Art of Reasoning

Let’s say that I ask you a simple question: “Do you have hands?”

Do you even need to look before you answer? Of course not! You have used them, you have seen them, and you have felt them too many times to even describe. The question seems outlandish! Of course you have hands.

In short, you have “observed” your hands, empirically speaking. You have a very intimate, experiential relationship with your hands. They are literally “part of you,” you know them so intimately. Your hands (or feet or so on) are the closest things to what we would call “direct observables.” If you are able to “observe” anything at all, you can “observe” your hands!

Now, let’s contrast that example with another one. Let’s say that I ask you another simple question: “Does your car have an idler arm?”

Ah, now that one is much harder for most people. It is the same type of question, but for most people there is some “distance” between the question and the answer.

Even if you have enough mechanical background to know what an idler arm is, you have probably never checked to see for yourself. Or, perhaps you don’t know that cars with rack and pinion steering have no idler arm. You might know that your car has rack and pinion steering, but you didn’t know that such cars have no idler arm. So, perhaps you thought that your car could have both rack and pinion steering and an idler arm.

Notice that in this case you might “know” something just in virtue of some theoretical knowledge without having directly observed the state of affairs for yourself. You can legitimately say, “I know that my car has rack and pinion steering, and I know that such systems don’t employ an idler arm. So, my car does not have an idler arm.” But now you are appealing to inference rather than actual observation! (I’m not saying that this is bad. I’m just noting that there is some “additional distance” between your “knowledge” regarding idler arms compared to the “distance” regarding your knowledge of your hands.)

Here is an important point: Even though you might “know” about your car’s idler arm or lack thereof only inferentially, you could in principle go actually observe the state of affairs directly. You could hoist the car up, go under it, inspect the steering apparatus with your eyes and hands (and tongue, if you were so inclined). In short, you could in principle put yourself in very direct contact with the very entity that is under consideration, and you could experience it as that very entity.

We call things like hands and idler arms “in-principle observables.” You can in principle directly observe these entities being exactly the things they are purported to be. The terminology you would use to describe such entities would be laden with propositions that could be themselves directly experienced: “Hands have fingers that bend,” and, “Fingers have knuckles that facilitate the bending of the fingers,” and, “fingers have skin and nerves that convey hot and cold and touch,” and so forth. And each of the entities of the sub-descriptions would be directly observable as the very things that they are described to be.

However, scientists often refer to entities that are “in-principle unobservables.” When you keep reducing and reducing your descriptions to more and more “fundamental” entities (as physicists love to do), you find your “hand” descriptions starting to consist of not fingers and knuckles and so forth but consisting instead of exotic entities that can never be directly observed as they are claimed to be. Sub-atomic particles are classic examples of such entities.

There is no way in principle to directly observe such entities; their existence can only be inferred. And they can never be experienced by us as the very things they are purported to be. In fact, quantum theory proposes that our very acts of “observing” (even at great distance from these things as they “really” are in themselves) changes the entities. We cannot observe a quark in itself, in isolation, like we observe our own hand. Instead, we can only “observe” quarks as readings on gauges and as probabilistic equations that predict yet more readings on gauges. (That’s a bit of an oversimplification, but accurate enough for the point we are making here.)

So, empirically-known entities can be known on a continuum from “direct observable” to “in-principle observable” to “in-principle unobservable.”

The exemplar cases of direct observables are those entities that we most intimately perceive as parts of ourselves with as little intervening inferences as possible. At the other end of the spectrum, the exemplar cases are in-principle unobservables that we never “perceive” at all but that we believe exist strictly because of inferences.

Of course, it can be argued that there is no such thing as even an exemplar case of a direct observable. Perhaps you do not really have hands. What you really have are ideas of hands that emerge due to sensations of hands that are themselves merely products of your mind. And even if you do not go all the way into idealism, the “veil of perception doctrine” is that we never directly perceive reality at all; we only perceive our sensations of whatever is causing those sensations, and we really have no idea if those causal agents even remotely resemble the things we conveniently call “hands” and so forth.

Actually, the veil of perception doctrine is a serious issue, and we shall return to it in some depth later in this seminar. However, for now, I will say that this is a “distinction without a difference” for the purposes of the continuum I’m talking about. Here is why.

The question about the role of inference in “direct observables” does not help to explicate the fact that there is still an obvious continuum between entities that we do perceive (however we do it) as the very things we are referring to when we talk about them and those entities that we in principle cannot ever perceive as the very things we are referring to when we talk about them. Regardless of what inferential steps are involved in the direct observables, there are obviously layers of whole different types of inferences in in-principle unobservables. And that fact is enough for our purposes now.

Finally, notice that this continuum is not limited strictly to “entities.” The same principle can be employed when talking about events, for example. An event I have just witnessed personally is a very different event from one described to me by somebody else. And, as more and more “distance” (in time) occurs, the more inferences are required to believe in the event. I can have a direct observation of an event in the second in which it occurs. I cannot in principle directly observe any event that I did not observe for myself in the second of occurrence. So, events that I did not directly witness for myself are all in-principle unobservables for me, and I must rely upon inferences that are not required when I directly observe an event.

Regardless of the veil of perception doctrine (which introduces a “constant,” if you will), the more “layers” and “distance” there is between the entities and our ability to directly perceive them as they are purported to “really exist,” the more inferences are required to describe them. And those additional layers of inferences will themselves appeal to entities that themselves require inferences, because they are not direct observables.

So, as we move forward in this seminar, just be aware of the continuum of observables, and be aware of the nature of the inferences that must be employed to describe and “verify” the purported entities. Empiricism (science) is about observation in the strictest sense! So it is very important to be crisp and clear about exactly what is being observed, particularly in cases where the purported entities and/or events are in-principle unobservables.