I’ve not been commenting on the comments on this post about the Myers – Rue debate, but I have been reading them with great interest. The following, while not addressing most of the comments, arises from them.
In this post, I mentioned Loyal Rue’s linear hierarchy of … I’m not sure what he called it … let’s say certainty (in science). It ran something like this:
This is not exactly what he use, but close. Perhaps conjecture and speculation were reversed… but you get the idea. You think of something that might be true (speculate) and formulate it into a hypothesis, test it. Then, if you’ve tested it enough it moves to theory because it is reasonable to accept as likely true. Eventually such ideas move to the status of law, etc.
There are some ways in which this sort of linear structure is useful when describing scientific thinking, but I think most people who think about thinking (in this area) … in other words, philosophers, would see the linear arrangement, with the assumption that more work moves you up in line, glosses important aspects of the process of seeking and gaining knowledge.
For instance, these days, we see the formation of “laws” only infrequently, probably for a couple of reasons. One I think often overlooked reason is that all the good laws are taken. Laws that describe basic physical relationship (like the various laws related to fluids and gases, or that are mathematical models that accurately model physical processes (like the formula for acceleration) were worked out in the pre-nuclear days in physics. The other reason is that we are cautious about forming laws about more nuanced or organic processes. Cope’s Law, for instance, which is that body size increases over time in a given lineage, has been relegated to the status of Rule partly because it is not law-like (it does not always happen) and partly because it may engender teleological thinking.
So, if laws are not always desirable, then the transition from theory to law is not usually going to happen.
Another problem with this hierarchy, I think, is that while a theory can be as simple and stripped down as a hypothesis, it is difficult for a hypothesis to be as complex as some theories. At the same time, a hypothesis can be a hypothesis and be totally wrong, or it can be dead on correct. The key feature of a hypothesis is that it is testable. A theory may not be testable in the same way that you cannot assay the health of your automobile by checking the oil … you also have to kick the tires, bounce up and down on the shocks, and drive it around the block and listen to the engine, and each of these tests is going to contribute to your sense of quality of the car in a different, complex way. Theories tend to be like that … complex and heterogeneous in how they are supported, measured, described, and applied. Also, one seeks to have theories that are well supported and thus have reliable predictive power, but one actually seeks to have hypotheses that are clearly wrong (when tested).
Larson, commenting on my earlier post, also gives examples of increasing knowledge moving things down the hierarchy rather than up. This is pretty much what I had in mind.
Another problem with the simplified model is this: Rue suggested, maybe reasonably, that physics is the science that has the closest relationship between the scientific method (which is what this hierarchy is all about) and establishment of something we could call “truth” or at least, “knowledge.” Other sciences were less clean in this regard, perhaps in a sequence like this: Physics; chemistry; biology; psychology; anthropology. (One wonders where to put geology … or, egads, meteorology!) This may be true, and again, this may be a useful heuristic for understanding the relationship between various disciplines and theory-building. But it may also be true that the sciences over on the right hand side are fundamentally different than those on the left hand side. Gertz talked about “rich description” as a methodology in anthropology and other social sciences, for instance. Others have spoken of experiential investigation. It may be that there are areas of anthropology where science simply does not work, but rather, something more akin to what one finds in the humanities as appropriate.
I want to end with an example from human evolution of the quirky relationship between “theories,” “hypotheses” etc. The rise of bipedalism is seen as a major change in human evolution that must have been a big deal, and must have huge consequences. Therefore, there has been a great deal of attention to this adaptation, it’s morphology, and the reasons for it happening. My first grad school advisor, Glynn Isaac used to deal with bipedalism in an interesting way that was meant to be instructive of how science can work. The idea is to make a list of all the possible explanations for bipedalism in early hominids. Each explanation counts as a possible cause for this adaptation to arise, although of course, a given explanation may relate more to what bipedalism eventually became good for rather than why it arose to start with. But one must begin with something.
Now, you take this list and keep it handy, and every time you come up with a new possible explanation, you add it to the list. In the mean time, you attack each item on the list by formulating hypotheses, and testing the hypotheses.
Now, this immediately provides a critique of the linear hierarchical model. Bipedalism is an observation, a fact. It sits at the top of this hierarchy as an undeniable reality … early hominids were bipedal. But the list of explanations sit near the bottom of the list, ranging from absurdities to conjectures or speculations. But in order to form a hypothesis form one of these lower level assertions, one must have even a vague theoretical framework. For instance, one can look at bipedalism or positional behavior or stature in relation to other organisms. One idea is that a chimp sized organism has to get its head either above or below the zone in the savanna just at the top of the grass canopy, where it can be so hot as to be fatal during afternoon hours. Chimp heads, in their normal quadrupedal stance, would be right in this zone. So the chimp like ancestors of the hominids would have to either scoot around with their face in the dirt, or stand up. Why do we say this? We have an overarching theoretical construct regarding the physiology of cooling (of the brain) in mammals, a lot of data, a couple of natural “laws” (of thermodynamics of both animals and grasslands) all tied up in this theoretical structure. From this structure, we formulate a number of different hypotheses, for example, that bipedal hominids are typically found associated with proxyindicators of grassland environments, and so on.
Two interesting things happen here. One is that we are dancing around the “hypothesis” level of the linear hierarchy in interesting and complex ways, using theories to reshape conjectures into hypotheses, etc. The other thing that happens, in this case, is that we cannot eliminate the conjecture, nor can we strengthen it. It turns out that many (but not all) of the ideas proposed as explanations for bipedalism are not easily disposed of. The list gets longer when we add creative speculations, it gets shorter when we falsify enough hypotheses linked to these speculations (thus tossing out the speculation). But what happens very quickly is that you end up with a list of about six or seven ideas that you can’t get rid of. And it stays that way for decades and decades.
One possible conclusion from this is that the question of the origin of bipedalism is not answerable by science. The best we can do is to come up with a few interesting stories. This may be evidence that certain aspects of palaeoanthropology are not amenable to this sort of scientific knowledge or fact seeking. Another idea, that I prefer, is this: If bipedalism cannot be explained after decades of research, then there is a certain chance that bipedalism is either a) not real or b) not important.