in press, April 2007 

Epistemic Development and the Perils of Pluto 

David Moshman 
To appear in M. Shaughnessy, M. Vennemann, & C. K. Kennedy (Eds.), Metacognition. 
Hauppauge, NY: Nova. I am grateful to Daniel Abbott, Deanna Kuhn, Rick Lombardo, 
Annick Mansfield, Yolanda Rolle, Martin Rowley, and Les Smith for feedback on a draft 
of this chapter. Author's address: Department of Educational Psychology, University of 
Nebraska Lincoln, Lincoln, NE 68588-0345, USA (email: [email protected]). 
 
At the start of the year 2006 the Earth was the third from the sun among the nine 
planets in its solar system. By the end of that year, Earth was the third of eight planets on 
a list that now ended with Neptune. How did that come about? What happened to Pluto? 
One possibility is that Pluto was destroyed by a starship from another part of the 
galaxy, or perhaps from the future. Perhaps the intent was to destroy the Earth but the 
ship mistakenly targeted Pluto instead. Or maybe it was an accident. Perhaps Pluto was 
destroyed by a collision with some other object way out there. It’s a dangerous universe. 
Or perhaps it had some sort of internal, geological catastrophe. There are many potential 
scenarios in which Pluto is destroyed, leaving just eight planets. 
But none of this happened in 2006. Pluto was not blasted out of space. It was 
redefined. More precisely, the definition of planet was changed, after years of 
contentious debate among astronomers and others, so as not to include Pluto. The object 
of this astrolexical brouhaha meanwhile continues on its vast course around the sun, 
sublimely oblivious to all the turmoil surrounding its conceptualization by a primate 
species on the third planet. Nothing happened to Pluto out there beyond Neptune but 
plenty happened to it here on Earth. To understand why there are no longer nine planets 
we need to look into our own minds. 
But how can that be? Is the world a function of our minds? As it turns out, 
moreover, it’s not just Pluto that perplexes us, and not just the number of planets and 
not just astronomy. Things we thought we knew often turn out to be false. Sometimes we 
examine how we know that something is true. Sometimes we realize we don’t know 
what’s true. Sometimes we don’t know how to tell what’s true. Sometimes we think 
2 
 
about what it means for something to be true. Sometimes we consider how we can tell 
whether anything is true. Sometimes we wonder whether we can ever really know 
anything at all. Sometimes we worry about how we can live our lives in the light of that. 
Sometimes we find our way through such epistemic doubts. Sometimes we don’t. 
But is there any pattern to all this? Do we simply mull over such questions or do 
we make progress in our understanding of knowledge? Do we simply come up with a 
variety of beliefs about knowledge or does our knowledge about knowledge develop? 
Research indicates that children know something about knowledge as early as age 4 and 
make remarkable progress for many years thereafter. To the extent that they have 
knowledge about knowledge, they may be said to have epistemic cognition. To the extent 
that they make progress in such knowledge, their progress is epistemic development. 
That is: 
(1) Epistemic cognition is knowledge about knowledge. 
(2) Epistemic development is progress in epistemic cognition. 
Epistemic development, then, is progress in knowledge about knowledge. In this 
chapter I propose a theoretical account of epistemic development that draws on several 
discrete literatures: (a) a literature on epistemic cognition in adolescents and adults in the 
tradition of William Perry (1970; Chandler, 1987; Chandler, Boyes, & Ball, 1990; 
Chandler, Hallett, & Sokol, 2002; King & Kitchener, 1994, 2002; Kuhn, 2005; Kuhn, 
Cheney, & Weinstock, 2000; Kuhn & Franklin, 2006; Kuhn & Weinstock, 2002; 
Mansfield & Clinchy, 2002; Moshman, 2005); (b) a literature on young children’s 
theories of mind, including their understanding of the possibility of false belief 
(Wellman, Cross, & Watson, 2001); (c) a literature on the development of interpretive 
3 
 
and constructivist theories of mind in the elementary school years (Carpendale & 
Chandler, 1996; Fabricius & Schwanenflugel, 1994; Lalonde & Chandler, 2002; Pillow 
& Henrichon, 1996; Rowley & Robinson, in press; Schwanenflugel, Fabricius, & Noyes, 
1996); (d) a literature on the development of metalogical understanding (Miller, Custer, 
& Nassau, 2000; Moshman, 1990, 2004, forthcoming; Pillow, 2002); and (e) a literature 
on domain specificity in cognition (Komatsu & Galotti, 1986; Nicholls & Thorkildsen, 
1988; Wainryb, Shaw, Langley, Cottam, & Lewis, 2004). With regard to previous 
theoretical approaches, the present account draws on (a) Wainryb et al.'s (2004) 
conception of distinct epistemological domains differentiated even by young children; (b) 
Chandler's (1987; Chandler et al., 2002) distinction between "retail" and "wholesale" 
levels of epistemic cognition; and (c) the standard three-stage conception of advanced 
epistemic development as the reflective coordination of objectivity and subjectivity 
(Kuhn & Franklin, 2006, Mansfield & Clinchy, 2002; Moshman, 2005). 
The theory, in two paragraphs, is as follows: Children may be said to have 
epistemic cognition from the time they understand that beliefs may be false, about age 4. 
Over the course of the elementary school years children increasingly distinguish at least 
three epistemic domains: (1) an objective domain of truth, (2) a subjective domain of 
taste, and (3) a rational domain of reasonable interpretation. Questions about the truth, 
falsity, or justification of particular beliefs and inferences arise and are addressed within 
these domains. 
Adolescents and adults, unlike children, often theorize more abstractly about the 
nature and justification of knowledge in general. Development beyond age 11 or 12 is 
much less universal and predictable than earlier development. To the extent that 
4 
 
adolescents and adults construct advanced forms of epistemic cognition, however, they 
progress in a universal sequence from (1) objectivist epistemologies, which take 
verifiable facts and logical proofs as paradigm cases of knowledge; to (2) subjectivist 
epistemologies, which view knowledge (if we even call it that) as opinion, and opinion as 
a matter of taste; to (3) rationalist epistemologies, which construe knowledge, in a world 
of interpretation and inference, as justified belief. 
The Perils of Pluto 
Consider the following questions: 
1. Is Pluto a planet? 
2. How many planets are there in our solar system? 
3. Which planet is best? 
4. Which planet(s) can support life? 
For most of the 20th century, the first two questions seemed simple. Pluto, which 
was discovered in 1930, was a planet. There were nine planets. These were matters of 
fact. You could learn things like this from teachers, parents, and experts. You could look 
them up in books. In principle, with sufficient expertise and technology, you could 
observe for yourself that Pluto is a planet and could count the number of planets. On 
matters of fact, you can perceive the truth. (Or so it would seem.) 
Which planet is best? This seems a different kind of question. No planet is 
“better” than another, at least not in any general sense. We might instead ask, “Which is 
your favorite planet?” That formulation acknowledges that choosing a planet as best, in 
some general and undefined sense, is a matter of personal preference about which people 
5 
 
legitimately differ. Even with full access to information, careful reflection, and extended 
discussion, we cannot count on consensus. 
Which planets can support life? We know the Earth can, because it does. Can 
Mars? Venus? Jupiter, or maybe its volcanic moon Io? If we go beyond planets, how 
about one of the larger asteroids? Come to think of it, how about Pluto, whatever it is? 
These questions seem different from questions about which planet is best, but also 
different from questions about the number of planets. We can get evidence relevant to 
whether various planets (or other heavenly bodies) could support various kinds of life, 
but such evidence is open to multiple interpretations. There can also be different 
conceptions of what counts as "life" and what it means to "support" it. The question of 
whether a planet can support life is more factual than the question of how "good" it is, but 
less a matter of simple truth than the question of whether it is a planet. The question of 
potential support for life, and other questions of this sort, are neither matters of truth nor 
matters of taste. Instead, they are matters of interpretation. 
It thus appears that we can distinguish three categories: (1) matters of truth, (2) 
matters of taste, and (3) matters of interpretation. Matters of truth fall within an objective 
domain in which truth and falsity can be clearly and sharply distinguished. In this 
domain we expect agreement unless someone is making a mistake. Matters of taste fall 
within a subjective domain in which truth and falsity are irrelevant. In this domain we 
neither expect agreement nor try to achieve it. Finally, matters of interpretation fall 
within a domain of reasonable judgment in which some ideas are better justified than 
others but truth cannot be proven. In this domain we acknowledge that different minds 
6 
 
may reach legitimately different conclusions but we believe evidence and argument may 
generate progress in understanding. 
And now, in the present, let us reconsider the first two questions. Is Pluto a 
planet? No. How many planets are there? Eight. But what are we to make of the 
changes in our answers? Did Pluto used to be a planet, but now it isn’t? Did there used 
to be nine planets, but now there are eight? But how could that happen, given that Pluto 
is still there and nothing about it has changed? Maybe Pluto was never a planet. Did we 
used to think Pluto was a planet, but new evidence showed it is not? But Pluto's 
deplanetization was not the result of new evidence about it. So why is it not a planet, and 
how sure are we that this time we have finally got it right? All we really know for sure, it 
appears, is that we used to consider Pluto a planet and now we don't. Are all facts subject 
to such radical change? Can any truth turn out to be false? If so, do we really know 
anything at all? 
The purpose of the present chapter, of course, is not to set things straight about 
Pluto but rather to consider what people know about knowledge and how knowledge 
about knowledge develops. With that in mind, we can make an initial distinction 
between two aspects of epistemic cognition, both illustrated by the above questions. 
Epistemic cognition includes both (a) knowledge about the epistemic status and 
properties of specific beliefs and inferences and (b) knowledge about the general nature 
and justification of knowledge. Research on epistemic development shows that epistemic 
cognition regarding specific beliefs and mental processes shows substantial development 
over the course of childhood, whereas epistemological theories of a more general, 
abstract, and explicit sort are not seen in childhood but often develop over the course of 
7 
 
adolescence and early adulthood. It thus appears useful to distinguish two levels of 
epistemic development, one associated with childhood and the other with adolescence 
and adulthood (Chandler, 1987; Chandler et al., 2002). 
Epistemic Development in Childhood 
The origin of epistemic cognition lies in the recognition that beliefs can be false. 
This creates a distinction between truth and falsity and raises questions about the 
justification of beliefs. Developmental research indicates that comprehension of the 
possibility of false belief develops dramatically between ages 3 and 5 (Wellman et al., 
2001). By age 4 or 5, but rarely much before that, children clearly understand that lack of 
information can result in false beliefs, in themselves and others, and that they and others 
act on the basis of their beliefs even when those beliefs are false. Thus epistemic 
cognition begins in the preschool years (Kuhn, 2005; Kuhn & Weinstock, 2002; 
Moshman, forthcoming). Evidence from several distinct literatures indicates that over 
the course of middle childhood (roughly ages 6-10) children increasingly distinguish the 
three epistemologically relevant domains noted earlier: (1) a domain of truth, (2) a 
domain of taste, and (3) a domain of interpretation. 
The domain of truth is where objectivity prevails. Assertions are true or false. If 
I say there are nine planets and you say there are eight, and if we’re still in the 20th 
century, we can look in a book and determine that my assertion is true and yours is false, 
which settles the matter. If I say there are at least as many planets in our galaxy as there 
are in our solar system, it can be proved logically that this must be true. If I say harming 
8 
 
innocent people for no good reason is wrong, we can agree on the correctness of my 
claim. The domain of truth, then, includes factual, logical, and moral truths. 
With regard to matters of fact, children understand by the age of about 4, as we 
have seen, that they and others can hold demonstrably false beliefs (Wellman et al., 
2001). They understand that a belief can be true or false, and that its truth or falsity can 
be determined by examining the relation of the belief to reality. Beliefs that correspond 
to reality are true and those that do not are false. 
With regard to logic, preschool children also make inferences, including 
deductive inferences, but they are largely unaware of, and thus unable to evaluate, their 
inferences. By age 6, however, children recognize inference as a source of knowledge 
(Rai & Mitchell, 2006; Sodian & Wimmer, 1987). Over the next few years they 
increasingly recognize the logical necessities inherent in class hierarchies (for example, 
that there must be at least as many flowers as daisies), mathematical truths (3 is always 
and everywhere bigger than 2), and deductive inferences (as distinct from inductive 
inferences and partially informed guesses) (Miller et al., 2000; Moshman, 1990, 2004, 
forthcoming; Piaget, 1987, 2001, 2006; Piéraut-Le Bonniec, 1980; Pillow, 2002; Pillow 
& Anderson, 2006; Smith, 1993). 
Finally, with regard to morality, there is extensive evidence from social cognitive 
domain theorists that by age 5, if not before, children distinguish moral from other values 
on the basis of objective and universal standards of right and wrong. Failing to sit in your 
assigned seat, for example, is wrong if there is a rule that children must sit in assigned 
seats but would be acceptable in another school without such a rule. Hitting another 
child, however, is wrong regardless of whether there is a rule against this because it hurts 
9 
 
the other child. More generally, even children as young as 5 years see issues of rights, 
justice, harm, and welfare as distinct from issues of authority, convention, and personal 
choice as constituting a moral domain in which objective standards prevail (Wainryb et 
al., 2004). 
In the domain of taste, in contrast, subjectivity prevails. Idiosyncratic preferences 
are neither true nor false, and efforts to argue for them are pointless. These may be 
personal preferences for particular flavors of food taste in the most literal sense or 
cultural preferences for particular social conventions, but we do not expect to convince 
others that chocolate really tastes better than vanilla, or vice versa, or that a particular 
manner of assigning seats is the one true way. We can say that a food tastes good to a 
particular person or that a social convention feels right to those brought up with it, but 
there is no rational basis for claiming that all people ought to enjoy the taste of that food 
or ought to practice that social convention. 
Research shows that at least some children recognize the subjectivity of taste as 
early as age 5 and that matters of taste are increasingly distinguished from matters of 
truth over the elementary school years (Carpendale & Chandler, 1996; Kuhn et al., 2000; 
Rowley & Robinson, in press; Wainryb et al., 2004). Wainryb et al. (2004), for example, 
presented children ranging in age from 4 to 9 years with disagreements reflecting matters 
of taste (e.g., whether chocolate ice cream is "yummy" or "yucky"), fact (e.g., whether 
pencils "go up" or "fall down" when released), and morality (e.g., whether it is "okay" or 
"wrong" to hit and kick other children). Regardless of age, children were virtually 
unanimous in maintaining that there was one right answer with regard to matters of fact 
and morality. With regard to matters of taste, in contrast, even some of the younger 
10 
 
children saw both opinions as legitimate and justified their responses on the basis of the 
subjectivity of taste. Acceptance of diversity and acknowledgment of subjectivity with 
regard to taste increased substantially with age. The domains of taste and truth may be at 
least partially distinct as early as age 5, it appears, and become increasingly differentiated 
over the course of middle childhood (see also Komatsu & Galotti, 1986; Nicholls & 
Thorkildsen, 1988). 
The domain of interpretation is the domain where reasons may be less than 
definitive without being entirely personal, cultural, or arbitrary. Given the constructive 
nature of the mind, diversity of interpretation is to be expected. Within the domain of 
interpretation, however, such diversity is not associated with an ultimate and 
nonnegotiable subjectivity. On the contrary, interpretation of an object must in some way 
be true to that object in order to qualify as an interpretation rather than a pure act of 
imagination. Thus interpretation has an objective aspect of accommodation to reality in 
addition to its subjective aspect of assimilation by an active mind. 
Carpendale and Chandler (1996) presented children of ages 5 through 8 with tasks 
in which two puppets disagreed with each other about the interpretation of (a) ambiguous 
figures (such as the classic “duck-rabbit”), (b) lexical ambiguity (e.g., “wait for a ring”), 
or (c) ambiguous referential communication (eliminating one possibility but leaving two 
others). They also provided control tasks in which (a) an additional puppet provided a 
deviant interpretation (e.g., interpreting the duck-rabbit as an elephant) or (b) the puppets 
disagreed about matters of taste (e.g., whether a particular soup tastes “good” or “bad”). 
In addition, every child received, and passed, a standard false belief task. The 5- and 6- 
year-olds seemed mostly unable to comprehend the legitimacy of different interpretations 
11 
 
in these tasks, though they were somewhat more able to recognize and explain 
differences in taste. The 7- and 8-year-olds, in contrast, often recognized and explained 
that the two interpretations were both reasonable given the ambiguous nature of the 
stimulus and that the response of a new child could not be predicted. They usually 
recognized, moreover, that the deviant interpretations were not reasonable. Further, 
children’s explanations showed an emerging distinction between the domains of 
interpretation and taste. At least some of them recognized that in the domain of 
interpretation, in contrast to that of taste, responses must be grounded in evidence and 
reasons concerning the object to be interpreted. Thus they distinguished the domain of 
interpretation not only from the objective domain of truth, in which everything is true or 
false, but also from the subjective domain of taste, in which nothing is true or false, or 
even better or worse. It appears that by age 7 or 8, many children are beginning to see 
interpretation as distinct from both matters of truth and matters of taste (see also Lalonde 
& Chandler, 2002; Pillow & Henrichon, 1996). 
By age 7 or 8, then, children are beginning to distinguish (a) an objective domain 
of truth, (b) a subjective domain of taste, and (c) a domain of reasonable interpretation 
that is distinct from both. These distinctions become increasingly clear over the next few 
years (Rowley & Robinson, in press; Wainryb et al., 2004). Children’s epistemic 
cognition, however, remains focused on particular beliefs, inferences, interpretations, and 
judgments. Children do not think about knowledge in the abstract or in general 
(Chandler et al., 2002). Children intuitively distinguish at least three epistemic domains 
in thinking about questions and issues of knowledge, but they do not think about the 
domains themselves, they do not reflect on the interrelations and coordinations of 
12 
 
subjectivity and objectivity, and they do not think about the general nature of knowledge 
and justification. Many adolescents and adults, however, do indeed think about such 
matters. 
Epistemic Development in Adolescence and Beyond 
Suppose you are a child of 9 or 10 in 2006. How might you react to the news that 
Pluto is no longer a planet and that there are now eight planets? If you hear it first from a 
friend you might assume you are being tricked. Assuming you also hear this from 
teachers, parents, television, and other sources, however, you would soon accept the 
change. You already knew that people can turn out to be wrong. Now you know, if you 
didn’t already, that even scientists can turn out to be wrong. But you are unlikely to 
perceive a problem with the distinction between truth and falsity or with the status of the 
domain of truth as a distinct domain. If you are sufficiently interested in astronomy you 
may learn, perhaps with surprise, that nothing whatsoever happened to Pluto, and that 
scientists have not learned anything new about it that changed their minds regarding its 
planetary status. In response to your questions, you may learn that scientists decided 
Pluto is no longer a planet because they have a new definition of planet. Even if you find 
this perplexing, however, you almost surely see it as a problem of astronomy, not 
epistemology. 
Suppose now you are in your early teens, or beyond, and very interested in 
astronomy. You like it, perhaps, because it is so scientific. You like it better than, say, 
art or literature, where everyone just spouts their opinions. Scientists work with facts. 
Even their theories are based on facts and tested against facts, so we can tell what’s true 
13 
 
and what’s false. This is real knowledge. Of course you can have knowledge about a 
novel, such as the names of the characters and what they did, or about a work of art, such 
as the colors and the name of the artist, but once you get into the realm of opinion, it’s all 
a question of taste, not knowledge. This objectivist epistemology takes the objective 
domain of truth as paradigmatic of knowledge. Matters of taste are not matters of 
knowledge. As for interpretation, to the extent that it involves logical inference and 
affords the potential for consensus, it qualifies as knowledge. All the rest is mere 
opinion, not knowledge. 
But what about Pluto? Is it a planet or isn’t it? Are there eight planets or nine? 
These are, it would seem, simple matters of fact. If we had learned something new about 
Pluto such that it no longer qualified as a planet, this would be simply a case of correcting 
a mistake, and would not raise any epistemological issue. If instead a new planet had 
been discovered, bringing the total number to ten, that would have been astronomically 
exciting but of no epistemological concern. Alternatively, if Pluto had been destroyed in 
one of the scenarios suggested at the opening of this chapter, that would have been 
astronomically and psychologically dramatic but epistemologically ho-hum. What makes 
the new status of Pluto and new number of planets epistemologically interesting is that 
Pluto is still there, unchanged, and we are not reacting to any new information about it. 
The change is a matter of definition. As an objectivist you might assume definitions 
sometimes need to be corrected, and in any event this is a rare and special case. Thus you 
see the problem of Pluto as an anomaly and preserve your objectivist epistemology. 
But if you have enough experience with matters of this sort you may begin to see 
that many things are “a matter of definition,” even in the so-called “hard” sciences. If 
14 
 
basic facts of astronomy are just matters of definition, what does that say about 
knowledge in general? What does that say about the possibility of really having any 
knowledge at all? Further consideration of these matters may convince you that all 
knowledge is relative to definitions and concepts imposed by the knower. Facts, if we 
even call them that, are determined by our concepts and definitions; nothing is really true 
or false. Under one definition there are nine planets; under another there are eight. If we 
want there to be, say, 15 planets, we could come up with a set of criteria of planethood 
that would hold for 15 astronomical objects, or whatever number we desired. If Pluto is 
our favorite, we can use a definition that makes it a planet. If we think it’s a puny upstart 
unworthy of being in the same class as Jupiter, we can define it out of that class. 
Planethood is just a matter of definition. Some people may prefer one definition to 
another, but such preferences are ultimately arbitrary. Sharp distinctions between truth 
and falsity have little or no basis. Knowledge is a matter of preferences and opinions, 
which are ultimately a matter of taste. Thus a subjectivist epistemology takes the domain 
of taste as paradigmatic of knowledge, if we even call it that. Subjectivism can take root 
in very general ways, with profound consequences for people’s lives (Chandler, 1987; 
Chandler et al., 1990). 
But subjectivists, in some cases, construct a rationalist epistemology that 
coordinates objectivity and subjectivity. Despite your encounter with the perils of Pluto 
you may reflect on subjectivity itself and find some basis for objectivity. The 
classification of Pluto and the number of planets do indeed depend on the definition of 
planet. Definitions are matters of conventional usage, rather than empirical claims, and 
thus cannot be evaluated as true or false. Given that no definition is the one true 
15 
 
definition, there is no simple objective answer to the question of whether Pluto is a planet 
or the question of the number of planets in our solar system. But it does not follow that 
any answer is as defensible as any other. Some definitions, and thus some conclusions 
about planethood, are more justifiable than others. When the definition of planet was 
changed in 2006, this was not done on the basis of new empirical evidence about Pluto, 
but it did take into account new discoveries concerning the existence and nature of other 
astronomical objects in our solar system. Based on the accumulating evidence, there 
appeared to be no basis for classifying Pluto as a planet and excluding other objects, such 
as asteroids, that were as large as Pluto and no different in any characteristic that could 
reasonably be deemed relevant to planethood. Thus any definition that included all nine 
traditional planets would include additional objects, some already known and others 
perhaps yet to be discovered, with the result that there were at least 12 planets, and 
perhaps many more, in the solar system. It was possible, however, to devise a 
parsimonious definition that included eight of the traditional nine planets, all but Pluto, 
and no other object that was already known or likely to be discovered. The official 
adoption of this definition by the International Astronomical Union did not make it the 
one true definition in any objective sense and thus does not justify an objectivist 
conclusion that we used to mistakenly believe there were nine planets including Pluto but 
now know there to be only eight. But the subjectivity of the definition does not justify a 
subjectivist conclusion that all definitions are equally good, or equally arbitrary, or the 
associated subjectivist conclusions that any object can be a planet or not, and that the 
number of planets in the solar system can be whatever we want it to be. We could retain 
the traditional set of nine planets by defining a planet as (a) any object that has the unique 
16 
 
combination of characteristics of any one of the nine objects that were considered planets 
in the late 20th century or (b) any object that either meets the new 2006 definition or has 
a five-letter name beginning with “P.” Such definitions would not be false. They are not 
inconsistent with evidence. Their arbitrariness, however, renders them less defensible 
than more parsimonious definitions based on astronomical criteria such as orbit, size, and 
gravitation. Thus we can reasonably conclude that there are precisely eight planets or 
that there are at least 12, but there appears to be no reasonable basis for concluding that 
there are nine. In deciding on a definition, astronomers were making a subjective choice 
within objective constraints. The rationalist sees this as typical of knowledge, thus 
situating knowledge within the domain of interpretation, neither purely objective nor 
purely subjective. 
The Epistemology of History 
Consider another example: history. A child of 9 or 10 years would readily 
understand the distinction between history and fiction. Both history and fiction involve 
stories but historical stories are true whereas fiction is not. If a work of history tells about 
Abraham Lincoln conquering Japan and freeing the Jews, it is presenting false 
information. Novels and short stories, in contrast, can tell any story they want. If you 
don’t like the story, you are entitled to your opinion, but someone else may reasonably 
hold a different opinion. History is a matter of truth, whereas fiction is a matter of taste. 
Of course, our beliefs about what happened in the past can be distinguished from what 
actually happened, and the former may not correspond to the latter. But this only means 
we have made a mistake. It does not undermine the basic assumption that history is a 
17 
 
matter of knowledge and that historical claims can thus, at least in principle, be judged 
true or false. 
Consider a seemingly simple question: How many genocides have there been in 
human history? It might appear that, at least in principle, we could answer this question 
by determining whether each of the events in history was or was not a genocide and then 
counting the number of genocides. We might be unable to reach a definitive conclusion 
because there are historical events of which we are unaware, or about which we have 
insufficient knowledge to determine whether they should be classified as genocides. 
Even if we had substantial knowledge about every event in history, we might mistakenly 
classify some events due to gaps or inaccuracies in our knowledge. In principle, 
however, as the objectivist sees it, there is a right answer to the question of whether any 
given event was a genocide, and thus a right answer to the question of how many 
genocides have occurred in history. 
But history cannot be objectively divided into some finite number of discrete 
events. Even to the extent that we can distinguish a particular event, whether it qualifies 
as a genocide depends on our definition of genocide. The literature of genocide is rife 
with disputes over the definition and conceptualization of genocide and the classification 
of various historical events. There is, it seems, no objective basis for parsing the flow of 
history into some specific number of discrete events and no objective basis for 
determining which such events, if we even agree in distinguishing them, constitute 
genocide. Definitions of genocide differ with regard to dimensions such as the nature of 
the perpetrator, the intent of the perpetrator, the nature of the victim group, the means of 
destruction, and the totality of destruction. Whether some portion of history can be 
18 
 
differentiated as a discrete event, and whether that event is classified as genocide, is in 
the eye of the historian. Ultimately we see whatever stories we create for ourselves, in 
history as in fiction, and if some people prefer some stories to others, there is no reason to 
question their choices. Similarly, if some prefer to think of Pluto as a planet, or not, or 
prefer to believe there are some number of planets rather than some other number, who is 
to say who is right? Over the course of many such reflections in diverse domains, an 
objectivist may construct a subjectivist epistemology and thus become a subjectivist. 
Further consideration of definitions of genocide may, at least initially, reinforce a 
subjectivist conception of knowledge. The legal definition of genocide in the 1948 
United Nations Genocide Convention looks like a cobbled-together set of weirdly worded 
political compromises, which indeed it is. If you don’t like it, however, no problem: you 
can choose a definition more to your liking. There are many to choose from, and they 
differ along many dimensions. If you are concerned about undermining the status of the 
Holocaust by acknowledging too many genocides, there are definitions that recognize 
only the Holocaust, or the Holocaust and a small number of other events, as genocides. If 
you are concerned about some other set of events that seem to you genocidal, you can 
find a definition that includes them. If you are concerned that some group is being 
falsely labeled a perpetrator of genocide, you can find a definition under which its 
actions, however unfortunate, are not genocidal. Defining genocide, it appears, is an 
arbitrary and political matter, and thus the entire study of genocide is necessarily arbitrary 
and political. If even astronomy and even such basic factual matters as defining, 
identifying, and counting planets is subjective, a function of definitions and 
conceptualizations, it is not surprising that matters of history would be subjective too. 
19 
 
But further examination of definitions and conceptualizations of genocide shows 
some degree of consensus and some indications of objectivity, or at least rationality 
(Moshman, 2001, in press). Among scholars of genocide, there appears to be a consensus 
that the 1948 Genocide Convention was strongly influenced by the politics of the era and 
that its definition is deeply flawed, at least for purposes of research in history and the 
social sciences. There appears to be a consensus that what is needed is a more 
parsimonious definition that hasn’t been politically gerrymandered to include or exclude 
specific historical events of special concern. There appears to be a consensus that not just 
any definition will do and that some definitions are better than others. The lack of 
consensus on a specific definition is consistent with the subjectivist insight that there is 
no one true definition, but there are a small number of commonly used definitions that 
overlap substantially and there is agreement on the genocidal nature of many historical 
events, and on the nongenocidal nature of many others. 
To the extent that diverse conceptions of genocide provide divergent perceptions 
of history, moreover, we can enhance our understanding of history by recognizing and 
coordinating those multiple perspectives. Some acts constitute genocide under every 
available definition, but many other acts of group destruction and mass killing are or are 
not genocides depending on what definition one uses. From an objectivist point of view, 
this might seem a serious problem and it might seem the study of genocide must either 
find the correct definition or acknowledge that genocide research is only a source of 
idiosyncratic perceptions and perspectives, not knowledge. A rationalist, in contrast, sees 
the diversity of reasonable perspectives as a potential source of objectivity. To the extent 
that definitional disagreements focus our attention on issues of perpetrator, intention, 
20 
 
victim, process, and outcome the dimensions along which they differ we may better 
understand the actual events and their genocidal nature, and thus, without resolving the 
conceptual disagreements, attain greater objectivity through reflective coordination of our 
multiple subjectivities. 
With “genocide” as with “planet,” there may be more than one way to define what 
we have in mind, but not just any definition will do. Even to the extent that this leaves us 
with diverse definitions, moreover, our ability to coordinate and reflect on multiple 
definitions and conceptualizations may provide us with a metasubjective form of 
objectivity. To the extent that we recognize such a post-skeptical form of objectivity, we 
are constructing a rationalist epistemology. But in a world of epistemic peril, this may be 
a never-ending enterprise. 
Conclusion 
Epistemic development is the development of knowledge about knowledge. A 
central thesis of this chapter, following Chandler et al. (2002), has been that two levels of 
epistemic development can be distinguished, the first corresponding to childhood and the 
second to adolescence and adulthood. Childhood, I have suggested, is marked by the 
ongoing differentiation of three domains: a domain of truth, in which truth and falsity can 
be sharply distinguished; a domain of taste, in which no such distinction is possible or 
meaningful; and a domain of interpretation, in which the subjectivity of the mind is 
constrained by the reality of the object. Epistemic development in adolescence and 
adulthood, I have suggested, is largely a matter of determining which of these three 
domains is the domain of knowledge. Such development is neither universal nor tied to 
21 
 
age, but it tends to proceed in the familiar sequence extending from objectivist to 
subjectivist to rationalist epistemologies. 
We should be wary, however, of assuming there is a highest stage of epistemic 
development that we not only have attained but have fully understood. Rather than end 
on a complacent note of epistemic maturity, consider one last astronomical poser: What 
was the sixth planet to be discovered (and who discovered it, and when)? Thinking of 
discovery as an empirical process of finding what is there, one might readily surmise that 
the sixth planet to be discovered was likely the sixth from the sun, Saturn. But Saturn 
(along with Mercury, Venus, Mars, and Jupiter) is one of the classic five planets that have 
been known at least since the ancient Greeks. The sixth planet to be discovered was the 
planet Earth, which was discovered by Nicolaus Copernicus in the early 1500s when he 
determined that the five known planets all orbited the sun and that the Earth did as well, 
thus making it a planet and bringing the total number to six. 
We need not look as far as Pluto, it seems, for epistemological trouble. There is 
an obvious sense in which the Earth was known to people long before Copernicus, who 
thus cannot be credited with discovering it. But there is also a sense in which the Earth 
as we know it is the planet Earth, which was indeed unknown prior to its discovery by 
Copernicus. Discovery is conceptual as well as empirical, rendering truth uncomfortably 
problematic. Development may tend toward rationalist epistemologies but we should not 
assume there exists much less that we have attained a mature and final state of 
knowledge about the nature and justification of knowledge. 
22 
 
References 
Carpendale, J. I., & Chandler, M. J. (1996). On the distinction between false belief 
understanding and subscribing to an interpretive theory of mind. Child Development, 
67, 1686-1706. 
Chandler, M. J. (1987). The Othello effect: Essay on the emergence and eclipse of 
skeptical doubt. Human Development, 30, 137-159. 
Chandler, M., Boyes, M., & Ball, L. (1990). Relativism and stations of epistemic doubt. 
Journal of Experimental Child Psychology, 50, 370-395. 
Chandler, M. J., Hallett, D., & Sokol, B. W. (2002). Competing claims about competing 
knowledge claims. In B. K. Hofer & P. R. Pintrich (Eds.), Personal epistemology: 
The psychology of beliefs about knowledge and knowing (pp. 145-168). Mahwah, NJ: 
Erlbaum. 
Fabricius, W. V., & Schwanenflugel, P. J. (1994). The older child’s theory of mind. In 
A. Demetriou & A. Efklides (Eds.), Intelligence, mind, and reasoning: Structure and 
development (pp. 111-132.). Amsterdam: North-Holland. 
King, P. M., & Kitchener, K. S. (1994). Developing reflective judgment. San Francisco: 
Jossey-Bass. 
King, P. M., & Kitchener, K. S. (2002). The reflective judgment model: Twenty years of 
research on epistemic cognition. In B. K. Hofer & P. K. Pintrich (Eds.), Personal 
epistemology (pp. 37-61). Mahwah, NJ: Erlbaum. 
Komatsu, L. K., & Galotti, K. M. (1986). Children's reasoning about social, physical, 
and logical regularities: A look at two worlds. Child Development, 57, 413-420. 
Kuhn, D. (2005). Education for thinking. Cambridge, MA: Harvard University Press. 
23 
 
Kuhn, D., Cheney, R., and Weinstock, M. (2000). The development of epistemological 
understanding. Cognitive development, 15, 309-328. 
Kuhn, D. & Franklin, S. (2006). The second decade: What develops (and how)? In D. 
Kuhn & R. Siegler (Eds.), Handbook of child psychology, Vol. 2: Cognition, 
perception, and language (6th ed.) (pp. ) (W. Damon & R. Lerner, series eds.). 
Hoboken, NJ: Wiley. 
Kuhn, D., & Weinstock, M. (2002). What is epistemological thinking and why does it 
matter? In B. K. Hofer & P. R. Pintrich (Eds.), Personal epistemology: The 
psychology of beliefs about knowledge and knowing (pp. 121-144). Mahwah, NJ: 
Erlbaum. 
Lalonde, C. E., & Chandler, M. J. (2002). Children's understanding of interpretation. 
New Ideas in Psychology, 20, 163-198. 
Mansfield, A. F., & Clinchy, B. M. (2002). Toward the integration of objectivity and 
subjectivity: Epistemological development from 10 to 16. New Ideas in Psychology, 
20, 225-262. 
Miller, S. A., Custer, W. L., & Nassau, G. (2000). Children’s understanding of the 
necessity of logically necessary truths. Cognitive Development, 15, 383-403. 
Moshman, D. (1990). The development of metalogical understanding. In W. F. Overton 
(Ed.), Reasoning, necessity, and logic: Developmental perspectives (pp. 205-225). 
Hillsdale, NJ: Erlbaum. 
Moshman, D. (2001). Conceptual constraints on thinking about genocide. Journal of 
Genocide Research, 3, 431-450. 
24 
 
Moshman, D. (2004). From inference to reasoning: The construction of rationality. 
Thinking & Reasoning, 10, 221-239. 
Moshman, D. (2005). Adolescent psychological development: Rationality, morality, and 
identity (2nd edition). Mahwah, NJ: Erlbaum. 
Moshman, D. (in press). Conceptions of genocide and perceptions of history. In D. 
Stone (Ed.), The historiography of genocide. Hampshire, UK: Palgrave Macmillan. 
Moshman, D. (forthcoming). The development of rationality. In H. Siegel (Ed.), Oxford 
handbook of philosophy of education. Oxford, UK: Oxford University Press. 
Nicholls, J. G., & Thorkildsen, T. A. (1988). Children's distinctions among matters of 
intellectual convention, logic, fact, and personal choice. Child Development, 59, 939- 
949. 
Perry, W. G, (1970). Forms of intellectual and ethical development in the college years: 
A scheme. New York: Holt, Rinehart, and Winston. 
Piaget, J. (1987). Possibility and necessity. Minneapolis: University of Minnesota Press. 
Piaget, J. (2001). Studies in reflecting abstraction. Hove, UK: Psychology Press. 
Piaget, J. (2006). Reason. New Ideas in Psychology, 24, 1-29. 
Piéraut-Le Bonniec, G. (1980). The development of modal reasoning: Genesis of 
necessity and possibility notions. New York: Academic Press. 
Pillow, B. H. (2002). Children’s and adults’ evaluation of the certainty of deductive 
inferences, inductive inferences, and guesses. Child Development, 73, 779-792. 
Pillow, B. H., & Anderson, K. L. (2006). Children’s awareness of their own certainty 
and understanding of deduction and guessing. British Journal of Developmental 
Psychology, 24, 823-849. 
25 
 
Pillow, B. H., & Henrichon, A. J. (1996). There's more to the picture than meets the eye: 
Young children's difficulty understanding biased interpretation. Child Development, 
67, 803-819. 
Rai, R., & Mitchell, P. (2006). Children’s ability to impute inferentially based 
knowledge. Child Development, 77, 1081-1093. 
Rowley, M., & Robinson, E. J. (in press). Understanding the truth about subjectivity. 
Social Development. 
Schwanenflugel, P. J., Fabricius, W. V., & Noyes, C. R. (1996). Developing 
organization of mental verbs: Evidence for the development of a constructivist theory 
of mind in middle childhood. Cognitive Development, 11, 265-294. 
Smith, L. (1993). Necessary knowledge: Piagetian perspectives on constructivism. 
Hillsdale, NJ: Erlbaum. 
Sodian, B. & Wimmer, H. (1987). Children’s understanding of inference as a source of 
knowledge. Child Development, 58, 424-433. 
Wainryb, C., Shaw, L. A., Langley, M., Cottam, K., & Lewis, R. (2004). Children's 
thinking about diversity of belief in the early school years: Judgments of relativism, 
tolerance, and disagreeing persons. Child Development, 75, 687-703. 
Wellman, H. M., Cross, D., & Watson, J. (2001). Meta-analysis of theory-of-mind 
development: The truth about false belief. Child Development, 82, 655-684. 
26