B Kinds of Science: Subject Matter

Thinkers have long been divided into “lumpers” and “splitters,” people who find similarities and group things together and people who focus on distinctions and divide things into categories.²¹ Lumping sciences together by identifying com­monalities or relationships, as we've been doing, only gets us so far; we'll have to do some splitting if we are to understand the roles of the hypothesis in sci­ence.

In the next section, I'll review a few of the conventional ways in which science has been split up and argue that they are not the primary divisions that we need to make; I'll get to more fundamental ones in Section 4.C.

4. B.1 Hard Versus Soft

Hard and soft sciences (e.g., physics and chemistry versus psychology and soci­ology, respectively) differ in the kinds of objects they investigate, the degrees of quantitative rigor they demand, and the predictive power of theories they pro­duce. At the same time their ideals and goals of achieving rational understanding of the world are largely the same. Scientists in hard and soft fields rely on the

Scientific Method, formulate and test hypotheses, strive for objectivity and uni­versality in their theories, and are open to correction and rejection. The hard- soft split, though common and sometimes convenient, is not fundamental.

4. B.2 Historical Versus Ahistorical

Can you do a properly controlled experiment where you have great influence over the experimental conditions? If so, you’re doing ahistorical science. Ahistorical sciences include all the laboratory sciences plus a few more (e.g., biology and botany). If you have to deal with the universe as it is and cannot manipulate the variables you’d most like to manipulate, then you are probably doing historical science (e.g., evolutionary biology, astronomy, geology).

Still, the similarities of background information, values, and methods of evo­lutionary biologists and geologists are greater than their differences. The famous study of the extinction of the dinosaurs is a good example. The father-son team of scientists, Walter and Luis Alvarez, hypothesized (the Alvarez Hypothesis) that a large asteroid struck the earth about 65 million years ago²² and its impact triggered cataclysmic environmental events, including an earth-enveloping dust cloud that blocked the sunlight for several years, as well as fires, acid rain, and increased vulcanism, all of which combined to cause worldwide extinctions. This is clearly historical science.

Despite the antiquity of the event that the Alvarez Hypothesis tried to ex­plain, the hypothesis was consistent with known geological data, and it made predictions that could be tested by future observations. In other words, the mere fact that historical sciences study past events doesn’t mean they can’t test hypotheses. For instance, the Alvarez Hypothesis accounted for the thin layer of iridium that is found everywhere in the world—iridium is an element in much higher concentrations in asteroids than is typical on earth—and for the presence of tectites—small, rounded, glassy rocks—associated with the iridium. The geological age of these phenomena was roughly in agreement with the onset date of the known extinctions. Most dramatically, the hypothesis predicted that there should be a huge asteroid impact crater somewhere on earth. Eventually a crater having the predicted age and size—110 miles across, 12 miles deep—was found near Chicxulub in the Yucatan peninsula of Mexico. The reasoning and procedures in historical science and ahistorical laboratory science are so much alike that it makes little sense to classify them as different kinds of science.

4. B.3 Natural Versus Social

For some authors there is an unbridgeable gap between natural and social sci­ences. Indeed, an extreme point of view is that social sciences are not, and should not to aspire to be, scientific in the way that natural sciences are. Under this in­terpretation, the subject matter of social sciences—society or culture—should be interpreted as a “text” analogous with the study of literature rather than rigor­ously interrogated for empirical scientific truths.²³ Many social scientists would not go so far; nevertheless, they would argue that interpretations in their fields must be conducted with qualitative methods that are necessarily less objective than the quantitative methods of the natural sciences. And indeed, it does seem reasonable to conclude that, if there is no way of making or desire to make social studies scientifically objective, then these studies truly have little or nothing in common with natural sciences. But then, what would be the purpose of calling them “sciences” in the first place? Why not group them with other careful, schol­arly inquiries such as art, music, or literature that are not considered sciences?

On the other hand, many social scientists do embrace the ideals of modern science, including hypothesis testing and the standard of, at least, “weak objec- tivity.”²⁴ And, increasingly, social scientists are adopting Big Data analyses and Bayesian methods to propose and test hypotheses (example in 4.E.2). On what grounds could we deny that these social scientists are engaged in real science?

In summary, although intuitively appealing, traditional divisions among kinds of science won't provide much guidance in following the debates about sci­ence that this book is concerned with. Instead, I suggest that an alternative set of distinctions will be more helpful.

4.