Theory and observation
This simplified version of Mendelian genetic theory will allow us to examine many of the features of scientific theories that philosophers of science have discussed.
The first important thing to say about Mendel's theory is that it was a great feat of creative imagination. He couldn't see (or touch or hear or taste) genes, so he had to postulate them in order to try to explain the results of his experiments. To postulate the existence of entities is to hypothesize that they exist.What, exactly, was involved in hypothesizing that genes exist? Certainly Mendel had to do more than say that he thought there were things called “genes.” What he had to do as well was to say what some of their properties were. Frege's theory of meaning can show us why he had to do this. In order for us to understand a term like “gene” it has to have a sense, which is, as you will remember, an associated mode of presentation. So Mendel had to say what something would have to be like in order to be a gene. You understand the name “the Morning Star” because you know that something is the Morning Star if and only if it is a heavenly body that usually appears at a certain point on the horizon in the morning. Mendel had to associate a similar sense with his word “gene.”
Because the word “gene” had no established sense associated with it, the outline of the theory I presented above is a sort of implicit definition of the word. To make it an explicit definition we have to remove the word “gene” from the theory as I summarized it above. We can then introduce the idea of a gene in a way that is equivalent to using a Ramsey-sentence, just like the one we used in chapter one to develop functionalism. Once more, we write out Mendel's theory as a single conjunction of the eleven claims and the laws: call this very long sentence MG (for “Mendelian genetics”).
Then we replace the word “gene” throughout with a variable, “X”, and other new terms, such as “allele,” with other variables. Let's suppose that these were the only new terms. We can now define genes and alleles, quite simply, as the two kinds of thing that satisfy this complex open sentence, MG*. Genes and alleles are, so to speak, any X's and Y's that make all thirteen of Mendel's propositions true at once. This way we can define the word “gene” in terms of notions that we already understand: notions such as phenotype (which just means the visible characteristics of the organism), organism, and parent. Of course, this isn't like an ordinary definition, where we define one word only in terms of others we already understand. There is the other term—"allele”—that we don't already understand. But just as the Ramsey-sentence of 1.7 allowed us to interpret “pain" even though the definition involved the concept of “worry," so here we have replaced the words “allele" and “gene" by variables at the same time, and come to understand “allele" along with the term “gene." Mendel's theory that there are genes and alleles thus amounts to saying that there are two kinds of entity that together satisfy MG*.The reason all of this is necessary, of course, is the fact that I mentioned at the beginning of this section: Mendel couldn't see or otherwise sense genes. They were not observable. Because of this he could not introduce the term “gene" in the way we can define the name “the Evening Star" or the predicate “is red," by pointing to something. (This was what Wittgenstein meant by an “ostensive definition" in 1.3.) That is why unobservable entities have to have their names introduced in terms of things that we can observe. For if we didn't connect their names in this way with things we could observe, we could never use the names. There would be no role for the names in our language because there would be no circumstances in which experience would lead us to use them.
The term “gene" refers to something that Mendel couldn't observe, but it is also what is called a “theoretical term." It is a theoretical term because it is introduced by way of a theory, in this case MG.
Philosophers have sometimes thought that all unobservable things had to be referred to by theoretical terms. Whether this is true is partly a question of definition. If any set of propositions— such as (1) to (13) in MG—that plays the role of introducing a term can be called a “theory," then all names for things we can't observe will be theoretical terms, by definition. But if we restrict the word “theory" to relatively complex sets of propositions, or to propositions that we still regard as speculative, then some terms for unobservables won't be theoretical. Until we developed manned space flight, for example, we couldn't see the other side of the moon. Some astronomer might have introduced the term “Moonback Mountains” to refer to mountains on the other side of the moon. Thus “There is a Moonback Mountain” would be explained by a simple Ramsey-sentence:MM: There exists an X such that X is a mountain on the other side of the moon.
Moonback Mountains would have been unobservable, but in one sense, their name wouldn't have been terribly theoretical. Normally, we call a term “theoretical” only when the sentence by which we introduce it is complex or hypothetical. Is it a theory that the large circular source of light that we see in the sky is a large heavenly body that radiates light? If it is, “sun” is a theoretical term. If it isn't, “sun” isn't a theoretical term. It's as simple as that.
The issue is complicated by the fact that as we get used to theories we are less and less aware that they are theories at all. When the earliest astronomers first proposed that the little yellow disk in the daytime sky was a large spherical object, this was a theory. But gradually, over time, it has become part of common sense. Every child (in our society) learns that the sun is a large three-dimensional body and not just a disk in the sky.
The point is that even commonsense beliefs often were once new theories. Indeed, philosophers of science have tended to argue that common sense on any particular matter is just another theory.
If we don't call the view that the sun is a heavenly body a “theory,” it is because we are not aware of the fact that this was once an exciting and original discovery. In ordinary life, we tend to use the word “theory” to refer to claims that we are still unsure about or that we know we were once unsure about. We tend not to use it for beliefs that we have come to take for granted. In this usage, the distinction between theoretical and nontheoretical terms belongs to the context of discovery: it has to do less with how we came by the terms and more with how secure we have become in our use of them. But philosophers use the word “theory” to mean any set of beliefs about how the world is, even if those beliefs are relatively simple or obvious or familiar. The point about a theory is that it is a set of propositions that might or might not be true. The way philosophers think about the question, whether something is a theory is an issue about the context of justification.Even if the question whether all terms for unobservable entities are theoretical is partly a definitional question, however, there is no doubt at all that some highly theoretical terms refer to things that are perfectly observable. The term “electron microscope” describes a perfectly observable thing. You can observe one in many biology laboratories. But it is certainly a theoretical term. It can be understood only by way of a theory about electrons.
Many philosophers of science, especially since the logical positivists, assumed that all unobservable entities are referred to by theoretical terms, and all theoretical terms refer to unobservable entities. You can see why they might have been led to think this. If we are to refer to unobservable entities, we have to introduce them by way of sentences such as MM. Because of the way philosophers use the word “theory,” they would say that it is a theory that there were mountains on the other side of the moon.
That makes “Moonback Mountain” a theoretical term. You should keep in mind that, on this usage, when I say that a term is theoretical I do not mean that it can be understood only in terms of an elaborate or complicated theory.Because of this, the connection between the question whether a belief is theoretical and the epistemological concern that our beliefs be based on observation is not a simple one. It was simply a mistake to suppose that because a term was introduced by way of a theory the thing it referred to could not be observed. This mistake shows up in the case of Mendel's theory. Though Mendel couldn't see genes, when light microscopes and staining techniques improved in the early twentieth century, geneticists came to believe that they could see them. It turned out that some genes (in the salivary glands of fruit flies, for example) were much bigger than others and could be stained so as to reflect light under a microscope. They looked like colored bands on the chromosome. This didn't make the term “gene” any less theoretical, but it did make genes observable.
Philosophers call things that we can observe phenomena. A phenomenon is something like a phenotype (which, as you may have guessed, shares with the word “phenomenon” a Greek root meaning “show” or “appear”). A phenomenon is something you can experience with your senses. As far as Mendel was concerned, the claims he made about genes were not just about phenomena, they were about unobservable reality.
Nevertheless, there is an important connection between theoretical terms and observability. As I said just now, if there were no connection between a theoretical term and observable things, we would have no way of using it to refer to things in the world. As the empiricists (whom we discussed in Chapter 2) argued, it is only through experience that we can justify our beliefs about the world.
When I began my discussion of empiricism in Chapter 2, I said that its rise came along with the rise of science.
Because of this, empiricism has often been the unofficial philosophy of scientists. One of the reasons that philosophers of science have insisted on a connection between theoretical terms and the observable world is that they have mostly been empiricists who were impressed by the considerations that led to the development of foundationalist epistemologies. You will remember that I also said in Chapter 2 that foundationalist epistemologists insista) that we must find some class of beliefs of which we have secure knowledge; and
b) that once we find this class, we can then honor some of our other beliefs with the special status of knowledge by showing that they are properly supported by the members of this class of foundational beliefs.
For most traditional empiricists, the foundational class of beliefs encompasses beliefs about the observable world, expressed in observational terms. That is why it is important for empiricists that we can introduce those theoretical terms that refer to things we cannot observe by way of Ramsey-sentences that connect them with objects and properties that we can observe. For then we have some prospect of being able to justify our theoretical beliefs by reference to observation, in exactly the way empiricism requires, even if our theoretical terms refer to unobservable entities. Connecting theoretical terms with observation offers empiricists the prospect that science can lead to genuine knowledge.
4.5