge that h is a necessary but not a sufficient condi­tion for speculating that h. By extension, if you introduce h under the “theorizing” conditions, and unbeknownst to you h is false, the fact that h is false doesn't make h a speculation for you, since you may know facts that you know to be evi­dence that h is true.^[21]

A good deal more needs to be said about various ideas associated with (Spec).

In (Spec), h stands for an individual assumption. Now, as is the case with the three examples of speculations from physics at the beginning of this chapter, theories usually con­tain sets of assumptions, some of which may be speculations, some not. It is not my claim that each assumption in a spec­ulative theory is necessarily a speculation.

I have spoken of speculators introducing assumptions. As noted earlier, normally this is done in the course of “theorizing” activities such as explaining, predicting, calculating, etc. But occasionally it is done with little if any “theorizing” on that occasion by writing them down and calling them “assumptions,” “speculations,” or “hypotheses” that the speculator believes to be true or possible and worth further investigation at some point. Newton does this in his Queries in the Opticks when he speculates about the particle nature of light, though even there he theorizes a bit by of­fering a few arguments against the rival wave theory. I will call both cases speculating while “theorizing,” even if the theorizing is intended for a future occasion.

Among the speculations introduced in a theory consisting of many assumptions, some may be intended as literally true, some may be thought of as approximations with different degrees of closeness to the truth, while still others may be regarded as just possibilities worth considering. In chapter 3, I will argue that, despite Newton's claims to the contrary, his law of gravity was a speculation, but one he believed to be lit­erally true. His assumption that the only gravitational force acting to produce the orbit of a given planet was the gravita­tional force of the sun acting on the planet was also a specu­lation, but an assumption he regarded as only approximately true. And the speculation that “the center of the system of the world is at rest”—which he explicitly classified as a specula­tion (or “hypothesis”)—he perhaps regarded as a possibility.

Whether speculations are introduced boldly with the idea that they are true, or more cautiously with the idea that that if not true they are close to it, or even more cautiously with the idea they are possibilities worth considering, does not affect their speculative status if the speculator does not know that there is evidence for them. What it does affect is the question “Evidence for what?” Evidence that h is a pos­sibility, or that h is close to the truth, or that h is true will usually be different. If I introduce the assumption that h is a possibility—meaning that it is not precluded by known laws or facts—without knowing that there is evidence for the claim that it is not so precluded, then I am speculating when I claim that h is a possibility.

Accordingly, we could modify (Spec) by introducing distinctions between types of speculations: speculating that h is true, speculating that h is close to the truth, speculating that h is a possibility, and perhaps others. But I will not do so. When some assumption h is classified as a speculation, the focus of the scientific community is normally on the truth of h, even if the speculator is introducing h only as a pos­sibility. When Brougham criticized Young for producing a speculation (“a work of fancy”), it is not very convincing for Young to reply: “No I am not, I am only saying that the wave theory is a possibility, that it is consistent with Newtonian mechanics, and for that claim I can provide evidence.” Even if this reply correctly represents Young's intentions, the main interest of scientists, including Brougham, is the question of whether the theory is true. Scientists want to know if there is evidence that light is a wave motion in the ether, not simply whether this is a possibility. So, in such cases I will retain (Spec), and say that even if the speculative assumption h was introduced by P with the idea that it is a possibility, h is a speculation for P with respect to the truth of h, since P does not know that there is evidence that h is true.

What about “closeness to truth”? Newton in defending Phenomenon 1 pertaining to the Keplerian motions of the moons of Jupiter introduces the assumption that the orbits of these moons are circular. The assumption itself is literally false, which Newton realized, but it is a good approximation. Since Newton had evidence that this is a good approximation, the latter claim was not a speculation for him. In view of such cases we could keep (Spec) as is but add another type of speculation for cases in which no evidence is known for “h is close to the truth.” Or we could simply retain (Spec) without adding other types, and understand “true” in a broad way to include “true or close to it.” I prefer the latter.

There is a view about the introduction of speculations that I want to reject. It is based on a distinction that philosophers once regarded as important (perhaps some still do) between the “context of discovery” (when one first gets the idea of the hypothesis) and the “context of justification” (when one is attempting to test or defend it by providing evidence). Those wedded to this distinction claim that speculations ap­pear in the first context, not the second. On the view I am defending, whether something is a speculation does not depend on whether it is introduced when one first gets the idea, or afterwards when one is (or is not) attempting to de­fend it. For speculations, what matters is only that they are introduced in the course of “theorizing” activities with the idea that they are truths or close to it, or at least possibilities worth considering.

Finally, in accordance with (Spec), a “theorizing” as­sumption h introduced by P can be a speculation even if P does not believe that h is the sort of assumption for which evidence is possible. Assumption h might be regarded by P as “metaphysical,” or “theological,” or of some other sort for which there can be no evidence. In such a case, it follows trivially that P does not know that there is evidence that h, so that if P introduces h under the “theorizing” conditions, then h is a speculation for P. At the other extreme, what if P regards h as self-evident, having and needing no evidence? Whether there are cases of self-evident propositions having and needing no evidence is controversial. But if there are, ad if P knows that h is one of them and “theorizes” using h, then, since P knows that h is “self-evident,” and therefore true, I will say that P is not speculating.

6.