Form in Explanation

Return to Critical Thinking Reference

Return to Instructional Notes on Method from the Philosophy Proseminar

Introduction

What does a famous philosopher like Daniel Dennett and thousands of campfire and scouting kids have in common? They were all impressed by the following song when they were kids:

Tell me why the stars do shine

Tell me why the ivy twines

Tell me why the sky’s so blue

And I will tell you just why I love you

Because God made the stars to shine

Because God made the ivy twine

Because God made the sky so blue

Because God made you, that’s why I love you

One of the most poignant signs of the development of the reflective attitude in children is the asking of “Why?” questions. Why do the stars shine? Why is the sky blue? Why are we here?

Recall that explanations are rationales in which causes are offered to account for facts that we observe. Unlike arguments, explanations have conclusions that are not in doubt. We do not doubt that the stars shine or the sky is blue when we ask for an explanation of these facts. What we want in an explanation is an answer to the question “Why?” What counts as a good answer?

To see the difficulty of answering that “What” question, consider the tremendous variety of “Why” questions:

1. Why is Mars red?
2. Why do I feel the way I do?
3. Why do $2000 bicycles weigh so little?
4. Why does toast seem to land butter side down when you drop it?
5. Why does the AIDS virus resist drug treatments?
6. Why are humans so aggressive?
7. Why did my friend treat me rudely the other day?
8. Why do humans laugh?
9. Why are some forms of poverty difficult to alleviate?

We ask for explanations about the mundane and the cosmic, about events in our personal life and the life of our societies, about the present, past, and future. We can sometimes offer pretty reliable explanations with very little evidence and other times we can be drowning in data and not have even a promising hypothesis.

Explanation in Everyday Life, Science, and Philosophy

In order to see how explanations are structured, we should look at two examples, one drawn from everyday life and the other from scientific explanation. We want to be able to see explanatory rationales in any context in which they arise. By noticing the similarities and differences between these two cases, you can begin to see the general structure of explanations.

Saturday Morning

Kylie: So, Andy, where were you last night. I’ve been trying to call you. Why didn’t you have your cell phone on?

Andy: Umm, well, I wanted to tell you about that. . . The thing is, um, I forgot to charge it up.

Kylie: It’s just that we usually meet up on Friday after we are done with classes. I went to your room when I didn’t see you, but no one told me where you were. We’ve been seeing each other for three months now. I’m just surprised that’s all.

Andy: Well, I had to meet a friend who came into town.

Kylie: That’s funny. Later that night I ran into Beth who said you were with some woman who she didn’t recognize at Shenanigan’s Sports Bar. You hate sports bars. Why did you go there? Who were you with?

Andy: Oh! Beth saw me there huh? Well, that’s what I needed to talk to you about. I mean, . . . (stammering). . . I wanted to call you and all, but I guess the time just got away from me. That was my cousin, Darla, She surprised me by coming here from the West side. She goes to school there, but I haven’t seen her for a while. She said she wanted to see her school’s team play and we had to find a sports bar showing it. I wasn’t trying to avoid you, but I was pretty sure you wouldn’t want to go there. It’s awfully loud and smoky.

Kylie: Your cousin? Really? You’ve never mentioned her. Beth said you two were having a pretty good time. It sure didn’t occur to her to think you were with a cousin based on what she saw. I’m glad I didn’t get in the way of your evening. To tell you the truth, Andy, I’m not sure I buy your explanation. You don’t sound so sure of it yourself.

Andy: Well, it probably looked pretty suspicious. I mean, Darla dresses kind of differently. She really likes to show off a bit. And she’s always been pretty friendly. Beth probably made some reasonable inferences, but they don’t really explain what’s going on. I might have been a little embarrassed of bringing her around you, but I probably should have told her I needed to find you first. Besides, you’re going to meet her when you come to my folks’ house for Thanksgiving next month. My Aunt and Uncle and Darla always come over.

Kylie: Well, maybe things aren’t what they appeared to be. I’ll look forward to meeting her at your parents’ house.

How were you assessing the explanations in this dialogue as you read it? Did you arrive at the same conclusion as Kylie before Andy’s last speech? Along the way, did you doubt Andy’s explanation for not finding his girlfriend Friday afternoon? What about cousin Darla? Did Andy’s explanation seem remotely plausible? If so, at what point?

If you were as suspicious as Kylie, you certainly had good reason. Andy was stammering, he didn’t volunteer much information and his behavior fell out of the normal pattern Kylie had become used to. Something needed explaining. The first thing to notice about explanations is that they arise out of real doubt about how something works, comes about, or fits together. Nothing needs explaining until someone starts wondering about things in one of these ways. So explanation starts with a psychological condition of doubt about some aspect of our experience or belief.

Andy's last speech probably relieved your suspicions. By the end of it, the most plausible explanation seemed to be that Andy is a somewhat awkward boyfriend who might have been completely innocent in going out with his cousin. He might have handled the situation better, but he doesn't seem so bad after all. But what made the difference? The fact that he had a plausible account of what Beth saw might have helped, but the fact that they are all getting together for Thanksgiving is probably the most important detail adding credibility to Andy's story. That detail adds something that the most ordinary everyday explanations share with the most sophisticated scientific experiments: a verifying experience. As soon as Andy mentioned that Kylie would meet Darla at Thanksgiving, his behavior on Friday and in the dialogue could be reinterpreted. Kylie might still be upset at Andy for his lack of sophistication, but she knows his story is going to have to check out. Unless his parents and other relatives are involved in a massive cover-up of his two-timing with Darla, the odds are that she's really his cousin. If you are a trusting sort, you probably just went along with this way of interpreting the story. Unfortunately, the truth is worse than the fiction. Andy was a rat. A week later he told Kylie that his parents got in a big fight and the Thanksgiving plans had changed. Kylie was pretty sure that was an excuse and, since he seemed aloof during that week, she decided to call him on it. He confessed to the whole thing.

If explanations usually start with doubt about how something works, came about, or fits together, they usually end with a verifying experience, or, in this case, the prospect of one. Verification, or testability, is not the same as proof, but under the right conditions it raises our confidence in the explanatory power of the account offered to such a degree that you just stop wondering about it. While we sometimes talk as though explanations begin with the search for truth and end with it's discovery, it is more precise to say that explanations begin with doubt and end with the cessation of doubt. Something could always reawaken doubt. When you stop doubting depends upon a lot of factors in the context of inquiry - how big a claim you are making, how urgent the need for action is, the level of evidentiary support, etc. Hopefully, we stop doubting when we have a reliable story in hand that we could test in experience. But whether our story is the absolute truth or not is much harder to say. In that sense, explanations are more like inductive arguments which also aim at a "degree of confidence" in their conclusion. Good explanations offer the most probable account of how something works, comes about, or fits together.

To see an example of a scientific explanation, consider the question. "Why is Mars Red?" The online magazine Slate has a regular column called "The Explainer" which ran a short summary of explanations on this question. [Slate, "The Explainer"] The puzzle leading to the question is that Mars appears to be red all over, in every bit of its surface. Since planets are made of lots of different substances, that calls for explanation. The short answer is that Mars is red because it is covered with iron oxide. Notice, however, that even if this is the truth, it is not very explanatory, because we immediately want to know, "Why is Mars covered with iron oxide?"

One explanation, which is now being challenged, is that the iron oxide comes from water interacting with iron in the planet's rocks. This requires us to believe that Mars was substantially covered in water early on in its formation. This may also explain the long, empty channels that cover large areas of the planet and which inspire bad Martian science fiction novels. Water suffused with iron deposits might have gotten into the hydro cycle, dispersing iron oxides in raindrops all over the planet.

Notice what this explanation has going for it. It suggests one mechanism to explain two things, the presence of iron oxide and the channels. It relies on a chemical process that we know about and it explains the dispersal of iron oxide around the planet. The weakness is, of course, that we don't know if there really was an abundance of water on Mars. So this explanation needs a verifying experience to confirm it. Otherwise it might be a case of explaining the "obscure with the obscure." If you try to explain something puzzling with something you don't have any clearer knowledge of, you really are not making progress.

A newer, alternative explanation is being offered by NASA scientist Albert Yen. Using some of the data from the 1997 Pathfinder mission, Yen noticed that Martian soil, not just the loose dust on top of it, is also full of iron, suggesting that impacts from meteorites could be responsible for the amount of iron on the planet.

The interesting thing about Yen's explanation is that it doesn't require the presence of water. In an experiment in the year 2000, he placed a small chunk of labradorite, a mineral commonly found in Martian soil, in a test tube, simulated the Martian atmosphere inside the test tube, and bombarded it with UV radiation, which is strong on Mars. After a week, he found the presence of "super oxide ions," which could cause iron oxidation without water. Yen's results also might explain why the earlier Viking Lander did not find evidence of organic materials on Mars - super oxides break down organic materials. The process Yen proposed may not work quickly enough to account for the abundance of iron oxide on Mars, so it could be that both processes were at work, one involving water and one without.

Neither of these explanations is sufficiently strong right now to resolve doubt about the process that produced the iron oxides that make Mars red. The more recent explanation has the virtue of not depending upon water and explaining the absence of organic materials, something that might have been expected if the first explanation were accepted. Together, the two explanations might be stronger than either one alone. To see just how "fluid" the situation is, the day after these pages were written NASA announced new evidence from the Mars rover "Opportunity" that the planet almost certainly had an abundance of water on it. The existence of water in Mars' past plus the absence of signs of organic hfe actually supports, weakly, the idea that both explanations might be right - the first because it requires water and works quickly and the second because it could have worked after Mars became dry and it breaks down organic materials.

Another interesting feature of the second explanation is its use of experimental simulation. If you want to posit physical processes that cannot be observed, in this case because they happened in the past, you might try simulating the process experimentally. The same thing could be done with psychological phenomena. You can test a hypothesis about how people react to stimuli or think about a problem by simulating the psychological mechanism you think is at work in the phenomenon and then testing it through controlled experiments.

"Why" vs. "How" and the problem of teleological explanation

If you talk to very many scientists about explanation and you tell them that explanatory questions are "Why?" questions, you are likely to be corrected now and then. Some scientists will tell you that explanations divide neatly into "Why" questions and "How" questions. Science doesn't answer "why" questions, it only answers "how" questions. Science is not trying to tell us why there are human beings on earth, only how they might have gotten here. If you want to know the answer to the question, "Why is there something rather than nothing?", you should go to a theologian or a philosopher, so a scientist once told me, but if you want to know how some of the relatively constant properties of the universe were determined in the first few fractions of a second after the big bang, you should talk to a cosmologist, which is, these days, much more likely to be a physicist than a theologian or philosopher.

Both "why" and "how" questions are clearly explanatory from a psychological point of view, but scientists who insist on the distinction between scientific and non-scientific explanation have a good point when they say that scientific explanations only have to do with "how" something works, came about, or fits together. Their worry is that we will make use of the concept of "telos" or purpose in a way that will prevent us from offering more compelling explanations that can only be offered if we focus on "how". First, we need to know more about what a "teleological" explanation is. Then we will come back to the scientists' point about "why" versus "how," hopefully seeing it in a new light.

"Teleological" simply means "having to do with goals and purposes". A teleological explanation is one which answers a "Why?" question by identifying the goal or purpose of the thing in question. If you had asked Aristotle, "Why do fruit trees have leaves?" he would have said that they grow leaves for the purpose of protecting their fruit (Physics 199a 26-9). Since the scientific revolution in Western Europe in the 17th century this would not have been regarded as scientifically explanatory. A scientist today can tell you how fruit trees (or any other tree for that matter) produce leaves. To be fair to Aristotle, he would have understood this answer perfectly well. He understood "how" answers as "efficient causes". But he also thought that all living reality had "final causes" or purposes. A very small number of philosophers still advocate a return to Aristotle's approach to causation, but that is another story!

While there is a solid consensus today among physical and life scientists that identifying the purpose or function of a living thing is not a basic explanation, there are still disputes about whether and under what circumstances "purpose talk" should be introduced into scientific practice. After all, if you are trying to understand how some organelle inside a cell works, it might be helpful to ask, at some stage of the investigation, what its function is. A biologist just would not stop there.

If we take the controversy over teleological explanation out of the arena of scientific discourse and put it into a larger cultural context, we might get an additional insight into the problem. The broader reasons why biologists, for example, may be concerned to avoid "teleology" have to do with the remarkable success of explanations in contemporary evolutionary biology that, since Darwin, have been opposed in the culture at large precisely because they offer compelling accounts of speciation and change in biological characteristics of organisms without making any reference to ultimate purposes. Scientists who distinguish carefully between how and why questions may simply be trying to focus on a scientific paradigm that has been tremendously productive since the scientific revolution 400 years ago. They may also be trying to police the borders of a cultural conflict which still flares up today in popular U. S. culture, just when more and more people are starting to wonder if there is any need for explanations to connect phenomena to ultimate purposes. Certainly there is no such need within the contemporary paradigm of the physical sciences.

Whether critics of that paradigm will be successful in convincing people to pursue another kind of science only time will tell. But many scientists are like most non-scientists in wanting ultimate answers (final causes) to questions about the purpose of their existence. They do not all expect answers to that question to come from science. Not surprisingly, the question, "Is there any need for explanations to connect to ultimate purposes?" is itself one about which we can engage in critical thinking and philosophical reflection. If you want to pursue this question you should start by talking to friends about it and noticing the reasons they give for feeling or not feeling the need for teleological explanations in their lives.

Seeing Form in Explanations

The practical skill of seeing form in explanations should start to become clear if you compare some of the features of our two examples -- Andy's cheating heart and the redness of Mars. When you are trying to decide whether to believe someone's explanation, a great deal depends upon how their story hangs together. Explanations are causal stories about the world. Stories cohere when the events are connected in a plausible way and, frequently, when the events unfold in a way that fits with other things we think we've explained well. Of course, we have to be ready to acknowledge when a new explanation upsets some of our background knowledge. In the case of Andy, his story was pretty shaky all along, based on the background expectations Kylie felt justified in having. The causal story should be reconstructed in terms of the psychological expectations that Andy failed to meet (calling, being in a bar with another woman) as well as the explanations he offers for "appearances" (his embarrassment about cousin Darla, for example). In the Mars explanation, the explanation has to link physical processes that we know, or conjecture, could have happened in the history of Mars' existence and it has to lead to the production of a lot of iron oxide. Competing explanations are judged on how well they accomplish this. As different as the underlying content of these example is, they are both explanatory rationales, so they both tell a connected causal story about how a particular situation came about, Andy's date with Darla or the redness of Mars.

When you reconstruct explanations, you should clearly identify the thing to be explained, the steps or links in the author's explanation. If causal principles or general regularities to appeal to, such as "Dating couples stay in touch about their weekend plans," or "The production of iron oxide usually involves water", these should be identified at specific points along the story. In light of your reconstruction, it should be clear why the author thinks that reasonable doubts have been satisfied about how the thing to be explained has come about.