单项选择题

TEXT D
Theories are like apples; there are good ones and there are bad ones. A good apple is flavorful and nutritious. A good theory is one that we have good reason to believe is true, or at least likely to be true. Apples have all sorts of features that are indicative of goodness and that can be used at the store for making smart choices. An apple’s color, shininess, shape, and firmness are related to its flavor and food value. Similarly, theories have features that are indicative of their truth, and the task of justification is to identify these features and use them to guide choices as to which theories to believe. In this way, responsible theory choice and responsible science are not unlike responsible shopping. The big difference is that you get to go home and eat the apple and find out if you were right and it really does taste good. With theories, however, there is no feature that is the taste of truth, at least not that we get to sample. There are only features that indicate truth. You never get to go home and see unambiguously that atomic theory, for example, is true. But you do the best you can.
Theories have lots of different properties and can differ one from another in a variety of different ways. A few have been mentioned already, namely the degree of generality and hypothetical status of a theoretical claim. An account of the properties of a theory is an answer to the inquiry, "So tell me what this theory is like." There are plenty of things to say, that is, plenty of features to mention. There is the feature of how it was discovered, by whom, at what time of day. Some theories have the feature of being proposed by Einstein; some lack his feature. Some have the feature of being liked by Einstein; some don’t. There are also features of the content of a theory or of its form. Being logically consistent is a feature of form. Making reference to action at a distance, or to evil spirits, or to gluons are distinct features of content. There are even pragmatic features. Some theories are likely to be money-makers; others are not.
Clearly, some of these features of theories are irrelevant to an analysis of scientific knowledge and irrelevant to doing science. We want to focus on the important features, those that meet the following two criteria. First, they must be relevant to the likelihood of the theory being true. That is, they must be reliable indicators of truth so that our seeing whether the feature is present or not will be part of our warrant for believing the theory. Some theories have the property that you learned them on a Tuesday, but that is unimportant because it is neutral to their being true or false. Some theories have the property of describing goblin maintenance men to keep the world in good working order. That’s an important feature because it helps us decide whether the theory is true or false.
The second criterion of importance is that the property is something we can in fact evaluate. The information must be accessible to us if it is to be of any use. To get back to apples for a moment, even if it’s true that apples with darker seeds taste better, this isn’t any help there in the store where information about the seeds isn’t available. Similarly, a theory about the demise of dinosaurs is true if it is an accurate picture of the past events, but the feature of being an accurate picture is not something we can evaluate. We can’t see what this theory is a picture of, and so we can’t assess its accuracy. Much as we would like to use accuracy-of-picture or correspondence-to-facts as features that indicate truth, information about these features is unavailable. Features that are important to the justification of theories must be both truth-conducive and accessible.
It can be inferred from this passage that ______.