Although it is helpful to appreciate the general nature of explanations, we might reasonably want more than this. As this book is part of the Understanding Life series, we may expect to delve into details about kinds of explanations that are specific to the life sciences.

It is widely held that science is a (if not the) primary source of our knowledge of the world around us. Further, most accept that scientific knowledge is the best confirmed and well-supported kind of knowledge that we have of the world. But, how do scientific explanations lead to scientific knowledge? The short answer is that they do so via a method known as “inference to the best explanation” (IBE), sometimes called “abduction.” Before we get into the details of IBE, let’s take a quick look at an obvious way that scientific explanations give us scientific knowledge.

A general way of appreciating some of the main ideas of the previous chapter is to recognize that explanations aim at providing understanding. Scientists and philosophers agree that understanding is a (if not the) primary epistemic goal of scientific inquiry. Both explanation and prediction tend to be closely related to understanding. We want explanations in science because we want to understand why the world is as it is and how things happen. And, once we understand various phenomena, we can make accurate predictions about them. One simple, and widely accepted, way of assessing the quality of a given explanation is to look at the understanding it provides. Roughly, the better an explanation, the more understanding that explanation (if true) would provide. As philosopher Peter Lipton explained, the explanation that is the best is simply the explanation that, if true, would provide the deepest understanding of the phenomena being explained. That being said, some worry because it seems that we might misjudge how well we understand something.

In this chapter we’re looking at the relation between scientific explanations and predictions. It is tempting to think that the only difference between explanations and predictions is that one looks back and tells us how or why things happened as they did, and the other looks forward and tells us how or why certain things will (or are likely to) happen. This thought can seem particularly plausible when we consider that in many cases a good scientific hypothesis will both explain phenomena and allow us to make accurate predictions. Despite its initial plausibility, the idea that explanation and prediction are symmetrical is mistaken. The way to see this is to take a look at a particular theory of scientific explanation that entails this relationship between explanation and prediction. The particular theory of scientific explanation in question, the covering law model, which we discussed in Chapter 2, is false. One of the reasons that this theory of explanation fails helps illustrate the fact that explanation and prediction are not symmetrical.

Explanation is central to our lives, in general. We seem to have an innate (or nearly so) drive to explain and seek explanations. When our favorite app is not working, we want to know why, and we want to know how to fix it. When trying to understand why people engage in an odd behavior – refusing to wear masks during the COVID-19 pandemic, say – we want an explanation. What reasons do they have for doing something that seems so clearly misguided? Why are they resistant to expert advice on the issue? Ultimately, we seek explanations to help us understand and navigate the world around us.

While it isn’t necessary to do so, it’s often good to start a book by saying something that is clearly true. So, let’s do that. Science has had (and continues to have) a significant impact upon our lives. This fact is undeniable. Science has revealed to us how different species arise, the causes of our world’s changing climate, many of the microphysical particles that constitute all matter, among many other things. Science has made possible technology that has put computing power that was almost unimaginable a few decades ago literally in the palms of our hands. A common smartphone today has more computing power than the computers that NASA used to put astronauts on the Moon in 1969! There are, of course, many additional ways in which science has solved various problems and penetrated previously mysterious phenomena. A natural question to ask at this point is: why discuss this? While we all (or at least the vast majority of us!) appreciate science and what it has accomplished for modern society, there remain – especially among portions of the general public – confusions about science, how it works and what it aims to achieve. The primary goal of this book is to help address some specific confusions about one key aspect of science: how it explains the world.

In the previous chapter we discussed the importance of accurate explanations. Without explanations that are in fact accurate we cannot have genuine understanding. In this chapter we will explore whether false scientific theories can be used to generate accurate scientific explanations. Before jumping into this, let’s first briefly recall the relationship between scientific theories and scientific explanations. Scientific theories consist of laws, models, and principles. Together these components of scientific theories offer broad generalizations about the nature of the world.

Even though explanation plays a central role in science, it is not enough to simply come up with explanations. Scientists (and everyone else) must also evaluate explanations. After all, it’s clear that not every explanation is a good one, as well as that some explanations are better than others. For example, evolutionary theory provides a much better explanation of the diversity of life than, say, the hypothesis that all organisms appeared at the same time in their present form. But what makes one explanation better than another? Relatedly, how can we tell which of a set of competing hypotheses provides the best explanation?

To the extent that we can make education a science, we will gain some power to predict future directions for educational improvements. This chapter begins with quotations from some famous people that indicate that in the past, we have not learned from our mistakes. If we can succeed in creating a viable science of education and apply this in all educational settings, we may change the course of history in a positive way. This chapter presents a critique of some of the things we have done, and a description of more promising alternatives.

The chapter begins with a description of the first chance experience that shaped the future of my career, a meeting with a former Cornell PhD student, Bruce Dunn, who was interested in collaborating on research and invited me to do a sabbatical leave at the University of West Florida in 1987-1988. This in turn led to conversation with Dunn’s friend, Kenneth Ford, a new faculty member interested in artificial intelligence. We found that the use of concept mapping was highly facilitated for capturing expert knowledge in a fashion that rendered the knowledge easily applied in artificial intelligence settings. Ford became the director of the Institute for Human and Machine Cognition (IHMC) and he invited his friend, Alberto Cañas, to serve as associate director and to lead a team to create computer software for making concept maps electronically. We soon had available to us software that would work on almost any computer and that would not only allow the construction of concept maps, but also permit attaching digital resources to any map that could be accessed by simply clicking on icons on individual concepts. The software suite created became known as CmapTools, and this software suite is now used all over the world in virtually every field where organized knowledge is important.

In part to illustrate the slow progress in secondary school facilities and programs, I introduce findings from a study done some 50 years ago. Most of the positive changes that occurred in the last 100 years are the result of an occasional creative administrator or school leader. To the best of my knowledge, none of these innovations were introduced on the basis of a comprehensive theory of education. I present evidence to suggest that this situation is changing.

The chapter begins by addressing the question: Why do young children learn so quickly? The short answer is that they are learning names for objects and events they are experiencing directly. These words are concept labels and they are engaged in what we call meaningful learning. In contrast, school learning is too often rote learning where the concepts and principles children are learning are not related to direct experiences with objects and events. David Ausubel’s cognitive psychology was introduced in 1963 and we immediately applied this new psychology as the foundation for all of our future work. We rejected totally the behavioral psychology that had dominated the field of education for some one hundred years. We also rejected positivist epistemology in favor of the emerging constructivist epistemology. It was not until the late 1980s that cognitive psychology and constructivist epistemology became widely adopted.

This chapter opens with the question: Can education become a science? I seek to answer to answer this question by asserting that education is a human activity and like any other human activity, it can be studied scientifically. This means that we can construct concepts, principles, and theories that explain how human beings acquire, use, and construct new knowledge. A comprehensive theory of education must address the question of the nature of knowledge and how human beings build new knowledge, and how to organize education to facilitate these processes. I argue that the major problem with education in the past has been the use of faulty theories of learning and invalid theories of knowledge and knowledge creation, resulting in inadequate instructional practices.

Behavioural studies aim to discover scientific truths. True facts should be replicable, meaning that the same conclusions are reached if the same data are analysed, if the same methods are applied to collect a new dataset and if different methodological approaches are used to address the same general hypothesis. The replication crisis refers to a widespread failure to replicate published findings in the biological and social sciences. The causes of the replication crisis include the presence of uncontrolled moderators of behaviour, low statistical power and dubious research practices. Various sources of information can help to distinguish good research from bad. An evidence pyramid ranks different study types according to the quality of evidence produced. The Open Science movement encourages replication, preregistration and transparency over materials, methods and data, all of which should improve the quality of science and the likelihood that findings will be replicated.