Skip to main content
Humanities LibreTexts

1.15: Explanatory Virtue and Truth

  • Page ID
  • \( \newcommand{\vecs}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \)

    \( \newcommand{\vecd}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash {#1}}} \)

    \( \newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\)

    ( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\)

    \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\)

    \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\)

    \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\)

    \( \newcommand{\Span}{\mathrm{span}}\)

    \( \newcommand{\id}{\mathrm{id}}\)

    \( \newcommand{\Span}{\mathrm{span}}\)

    \( \newcommand{\kernel}{\mathrm{null}\,}\)

    \( \newcommand{\range}{\mathrm{range}\,}\)

    \( \newcommand{\RealPart}{\mathrm{Re}}\)

    \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\)

    \( \newcommand{\Argument}{\mathrm{Arg}}\)

    \( \newcommand{\norm}[1]{\| #1 \|}\)

    \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\)

    \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\AA}{\unicode[.8,0]{x212B}}\)

    \( \newcommand{\vectorA}[1]{\vec{#1}}      % arrow\)

    \( \newcommand{\vectorAt}[1]{\vec{\text{#1}}}      % arrow\)

    \( \newcommand{\vectorB}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \)

    \( \newcommand{\vectorC}[1]{\textbf{#1}} \)

    \( \newcommand{\vectorD}[1]{\overrightarrow{#1}} \)

    \( \newcommand{\vectorDt}[1]{\overrightarrow{\text{#1}}} \)

    \( \newcommand{\vectE}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash{\mathbf {#1}}}} \)

    \( \newcommand{\vecs}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \)

    \( \newcommand{\vecd}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash {#1}}} \)

    Whenever we investigate anything—black holes or the causes of the First World War or the demography of the Cayman Islands or the ambiguity of Yeats’s poetry—our intrinsic goal is to find the truth about something. If we did not have that goal, we would not be inquiring.


    Two Huge Problems

    David H. Glass clearly articulates the two biggest challenges inference to the best explanation:

    Despite its intuitive plausibility, IBE faces two key challenges. First, how exactly is IBE to be understood and made precise? There are various conceptions of the nature of explanation, but assuming some of these are suitable for IBE this still leaves the question as to how one explanation should be compared against another so that the best explanation can be identified. Second, what is the connection between explanation and truth? Is there any reason for thinking that the best explanation is likely to be true? Or to put it another way, does IBE track truth? Of course, no approach should be expected to lead to the truth in every instance, but if IBE is to be accepted as a rational mode of inference, there must be some reason for thinking that it provides a good strategy for determining the truth.2

    Inference to the best narrative (IBN) inherits these same problems. How should one narrative be compared against another so that the best narrative can be identified? And is there any reason for thinking that the best narrative is likely to be true? Does it provide a good strategy for determining the truth? But inference to the best narrative invites a third challenge. Does it even make sense to talk about truth in contexts involving violent ex-husbands or constitutional success or failure? All three of these challenges must be addressed, if not definitively answered.

    Gilbert Harman foresaw Glass’s first challenge in his initial treatment of inference to the best explanation.

    There is, of course, a problem about how one is to judge that one hypothesis is sufficiently better than another hypothesis. Presumably such a judgment will be based on considerations such as which hypothesis is simpler, which is more plausible, which explains more, which is less ad hoc, and so forth. I do not wish to deny that there is a problem about explaining the exact nature of these considerations; I will not, however, say anything more about this problem.3

    One might ask why is there is any problem in the first place. Harman seems to answer his own question about explanatory virtue. The best explanation must be determined by the standards of simplicity, plausibility, completeness, and not being ad hoc. The superficial answer is obvious. His list of explanatory virtues is incomplete (“and so forth”), the virtues can work against one another—the simplest account may not be the most complete—and each one is vague and overly general. Just as with inference to the best explanation, we face the obvious question of what are the criteria for one narrative to be better than another. Here, I think Harman’s little checklist, however vague, is helpful. The better narrative will be the one that best exemplifies the following characteristics:

    • It will tend to provide the most complete story.
    • It will tend to provide the simplest story.
    • It will provide the most plausible story.
    • It will provide the least ad hoc story.

    But fleshing out these criteria for explanatory and narrative success is clearly unfinished business in the philosophy of science and narratology.

    As serious as this problem clearly is, I don’t believe that it is as serious as the skeptics make it out to be. I know how to speak, understand, read, and write English. I know that the English sentence “Colorless green ideas sleep furiously,” though nonsensical and probably self-contradictory, is grammatically correct. According to a dominant tradition in Western epistemology, if I am right about my linguistic skills, I should be able to plainly articulate the rules I have used to recognize the grammatically of the green ideas sentence.

    Obviously, every speaker of a language has mastered and internalized a generative grammar that expresses his knowledge of his language. This is not to say that he is aware of the rules of the grammar or even that he can become aware of them.4

    Chomsky concedes that the rules of this generative grammar may be cognitively inaccessible and certainly difficult to articulate. Jason Stanley vigorously demurs:

    Knowing how to do something is the same as knowing a fact. It follows that learning how to do something is learning a fact. For example, when you learned how to swim, what happened is that you learned some facts about swimming. . . . You know how to perform activities solely in virtue of your knowledge of facts about those activities.5

    Socrates clearly articulated this epistemological principle 2,500 years ago—“and that which we know we must surely be able to tell.”6

    I side, however, with Michael Polanyi when he says, “We can know more than we can tell.”7 He uses a very apt example:

    This fact seems obvious enough; but it is not easy to say exactly what it means. Take an example. We know a person’s face, and can recognize it among a thousand, indeed among a million. Yet we usually cannot tell how we recognize a face we know. So most of this knowledge cannot be put into words.8

    Polanyi introduces the technical term tacit knowledge to label knowledge or skills that “cannot be put into words.” Polanyi is surely engaging in purposeful hyperbole. Most skills can be put into words, but these words are usually vague and general, and at times, the words are downright misleading.

    The essence of Glass’s first problem—“how one explanation should be compared against another so that the best explanation can be identified”—is that most of the defenders and critics of inference to the best explanation seem to seek something that I believe is unattainable. They seem to be searching for a kind of mechanical algorithm that validates an objective determination of one explanation being superior to another explanation. Perhaps the biggest temptation for insisting on a list of necessary and sufficient conditions for being the best explanation or story (or a better explanation or story) is the persistent illusion that all things we are skilled at can be articulated in clear, concise recipes or formulae. This is precisely the Plato and Stanley article of faith—“that which we know we must surely be able to tell.” We should know that is a mistake.

    Consider how remarkable it is that major league hitters can hit ninety-five-mile-an-hour fastballs.

    A typical major league fastball travels about 10 feet in just the 75 milliseconds that it takes for sensory cells in the retina to confirm that a baseball is in view and for information about the flight path and velocity of the ball to be relayed to the brain. The entire flight of the baseball from the pitcher’s hand to the plate takes just 400 milliseconds. And because it takes half that time merely to initiate muscular action, a major league batter has to know where he is swinging shortly after the ball leaves the pitcher’s hand—well before it’s even halfway to the plate. . . . A batter could just as well close his eyes once the ball is halfway to home plate. Given the speed of the pitch and the limitations of our physiology, it seems to be a miracle that anybody hits the ball at all.9

    So how do they do it? There are the clichés—“Keep your eye on the ball,” “Don’t open up too soon,” and the like. But these don’t tell you how it’s done; they are mnemonics to help skilled hitters get back on track when they are in slumps. No one has yet, and I insist never will, articulate the logical criteria for hitting major league fastballs. This emphatically does not mean, however, that the hitting, not the describing, can’t be done. This skill, like many others, is a kind of tacit knowledge.

    My mentor, Larry Wright, tells an important story:

    Virtually everyone who has survived past infancy has a more or less well developed set of perceptual skills. These skills may be generally described as the ability to tell what’s going on (sometimes) simply by seeing it . . . This ability to tell what’s going on—or what’s gone on—even when we are not confronting it directly. We can often tell what has happened from the traces it leaves. We can tell there was a frost by the damaged trees; we know it rained because the mountains are green; we can tell John had some trouble on the way home from the store by the rumpled fender and the broken headlight. We reconstruct the event from its telltale consequences. It is this diagnostic skill we exploit in the most basic sort of inductive arguments; it is the foundation of our ability to evaluate evidence.10

    This quasi-perceptual skill is what allows us to see what’s going on and what’s true or at least what’s the best bet given what we know. And the fact that the precise nature of this skill has proven incredibly difficult to articulate in no way counts against its existence and utility. Can anyone seriously doubt that Pete Rose knew how to hit because he could not say how he was able to hit?

    Literary Darwinism

    Wright talks of a “diagnostic skill,” “the ability to tell what’s going on.” I’d characterize it as a skill at making sense of things. What is the source of this skill? The answer to this question leads us directly to Glass’s second worry—“Does IBE track truth?” I am committed, of course, to a resounding affirmative answer. But I certainly owe the inference-to-the-best-explanation and inference-to-the-best-narrative skeptics at least an outline of “some reason for thinking that it provides a good strategy for determining the truth.”

    What is sometimes called literary Darwinism traces human storytelling back to evolutionary origins of modern human cognition.

    Minds exist to predict what will happen next. They mine the present for clues they can refine with help from the past—the evolutionary past of the species, the cultural past of the population, and the experiential past of the individual—to anticipate the immediate future and guide action. To understand events as they happen, with limited time, knowledge, and computational power, minds have evolved to register the regularities pertinent to particular species and infer according to rough-and-ready heuristics.11

    This little narrative assumes that we are pretty good at “predict[ing] what will happen next.” But it explains much more than the ubiquity of human storytelling; it accounts for our general ability to make sense of things, to explain what’s going on.

    We can tell stories to explain things, from a child’s or a country’s pouty “They started it” to why the world is as it is according to myth or science. . . . Why has the richest explanatory story of all, the theory of evolution by natural selection, been so little used to explain why and how stories matter?12

    Inference to the best explanation (IBE) and inference to the best narrative (IBN) track the truth because they rely, at base, on quasi-perceptual skills that were selected for precisely to do this job.

    Consider this explanatory narrative:

    Babies may have little control over their bodies, but they can willingly move their heads and eyes. And what a baby looks at can tell you something about how it sees the world. This is because babies are like adults in some regards. If they see the same thing over and over again, they get bored and look away. If they see something new or unexpected, they look longer. Thus, analyzing looking time can tell what babies think of as being “the same thing,” and what they see as “new or unexpected.”13

    The above two-stage inference to the best explanation—differential gaze times being explained as boredom or surprise and then as “same” or “new”—is the methodological presupposition for a host of fascinating experiments in the study of infant cognitive development. Paul Bloom provides a nice summary of some of these results:

    1. 1. Cohesion: If a hand pulls at an object, babies expect the entire object to go with the hand; if it comes off in pieces, they are surprised, showing an expectation that objects are cohesive.
    2. 2. Continuity: Imagine a stage with two vertical barriers separated in space. A small object, like a box, goes behind the barrier on the left, continues between the barriers, goes behind the barrier on the right, and comes out the other side. Adults see this is a single object, and so do babies. Now imagine that a box goes behind the barrier on the left, there is a pause, and then the box emerges for the screen on the right, never appearing in the gap. Adults assume there are two boxes here, not one. Babies make the same assumption; they expect continuity.14

    Why do we find differential gaze times for the hand pulling the object and it remaining whole, and the hand pulling the object and it coming off in pieces? Babies expect objects to be cohesive. Why the perception of a single box in the first experimental scenario with the box and the barriers but the perception of two boxes in the second scenario? Babies expect continuity. But where do these expectations come from? Bloom’s answer is a classic blend of nature and nurture.

    These results show that although babies enter the world with a foundational understanding of what objects are and how they act, it is incomplete, and this foundation grows. Some of the improvement might be due to maturation of the brain—like the rest of the body, the brain changes rapidly in the early years of life, and this might cause corresponding increases in knowledge. But some of the improvement is plainly due to experience.”15

    And finally, what explains this foundational understanding of objects and how they act? This knowledge is clearly innate. Natural selection has hardwired infant brains to expect cohesion and continuity. It is easy to see the adaptive value for human infants having rudimentary understanding not just of objects and “folk physics” but also of agency and social relationships. Certainly, shared understanding of folk physics, agency, and social relationships are the cornerstones of the sort of the practical explanatory skill that would have been of value in hunter-gather times.

    Sally and Ann

    Let me tell you two stories about Sally and Ann. Sally prizes her special marble. When she leaves, she always places it in a basket and carefully covers it with a soft blanket. Ann has been hiding and watching Sally’s little ritual. After Sally has gone home for lunch, Ann removes Sally’s marble from under the blanket and hides it in a nearby box. Sally returns after lunch and goes to retrieve her marble. She goes right to the box and finds it there! Why? Well, because that’s where the marble is! The second story begins just as the first, but things take a turn when Sally returns from lunch. Sally goes straight for the basket and is heartbroken not to find her marble under the blanket. Why does she do this? Well, that’s where she remembers putting it before lunch.

    When shown a puppet version of the beginning of the Sally and Ann stories and then asked to predict where Sally will go to look for her marble, children younger than around four typically predict the box because they know that’s where the marble is. But between four and five, children’s predictions dramatically change. They now realize that Sally will look in the basket because that’s where she would remember putting it.

    Why do the more cognitively mature children simply recognize that the Sally-goes-to-the-basket narrative is significantly better than the Sally-goes-to-the-box account? They have begun to develop what is often called a “theory of mind.”

    Theory of mind allows a much more precise and multiperspectival understanding of social event. Because we understand beliefs as the basis for forming desires, goals and intentions, and because we understand the sources of belief, we automatically and effortlessly track what other might know about a situation and can therefore understand their behavior more finely. . . . Almost automatically we track what others can know, and that makes all the difference to our capacity to cooperate or compete.16

    Even back in hunter-gather times, our human ancestors were very skilled social explainers. Contemporary cognitive science provides a very plausible account of the origins of this skill.

    [Mind reading] is used by cognitive scientists, interchangeably with “Theory of Mind,” to describe our ability to explain people’s behavior in terms of their thoughts, feelings, beliefs, and desires. . . . This adaption must have developed during the “massive neurocognitive evolution” which took place during the Pleistocene (1.8 million to 10,000 years ago). The emergence of a Theory of Mind “module” was evolution’s answer to the “staggeringly complex challenge faced by our ancestors, who needed to make sense of the behavior of other people in their group, which could include up to 200 individuals.”17

    If this is right, and I certainly think it is, it suggests a somewhat surprising inversion in our thinking about explanation. Rather than extrapolating from the more “basic” notion of a causal explanation to account for our narrative skills, it might actually be that our ability to construct narratives about the behavior and motives of those in our social groups is what leads to the wider ability to construct scientific or causal narratives in situations where agents are conspicuously absent.


    But wait a second you may well counter. How can I possibly claim that our skills, both innate and learned, at explaining and making sense of things, are reliable enough foundation for a general logical procedure such as inference to the best explanation or inference to the best narrative? Clearly rival explanations and rival narratives are not just possible but strongly endorsed by equally intelligent and reflective evidence assessors. The district attorney and Abe held dramatically different interpretations regarding Hamilton and the murder. I believe Mary Ann and Wanda should have placed more confidence in the criminal justice system and not murdered Earl; the ladies saw things very differently. Intellectual disagreement seems to count heavily against my claims for explanatory and narrative skill. How can Justice Blackmun and Justice Scalia be skilled constitutional explainers and story judgers when they see things so dramatically differently with respect to the death penalty and the Constitution? These worries are legitimate and require attention and potential solutions.

    A big part of the story to be told here is one of simple intellectual modesty. One can be very skilled at something and at the same time fail spectacularly at exercising the skill. We are all skilled at recognizing faces. But we still misperceive all the time—“Hi Joanie! Oh, sorry, you look just like my mother-in-law.” Major league hitters perform the minor miracle of hitting ninety-five-mile-an-hour fastballs, but they also swing wildly, miss, and look foolish, and lest we forget, they fail to get base hits between two-thirds and three-quarters of the time. Furthermore, the skills that I am basing my argument upon were developed, honed, and tested in hunter-gatherer times. They can only be applied to science and the law by extension.

    Humans do not readily engage in [the highly abstract reasoning required in modern science, philosophy, government, commerce, and law]. In most times, places, and stages of development consists of quantities “one,” “two,” and “many” . . . Their political philosophy is based on kin, clan, tribe and vendetta, not on the social contract. . . . And their morality is a mixture of intuitions of purity, authority, loyalty, conformity, and reciprocity, not generalized notions of fairness and justice . . . Nevertheless, some humans were able to invent the different components of modern knowledge, and all are capable of learning them.18

    Please don’t misread my meaning here. I’m really good at spotting my mother-in-law, I’m in awe of the hitting prowess of the guys on my fantasy team, and as a teacher, I know firsthand that students, even the mediocre ones, can cast aside kin, clan, and vendetta and learn to embrace the social contract and justice and fairness.


    Let’s see if we can do a little better than the trivial definition of “truth” I offered in chapter 3—truth =df not-false. Inference to the best narrative is unapologetic about a close connection between narrative superiority and truth. The best story does not guarantee truth, but it does constitute evidence for what the truth is. Perhaps there is a better yet story that no one has thought to tell—that’s certainly been the case at specific points in the history of science. Perhaps, as I believe is often the case with many narratives, the best story is one that actually combines elements and insights from the competing narratives. But this is the nature of evidence generally. Even the strongest evidence can point in the wrong direction—evidence is not logical proof. But none of this implies that we should disregard evidence. Indeed, what choice do we really have but to base all our considered judgments, not just in law and scholarship, but in every aspect of our lives, on what the best available evidence tells us is likely true?

    Legal, constitutional, and scholarly truth, just like truth in science and regarding violent ex-husbands, remains philosophically problematic. I agree with Peter Kosso that the most intuitive sense of truth—at least in most explanatory contexts—is the correspondence theory, but that correspondence must be inferred from coherence.

    Though truth is correspondence with the facts it cannot be recognized by its correspondence. We cannot rely on the facts to guide proofs of scientific theories, since the facts are irretrievably at the outer end of the correspondence relation. . . . So any indicators of truth must be internal. . . . The process of justifying, then, is a process of comparing aspects of the system, and the accomplishment of justification is the demonstration of coherence among the aspects.19

    Such a model captures our intuitions about what really happened to Nicole and Ron or at the record hop. There aren’t just stories to be told about these happenings, but clearly, some stories are better than others—stories that point us to the truth. We believe that there’s a world out there, though we will never see it from the God’s eye perspective, and in this world, things happened involving O. J., Connie’s boyfriend, and the rest. These external happenings play a significant role in what counts as true.

    Things get much trickier, however, when we consider the best narrative concerning Brown v. Board of Education or Mary Ann and Wanda’s predicament. We are still confident that there is a best story or, at least, stories that are significantly better than others. But where does narrative superiority now point? What of the standard jurisprudential questions of how to interpret a statute, a line of precedent, or a constitutional text? Or even how to interpret the sad events confronting Mary Ann and Wanda? To reiterate the previous argument, I claim that in these cases we tell stories that try to make sense of the relevant texts and precedent as well as Earl’s violent behavior and Geneva’s story about Brown. When we tell these stories, we tell them with passion and conviction. We are convinced that our story is the best or, at least, a heck of a lot better than the other stories that are out there. Does inference to the best narrative not so much discover the truth but actually create the truth? This would be a mischaracterized insight. The insight, of course, is that few of us believe there is a Platonic heaven where moral and interpretive truth live and to where we can retreat to adjudicate controversies involving Earl’s murder or how we should understand Brown v. Board of Education. But it is mischaracterized because truth is not being created in the way Derrick Bell was able to make up his story about the disappearing black schoolchildren. It makes perfectly good sense to insist that there is an “objectively” best narrative, even when reasonable people disagree about what it is. And what other laudatory title would we bestow on such a superior narrative other than “true”?


    1. 1. How can a major league hitter possibly hit a ninety-five-mile-an-hour fastball if he can’t say how he does it?
    2. 2. What do the two Sally and Ann stories tell us about our ability to make sense of what others do?
    3. 3. Is there an objective truth about what happened to Hamilton’s partner? Is there an objective truth about what Mary Ann and Wanda should do? What does all this say about the notion of truth in the first place?


    What are the two problems for inference to the best explanation (and for inference to the best narrative) that were identified by David H. Glass? What is my proposed solution to these two problems? Do you think my solution works? Why, or why not?


    1. Ronald Dworkin, Justice for Hedgehogs (Cambridge, MA: Harvard University Press, 2011), 152.

    2. David H. Glass, “Inference to the Best Explanation: Does It Track Truth?,” Synthese: An International Journal for Epistemology, Methodology and Philosophy of Science 185, no. 3 (2012): 411–27.

    3. Gilbert Harman, “The Inference to the Best Explanation,” Philosophical Review 74, no.1 (1965): 88–89.

    4. Noam Chomsky, Aspects of the Theory of Syntax (Cambridge, MA: MIT Press, 1956), 8.

    5. Jason Stanley, Know How (Oxford: Oxford University Press, 2011), vii.

    6. Plato, The Collected Dialogues of Plato (Princeton, NJ: Princeton University Press, 1961), 133.

    7. Michael Polanyi, The Tacit Dimension (Chicago: University of Chicago Press, 1966), 4.

    8. Polanyi, 4.

    9. David Epstein, The Sports Gene (New York: Penguin, 2013), 4.

    10. Larry Wright, Better Reasoning (New York: Holt, Rinehart, Winston, 1982), v.

    11. Brian Boyd, On the Origin of Stories (Cambridge, MA: Harvard University Press, 2009), 134.

    12. Boyd, 1.

    13. Paul Bloom, Descartes’ Baby (New York: Basic Books, 2004), 9–10.

    14. Bloom, 11–12.

    15. Bloom, 13–14.

    16. Boyd, Origin of Stories, 149.

    17. Lisa Zunshine, Why We Read Fiction: Theory of Mind and the Novel (Columbus: Ohio University Press, 2006), 7.

    18. Steven Pinker, Language, Cognition, and Human Nature (Oxford: Oxford University Press, 2013), 359.

    19. Peter Kosso, Reading the Book of Nature (Cambridge: Cambridge University Press, 1992), 136.

    This page titled 1.15: Explanatory Virtue and Truth is shared under a CC BY 4.0 license and was authored, remixed, and/or curated by Jeffery L. Johnson (Portland State University Library) via source content that was edited to the style and standards of the LibreTexts platform; a detailed edit history is available upon request.

    • Was this article helpful?