of science who is trying to find the universal norms that govern science, can be learned from how scientists discover their hypotheses.
However, the context of justification is a different matter. Once we have an explanation, in the form of a hypothesis or a more encompassing theory, we can legitimately ask whether or not we are justified in accepting that hypothesis or theory. On most accounts, a theory is rationally acceptable if it gets the data right for which it was invented, and also new data that hasn’t been observed so far. In other words, we see what predictions the theory makes about the world and then check whether those predictions come to pass. If they do, good; if not, bad. How to spell out “good” and “bad” in this context is the subject of the next main section on theory confirmation.
Before we turn to those issues, we have to consider some complications in the role we assigned to observations so far, because especially in light of the earlier example, it makes it seem that what we decide to investigate is pretty unguided. In scientific practice, however, those things worthy of our attention are often suggested by already accepted theories. For example, in 1827, the Scottish botanist Robert Brown noticed that small particles suspended in a liquid were in constant, erratic motion, which could not be explained by any currents. This was noteworthy, because of the background assumption that things don’t move without a cause. It was only in 1905 that Einstein explained the motion as the effect of molecules colliding with the particles. On the other hand, there are many facts not worth knowing about, as, for example, the number of blades of grass bent to the left by at most 33 degrees on Saturday afternoons on north-facing slopes in cities whose name has an “l” as their third letter when translated into English. And there is also the fact of the number of such blades of grass bent by at most 34 degrees, or 34.5 degrees, or on Saturday mornings, nights, etc. Clearly, there is an unbounded number of facts in the world, all of which are potentially interesting to someone. Given this unbounded number of facts, it is a good thing that our observational interests are guided and constrained by theories we already accept. Otherwise, we would be wasting a lot of time and other resources trying to explain facts that are not particularly important for understanding the world around us.
However, this selection of what is worth observing by referring to our already accepted stock of theories has a somewhat surprising consequence. If it is true that what we deem worthy of observing today is constrained by theories we accept, and if those theories were based on observations deemed worthy of making, which in turn were constrained by prior theories, and so on, all the back to the “beginning of science,” it becomes quickly clear that our current theories are constrained by early observations through the historical trajectory of theories initiated by those observations. In other words, we have to acknowledge the possibility that what seems scientifically interesting today has been strongly shaped by the “starting point” of science. Had humanity been noticing things other than what it did notice, the scientific trajectory may have looked very different with the result that our current science might be different as well.
This thought may have consequences for the debate about scientific realism, which is the view that our best theories are at least approximately true. Even if this is so, it might be of little comfort for our belief that sciences provides us with a deep understanding of the world. We may well have true theories, but they may be true about facts that are as irrelevant to a real understanding of the world as are bent blades of grass.
3.2 Theory Appraisal
Suppose you found a theory which you consider to be pretty convincing. Others don’t agree. What can you do? One thing to do is of course to show that the theory accounts for all the data – evidence – that are relevant to it. In particular, you are in good shape if you can show that your theory correctly predicts data that have not been collected yet. For example, suppose that a theory explains reductions in the Earth’s temperature in terms of periods of volcanic activity during which emitted material reflects solar energy. Suppose further that in 1979, a scientist predicted that after an eruption of 0.5 to 1.0 cubic miles of ejecta, global temperatures would drop by 0.1 degrees Celsius. And then came the eruption of Mount St. Helens which spewed 0.7 cubic miles of material into the atmosphere – and the Earth cooled by 0.1 degrees. Well, the theory would be looking pretty good! How exactly does this process work – what is its logic, as it were?
3.2.1 Confirmation through Predictive Success
One influential proposal for how to understand this process is Carl Gustav Hempel’s model of theory confirmation. It basically reduces confirmation to a two-step process. First, a scientist generates testable predictions from a theory. Second, the researcher determines whether or not those predictions are borne out by appropriate observations, which of course include measurement results and experimental outcomes. For Hempel, the first step was a matter of deductive logic. Here’s an example.
Suppose that you want to defend the theory that the earth is spherical. The ancient philosopher Aristotle was one of the first to describe a test for this theory. The idea behind his attempt at confirmation is very simple. If the earth were spherical, we’d expect to see a ship that leaves the harbor to not only look smaller and smaller as it sails away, but to also disappear nonuniformly over the horizon. At the appropriate distance, we might be able to still make out the masts and the sails without seeing the hull of the ship anymore. So, the theory “The earth is spherical” predicts a partial disappearance of the ship. If you go to the harbor, you can actually make this observation – the ship disappears from the bottom-up, and the theory has been confirmed by this observation.
Hempel summarized and generalized this example, and others, into the following model of theory confirmation:
Theory predicts Observation
Observation is made
Theory is (better) confirmed
The first thing to notice here is that confirmation is different from proof. If a theory entails an observation and the observation is made, it doesn’t logically follow that the theory is true. To think otherwise would be to commit a logical fallacy. To see this, consider: “If it rains, the streets are wet. The streets are wet. Thus, it is raining.” But the conclusion doesn’t follow from the premises, because someone could have hosed down the streets, and that’s why they are wet. However, if the prediction is borne out, we can say that the theory has been confirmed to a certain degree. This in turn means that we have some reason for believing the theory.
Here’s another example following the same pattern. Suppose I want to confirm the hypothesis that all ravens are black. How would I do this? Following Hempel’s advice, I infer a prediction from my hypothesis and then determine whether it is true or not. What follows from the claim that all ravens are black? One thing that comes to mind immediately is that the next raven I see will be black. Thus,
Raven Theory predicts “Next raven is black”
Observation “Next raven is black” is made
Raven Theory is (better) confirmed
What’s important about this second example is that the theory talks about all ravens – past, present, and future, and everywhere they exist. Suppose I have seen 200 ravens, and all of them have been black. Thus, I inductively infer my Black Raven theory. Now I want to confirm it by using Hempel’s procedure, as we have just done. But, seriously, how much of a confirmation can that one black raven provide? Remember, my theory talks about all ravens.
3.2.2 Falsification to the Rescue
This problem led some philosophers to abandon any attempt to confirm theories and to look at the situation from a different perspective. The Austrian philosopher Karl Popper introduced the twin notions of falsification and corroboration to describe how we (should) go about selecting the theories we accept. The idea is fairly straightforward and starts with the observation that while we cannot conclusively confirm any theory, we can conclusively falsify a theory. Take our example involving the color of ravens. Perhaps 200 ravens might be good evidence for the claim that all ravens are black; 2000 black ravens would be even better evidence. However,