The received view of theories

Empiricists, then, place great importance on the thesis that the foundational class of beliefs, the class that justifies all our knowl­edge, is the class of observational beliefs.

a) Is it observable?

and

b) Do we use observational terms to refer to it?

Observability is an attribute of things and properties, not of terms. So empiricists need to give a definition of observational terms.

The obvious way to do this is to say that a term is observational if we can tell whether it applies simply by observation, without relying on any theory. Thus, “red” is an observational term because we can tell whether something is red just by looking, and “loud” is an observa­tional term because we can tell whether a sound is loud just by listen­ing. The reason “electron microscope” isn't an observational term is not that we cannot observe electron microscopes. Rather, it is that when we look at a piece of apparatus, we need some theory to inter­pret what we see and allow us to tell whether it is an electron micro­scope or not. To tell whether something is an electron microscope, you have to be able to find out whether it forms an image of an object by reflecting electrons to a detector, and to do this requires a good deal of theoretical knowledge.

I shall return to this issue again in the next section. For the moment, I am going to assume that we can make a distinction between observational terms, which we apply by using our senses alone, and theoretical terms, which we apply on the basis of obser­vations as interpreted by theory.

Given a distinction between theoretical and observational terms, we can divide all the terms in a theory into three, for along with observational terms and theoretical terms, we shall need logical terms, such as the connectives, the quantifiers, and, as we shall see, the modal terms “necessary” and “possible.” With these three kinds of terms we can build our theories; the logical positivists (who called themselves, you remember, “consistent empiricists”) developed an account of the structure of scientific theories that was based on these distinctions. That model has been so influential that Hilary Putnam, an American philosopher, once called it the “received view” of theories.

size=2 color=black face="Times New Roman">On the received view, a theory is stated in a language that con­tains, along with the logical terms, a vocabulary of observational terms and of theoretical terms. The observation language consists of sentences containing only observational and logical terms. The theoretical language contains only theoretical terms and logical terms. There will also be mixed sentences, containing both theo­retical and observational terms along with logical ones.

The theory itself will contain two parts. One part, the theoretical postulates, will be stated entirely in the theoretical language and will describe the relations between the entities and properties that the theory postulates.

We can see how this model works in the case of MG. The theo­retical postulates of MG will include, as we saw earlier:

8)     An allele, A, must exist in one of three relations to any other allele, B. A can either

a) be dominant with respect to B, or

b)          be recessive with respect to B, or

c)          interact with B.

This proposition certainly is not one we can confirm simply by direct observation. To connect it with observation, we have to include cor­respondence rules such as:

9)     If A is dominant with respect to B, then an organism that is heterozygous and has the genotype AB will have the A phe­notype.

The following two correspondence rules, (10) and (11), will also be important if we want to apply (8).

10)   If A is recessive with respect to A*, then an organism that is heterozygous and has the genotype AA* will have the A* phenotype.

11)   If A interacts with A*, then an organism that is heterozy­gous and has the genotype AA* will have neither the A nor the A* phenotype, but some other phenotype that is deter­mined by A and A* together.

But even these will not be enough, by themselves, to apply the the­ory in any particular case. To do that, we would need to replace the variables “A” and “B” with the names of specific genes and pheno­types.

Flower color in peas is determined by a gene that has alleles R, W, and P, which produce red, white, and purple flowers in the heterozygous plant. R and W interact to produce pink flowers. P is dominant with respect to W ...

and so on. Correspondence rules such as these connect the theoret­ical postulates with observation and make it possible to see what the theory says will happen in particular cases.

style='font-size:10.0pt;line-height: 112%'>The empiricist philosophers of science who developed the received view spent a great deal of effort trying to characterize the structure and functioning of theories. They did this because they were concerned with the epistemological problem of how we know about entities—like genes—that we cannot experience with our unaided senses. But they were interested in theories for another reason. Theories are one of science's most distinctive products. Of course, science has other important products as well. Airplanes and antibiotics, barometers and bazookas, cars and computers—the whole alphabet of modern technology depends for its development on the work of scientists. But we could imagine a (rather strange!) culture that pursued scientific research without much interest in its technological possibilities. What seems impossible is to conceive of science without theory. The development of theories about how dif­ferent parts of the world work is what science is for. If you don't want scientific theories, you don't want science.

To understand how theories work is to understand a large part of what science is about. But why do scientists want to construct theo­ries? What are they for?

One empiricist answer to this question is that we want theories in order to make reliable predictions.

Now, most people would say that the reason that Mendel's theory enables us to make these predictions is that it is true. There really are genes with alleles, which are transferred from parents to off­spring. The reason that Mendelian genetics gets predictions of flower colors right is that it is part of the correct explanation of how flowers get their colors.

This view of theories is called the realist interpretation of theo­ries. It says that the entities the theory talks about are real and the theoretical postulates and the correspondence rules of a good the­ory are as true as the sentences of the observation language. Of course, we can't observe the theoretical entities directly, so it is harder to get to know about them than it is to get to know about observable things. But because we have the correspondence rules that connect the theory with observation, we can find out about the­oretical entities in an indirect way. After all, doesn't the fact that Mendel's theory allowed us to predict the outcome of breeding experiments entitle us to think that genes exist? Or, to put the ques­tion another way, doesn't the success of Mendelian predictions give us reason to think that his theory provides the right explanation of how inheritance works, which requires the existence of the entities it postulates?

The close connection between successful prediction and expla­nation has led to the received account of how theoretical explana­tion works in science. This account of explanation starts from the received view of theories. It's called the deductive-nomological model of explanation, or the “DN model” for short; and it was developed by another member of the school of logical positivism, Carl Hempel.

4.6