










































ANALYTIC TEACHING AND PHILOSOPHICAL PRAXIS VOLUME 41, ISSUE 2 (2021) 

  
 

32 
 

 

How Can Rationality Empower Children? 

 
 

Anne Brel Cloutier   
 
 

 
  
f you had superpowers, would you use them to help yourself or to help the world? When asked 
this question, this 10 year-old girl made the following deduction: “Since I am part of the 

world, if I do something to help the world, then I am helping myself.” She avoided the false dilemma 
fallacy in which you have to choose between only two options. She was able to reason in such a way as 
to produce a third option without compromising the first two. This is not only a logical deduction, 
but also the type of reasoning that the world arguably needs for positive progress on environmental 
and political issues. Once people understand that helping the world is helping humanity and that they 
are, themselves, part of humanity, they might feel that is it their duty to do whatever they can to help 
the world.1 
 

Some might assume that children are too young to have rational thoughts and that philosophy 
might be more likely to confuse them than achieve any good. In more traditional views of education, 
the goal is for them to learn facts, rules and techniques. Some may fear that asking them to put things 
in perspective or challenge the foundations of their knowledge could lead to relativism and cause 
harm to their development. A possible cause may be the theoretical work of Jean Piaget, the first 
psychologist to study reasoning from a logician’s point of view as well as the first to conduct studies on 
rationality based on direct observations of children. His theory on the development of rationality 
claims that children are not born logical and that logical reasoning only appears progressively in 
adolescence. Though his theory has been criticised on numerous fronts, the false belief about children 
being illogical has remained strongly anchored in our conceptions of rationality.2 

 
In this article, we will argue that since Piaget, many advancements have been made in 

psychological studies about reasoning. Even though some parts of his theory are still very helpful to 
understanding the development of rationality, it is a mistake to view children as illogical beings. On 
the contrary, they happen to be logical at a very young age, as our argument and examples will show. 
Educational psychologist David Moshman offers a new reading of Piaget’s work by explaining the 
development of rationality at a metalogical level, not only in childhood but in adulthood as well. 
Metacognition is the mental act of thinking about thinking and is a fundamental cognitive process for 
the development of reasoning. 

 

                                                 
1 This example is drawn from a summer philosophy camp run in July 2014 by Brila (www.brila.org), an educational charity 
based in Montreal, Canada. The participant’s name has been omitted to protect her privacy. 
2 Moshman, 1998-a. 

I 

http://www.brila.org/


ANALYTIC TEACHING AND PHILOSOPHICAL PRAXIS VOLUME 41, ISSUE 2 (2021) 

  
 

33 
 

First, we will briefly summarize the main theories of reasoning, introducing Moshman’s pluralist 
rational constructivism theory. 3 On his view, rationality develops through our increased metacognitive 
understanding, which occurs during peer interactions. We will therefore argue that the Community of 
Philosophical Inquiry (CPI) method that forms the basis of the Philosophy for Children (P4C) 
program, when practiced with a special focus on metacognition, can constitute the perfect pedagogical 
approach to put Moshman’s theory into practice. A close examination of some of the behaviours that 
occur in a CPI can link to his theoretical approach of rational development: In P4C, not only do we 
regularly witness children expressing rational and logical thoughts, but we also see how the CPI’s 
metalogical aspects offer multiple strategies that can help foster their development. These 
metacognitive and metalogic strategies come from observations of children practicing P4C dialogues, 
from which adults could greatly benefit too.  
 
Developing Rationality: From Piaget to Moshman—Logic and Metalogic 

 
Piaget’s contribution to the psychology of reasoning is significant. He was the first to combine 

logic and psychology, and the first to understand reasoning as developing during childhood through 
understanding rather than the mere memorizing of facts. According to Piaget, we are born without 
logical capacities. At first, children do not have access to their cognitive operations—for example, the 
conservation of quantities. At four or five years old, children will perceive a flattened piece of dough 
as bigger than its original form, only because it seems larger. The same thing occurs with liquids: a 
single amount of juice presented in two different glasses will be interpreted as different because of the 
appearance of the glasses. At this age, children do not understand the transformation, even if it occurs 
in front of them.4 They therefore tend to explain the transformation by reference to magic. This clash 
between the phenomenon and their interpretation represents a logical contradiction. At first, children 
do not even realize the contradiction, but at about five or six, they start to sense a cognitive conflict. 
This conflict is essential to their development: it will motivate them to work toward understanding the 
origins of the contradictions.  

 
Different factors help children develop this understanding: maturity, repeated experience and 

social transmission. They repeat the operation until they finally understand the transformation that 
occurred—this is the condition for them to acquire reversibility, or the capacity to think through steps 
in reverse direction. However, they must be mature enough to possess the operational structure 
required for understanding and they need help from their social environment to be confronted to 
their contradictions. Balance between all these factors will eventually resolve the conflict; children 
then complexify their structure of understanding, acquiring a higher level of logical skill.5 As we will 
see in Moshman’s reading of Piaget, the principles of conflict and self-regulation are key to ensuring 
deeper levels of reasoning.  

 
Piaget identifies four stages of logical understanding, the highest of which is acquired at 

adolescence and remains optimal in adulthood—namely, the formal operational stage. This aspect of his 
theory is problematic since subsequent research has shown that children are able to accomplish logical 

                                                 
3 Moshman, 2004. 
4 See Inhelder, Sinclair & Bovet in Piaget, 1974. 
5 Piaget, 1974. 



ANALYTIC TEACHING AND PHILOSOPHICAL PRAXIS VOLUME 41, ISSUE 2 (2021) 

  
 

34 
 

tasks at an earlier age than Piaget predicted.6 Moreover, adult rationality is hardly always optimally 
logical: psychologist Peter Wason developed a selection task to detect logical fallacies that most adults 
cannot avoid committing, namely the affirmation of the consequent. This logical fallacy consists in 
understanding a conditional (A implies B) as if it were an equivalence (A equals B). The conditional 
“If A, then B” does not imply that if B (the consequent) occurs, then A (the antecedent) will occur 
too. In contrast, for the equivalence “A equals B,” when B occurs, A will occur too. Applied to 
children, this selection task could take the form, “If you want dessert, you have to eat your vegetables.” 
Empirical studies show that children will understand that in order to get dessert, they must eat their 
vegetables, but also that even if they ate their vegetables, they do not have to eat dessert (affirming the 
consequent does not imply that the antecedent will occur). In this case, they will eat dessert only if 
they want to, but eating the dessert is not part of the rule, only eating vegetables is. This task has the 
same logical form as the Wason task, but with different content. If the development of reasoning 
consists in the acquisition of logical skills, then some empirical studies would prove that kids have a 
logical understanding while Wason’s selection task shows that adults do not.7 Would logical 
understanding drastically drop with age? This does not seem entirely accurate.  

 
To explain these variances in logical understanding, Moshman points to the distinction between 

logic and metalogic. Logic requires us to be able to make simple inferences in familiar contexts, 
whereas metalogic requires us to be able to think and coordinate multiple inferences in unfamiliar 
contexts. We may be able to make valid logical inferences at a very young age, but reasoning is not 
restricted merely to capacity of making inferences.8 Children are perfectly capable of making logically 
valid inferences; they just do not know it yet. The development of reasoning requires awareness of 
inferences and explicit metalogical thinking. For Moshman, reasoning begins with logic and 
progresses without a final maturity state.  Reasoning begins with the capacity to make inferences, then 
develops into the capacity to apply these inferences to specifics goals—what Moshman calls thinking. 
Thinking then develops into the ability to coordinate different inferences, thus creating new 
knowledge—what he calls reasoning. Reasoning is thinking with epistemic prudence, that is, with 
concern for accuracy and adherence to appropriate inferential norms. 9 

 
 
 

                                                 
6 Hawkins & al., 1984; Markovits & al., 1989. 
7 A version of his selection task consists of showing four cards, each holding a letter on one side and a number on the 
other side. Since we can only see one side of each card, the subject sees cards with an A, a 7, a B and a 4. The experimenter 
gives a rule to the participant: ”If there is an A on one side, there is the number 4 on the other side.” Then, the participant 
is asked which card must be turned over to make valid or invalid the rule. Most of the participants choose to turn over the 
“A” card (which is correct) and the “4” card (which is a fallacy). In fact, in order to verify if the rule has been respected, 
one must turn over the “7” card, to make sure that there is no “A” on the other side since the rule applies to what is 
behind a “A” cards, but doesn’t mention anything about what is on the other side of a “4” cards. This selection task, when 
presented to adults is rarely correctly solved, even with subjects having a IQ higher than average (In « Regression in 
reasoning? » (Wason, 1969), one of the participants who failed at the selection task admitted being a member of Mensa, 
the International High IQ Society.). This demonstrates that logical reasoning is neither relative to age nor to competencies. 
8 Logical tasks correctly performed by young children are presented in familiar context, whereas correctly performed 
selection tasks, like Wasons’, appear in a formal context that requires a fully developed understanding of logical inferences, 
in order to make predictions and to coordinate diverse possibilities. 
9 Moshman, 2004, p. 224. 



ANALYTIC TEACHING AND PHILOSOPHICAL PRAXIS VOLUME 41, ISSUE 2 (2021) 

  
 

35 
 

Theories of Human Reasoning 

 
Evolutionist theories claim that we evolved with heuristics: although these have assured our 

survival, they have not always been logically correct. We reason in an adaptive rather than a logical 
way.10 This does not mean we never think logically, only that we sometimes make mistakes. We can 
acknowledge this as a step in the right direction, given some cognitive science and psychological 
theories offer an ameliorative view claiming that even if we produce incorrect answers, we have the 
capacity to correct them.11 

 
Evolutionary dual process theories offer a developmental view of reasoning where even if the 

biases and heuristics are really difficult—and in some cases, impossible—to avoid, we have the capacity 
to control them. 12 Reasoning comes from two essential processes that generally fall into two groups of 
proprieties. The first type, called System 1 (S1), generates heuristics, automatic and unconscious 
processes and do not require a lot of cognitive effort. A second type, called System 2 (S2) includes 
processes that are analytic, controlled and conscious, and require a lot of cognitive effort. Reasoning 
comes from heuristics that made us evolutionarily adapted, but not to the present time where 
evolution has moved too fast for us to adapt. Social, technical and scientific evolution grows a lot 
faster than the selection of our genes. This would explain why we are not necessarily adapted in such a 
way as to always provide good reasoning. S1 processes come from our evolution and even though they 
sometimes produce good conclusions, they are responsible for our mistakes too. S2 processes are more 
recent and they give us the capacity to control and correct S1 processes.13 For example, if one 
produces a logically invalid deduction, S2 processes can evaluate the answer and create a new one 
until the good answer is achieved and the task can then be successfully completed. 

 
In order to use S2 processes, one needs to possess adequate knowledge to produce logical 

answers. However, biases coming from S1 processes are deeply anchored in our reasoning; therefore, 
we cannot eliminate them, we can only inhibit them. This is why good learning consists not only in 
acquiring knowledge, but also in practicing strategies and developing awareness about possible biases 
that can come from S1 processes. But having strategies without knowing when to use them is 
pointless.14 We must learn how to detect situations in which they are needed. Thus, practicing 
epistemic prudence is as important as knowledge and strategies.15 Epistemic prudence consists in 
restraining ourselves from asserting with certainty when we do not hold enough information to 
conclude. Generally, this behaviour is observed when there is a possibility that one might have to 
override heuristic responses.  

 
In neuroscience, brain imaging shows that repeating rules does not help us avoid biases. We 

must be alerted to the possibility of bias in order to detect and override it.16 Identifying mistakes and 
                                                 
10 Gigerenzer & Goldstein, 1996, Cosmides, 1989. 
11 Stanovich, Toplak & West, 2008 in Beaulac & Robert, 2011. 
12 Dual process theories conceive reasoning as coming from two different systems in which different processes override 
others, allowing us to correct our incorrect answers. See: Evans, J.St. B.T. & K. Frankish, 2009 ; Stanovish & West, 2008. 
13 Samuels, 2009 in Evans, J.St. B.T. & K. Frankish, 2009 p. 131. 
14 Ibid, p. 687. 
15 Beaulac & Robert, 2011. 
16 Houdé et Moutier, 1996. 



ANALYTIC TEACHING AND PHILOSOPHICAL PRAXIS VOLUME 41, ISSUE 2 (2021) 

  
 

36 
 

strategies to avoid them helps us achieve better results when solving tasks, with the success rate rising 
from 10 to 90 per cent.17 We believe that from a dynamic cognitive point of view, neuronal activities 
can be produced by epistemic prudence, meaning once it is internalized, it becomes an automatic 
alert. In this article, we would like to suggest that metacognition is the procedure of coordination to 
override S1 processes. First, we need epistemic prudence to activate S2 processes; then, reflective 
reasoning—consisting of monitoring and coordinating inferences—results in overriding S1 processes 
toward a more thoughtful answer. We view the actual activation of S2 as a metacognitive process. 
 

For Moshman, reasoning is much too complex to be reduced to only two different processes. 
However, we will see that his constructivist theory of rationalist pluralism suggests that peers’ 
interactions is an effective way to put into practice cognitive strategies and epistemic prudence as part 
of a collective and internal metacognition act. 

 
Rationalist Pluralist Constructivism 

 

For both Moshman and Piaget, rational development begins in conflict. When faced with a 
contradiction between a phenomenon and its representation, children feel a discomfort that 
motivates them to understand what is missing in their reasoning. To find equilibrium, they must seek 
to understand the nature of the contradiction. Only then will they be able to achieve more complex 
forms of reasoning. Piaget’s experiments show that a more complex understanding structure can be 
achieved much faster when children repeat the operations by themselves.18 Similarly, for Moshman, 
children can achieve more complex reasoning by repeating the operation, but this recurrence can be 
understood as a metacognitive reflection on the reasoning experience. The development of reasoning 
progresses through explicit metacognitive reflections on their own inferences to make them clearer 
and more understandable.19 

 
We build our own metacognitive theory at a very young age, even if it is not fully conscious at 

first. Though we understand a metacognitive act as a conscious one, as children, we are able to 
understand how behaviours may impact our environment, even if we do not consciously view it as 
metacognitive understanding. This comprehension is therefore the tacit starting point of children’s 
own metacognitive theory construction. Making an inference does not require metacognitive 
awareness, yet inferring that “When I do this, that will happen” and acting accordingly constitute a 
form of metacognitive theory. Since a theory is simply the active use of knowledge to predict and 
explain empirical phenomena, developing our own metacognitive theory then consists in being able to 
build our own knowledge structure—it is a growing capacity to theorize and act on our own cognitive 
behaviours. The more we understand our thinking and organize it in reasoning structures, the more 
complex our comprehension structures will become. Thus, metacognition increases our level of 
reasoning, giving us more control on further reasoning—and our rationality development depends on 
it.20  

 

                                                 
17 Ibid. 
18 Piaget, 1974, p. 59. 
19 Moshman, 2004. 
20 Schraw & Moshman, 1995, p. 356. 



ANALYTIC TEACHING AND PHILOSOPHICAL PRAXIS VOLUME 41, ISSUE 2 (2021) 

  
 

37 
 

By becoming aware of our inferences, we are able to coordinate and control them appropriately.  
For Moshman, awareness and control of our inferences take place most often in social interactions: 
when we interact, we must justify ourselves and this justification process requires making our 
argument explicit. As he writes, “Reflection and coordination, moreover, often take place in the 
context of social interaction, and especially peer interaction.”21 

 
Moshman’s theory of rational development is based on a dialectic constructivist account, which 

holds that the construction of reasoning results from an interaction between external factors—like 
learned strategies and social environment—and internal processes—like maturation. For Moshman, 
metacognition depends both on the social environment and internal factors, so they complete each 
other, making both of them fundamental and richer. Internal construction explains metalogical 
comprehension, giving the capacity to understand foundations of logical strategies and to have a 
richer grasp on their integration and coordination. In the same way, external factors may help learners 
to have a better metalogical understanding, thus helping them to coordinate and reconstruct their 
thoughts at a deeper level. 

 
Control and coordination of inferences is a conscious metacognitive process made possible by 

three factors: cultural learning, individual construction and peer interaction.22 Through social 
learning, children internalize conceptions about the nature of cognition. Individual construction is a 
spontaneous attitude towards learning, in order to use strategies in an efficient way.23 Self-regulation, 
for example, could be an internal process used to improve their comprehension of their own 
cognition. When dialoguing in social contexts, we interact to justify ourselves, so that we may 
understand each other better. In order to be better understood, we must make our argument clear to 
others, but first of all, to ourselves. Most of the time, it is while arguing that we realize the weaknesses 
and strengths of our arguments, making dialogue crucial not only to being well understood but also to 
becoming aware of our own inferential processes as a metacognitive act. 24 Social interactions bring 
about the need to understand our own reasoning and others, to combine them and coordinate them 
toward a shared reality. 

 
Empirical studies have shown that collective thinking leads to better and more sophisticated 

answers.25 Dialogue generates better understanding of problems and better clarification of concepts, 
therefore facilitating the resolution of sophisticated tasks. However, let us imagine a situation where a 
child is dialoguing with her tutor. There is a very low possibility that she will doubt his conclusions 
since he has epistemic authority. In order to question her reasoning, she would need to interact with 
peers she deems as equals in terms of rational skills. 

 
                                                 
21 Moshman, 2004, p. 233. 
22 Schraw and Moshman, 1995, p. 362. 
23 Ibid., pp. 363-364. 
24 Moshman, 2004, pp. 233-234. 
25 For example, the rate of success rises from 9% to 75% for Wason selection task in collective thinking (Geil & 
Moshman, 1994 in Schraw & Moshman, 1995, p. 364). In this specific experiment, in order to make sure that the group 
was not relying on a group’s expert, some data reveal good results for groups that found the good answer collectively 
without anyone having found it individually. This shows the importance of a level of equality between participants. For an 
adequate coordination and reflection process, inferences must be understandable for every member of the group. 



ANALYTIC TEACHING AND PHILOSOPHICAL PRAXIS VOLUME 41, ISSUE 2 (2021) 

  
 

38 
 

Reasoning is much more complex than applying mere logical rules—it requires dialectical 
reasoning, or the ability to reflect on inferential diversity. Dialectical reasoning represents a mindful 
and intentional effort to reconstruct our own rules, principles, intuitions and conceptions in order to 
achieve coherence and conceptual progress. 

 
We may define dialectical thinking as the deliberate coordination of inferences for the 
purpose of making cognitive progress. Thus, the development of dialectical reasoning 
involves increasingly explicit knowledge about the nature of cognitive development, and 
increasingly deliberate efforts to further that process.26 

 
Since metacognition acts on itself, there is seemingly no limit to its development. Moreover, since 
awareness of mind begins at around three years old, the development of metacognitive understanding 
can start early in life and continue into adulthood.27  
 
CPI as a Pedagogy to Foster Logical Reasoning 

 

In light of the aforementioned theories, let us briefly summarize how logical reasoning works in 
five steps: 
 

• Learning adequate knowledge and strategies for logical answers. 
• Being alerted about the risk of being mistaken by possible biases. 
• Practicing epistemic prudence by doubting when we are not sure of an answer. 
• Thinking about our way of thinking, reviewing the information through 

a critical use of dialectical reasoning—that is, metacognition. 
• Correcting our mistakes when applying appropriate procedures toward an appropriate 
      answer. 

 

In this section, we will argue that the CPI method used in P4C is well suited to put the 
reasoning process into practice, because: i) its curriculum includes logical rules and procedures; ii) its 
dialogical approach regularly creates confrontations between different ideas and viewpoints, 
highlighting possible errors; iii) it fosters epistemic prudence, and iv) it constitutes in itself a 
metacognitive process. If practiced with metacognitive development in mind, P4C therefore stands a 
good chance of contributing to the development of logical reasoning, helping children grow in their 
capacity to apply appropriate procedures and arrive at appropriate answers. 

 
Logic in Education and P4C 

 

In the learning process, being open-minded is important in order to accept new information, 
but accepting too much without any process of doubting may lead to epistemological dogmatism. 
Questioning the information we receive is a necessity in order to protect ourselves from 
indoctrination. On the other hand, too much doubt can put us in a state of uncertainty, even when 
facing facts and empirical truth, thus risking epistemic relativism. Two types of logical errors are 

                                                 
26 Moshman, 1998-a, p. 961. 
27 Schraw & Moshman, 1995. 



ANALYTIC TEACHING AND PHILOSOPHICAL PRAXIS VOLUME 41, ISSUE 2 (2021) 

  
 

39 
 

possible in reasoning: fallacies, where we add too much information to the conclusion, and 
suppression of valid inferences, where we doubt too much and are not able to conclude anything. 
Reasoning logically is being able to open our mind to information with enough doubt to prevent 
fallacies and enough rules to prevent relativism. This is why logic is an important tool in education: it 
helps to open minds for learning while protecting children from indoctrination.  

 
This is not a way of insisting that children learn logic, but rather that the metalogical aspect of 

dialogue can enhance logical reasoning. Children who become progressively aware of the particular 
rules of reasoning and their impact on thoughtful expression gain a greater appreciation for the 
importance of developing thinking skills.28 A study by Robert & al. shows that P4C has a positive 
impact on the development of logical reasoning.29 Metacognitive and logic strategies in the P4C 
curriculum can help facilitators to guide and redirect children to enhance their thinking and help 
them develop their rationality on their own, while paying attention to the mental acts required to do 
so effectively.  

 
P4C and Epistemic Prudence 

 

Learning the rules of logic is not, as we have seen, sufficient to reasoning logically. Children 
must also be able to detect possible biases by becoming more alert and epistemically cautious. The 
Robert & al. study shows that children who have been practicing P4C for several years present 
significant signs of epistemic prudence.30 Doubting plays an important role in P4C inquiry in the 
development of logical awareness. In a CPI, children are not evaluated for their answers but for their 
capacities to inquire and question their own claims as well as those of others for the benefit of the 
inquiry. As P4C theorists Michel Sasseville and Mathieu Gagnon write, 
 

Doubting is a very important activity in a community of inquiry. To enter a community of 
inquiry is to make the decision to put one’s beliefs in check, that is, to accept that they may 
be questioned. For this to happen, it is important that we have the attitude of someone 
who, in the face of what they believe to be true, right or good, is willing to question some 
or all of their beliefs.31  
 

When doubting, children practice honest thinking, as they are not trying to convince others but 
rather striving to find the most valid and reasonable answer together.  
 

Inspired by socio-constructivism, which views knowledge as generated from collaboration and 
mutual aid,32 CPI dialogues represent a rationalist epistemological work of co-construction, that is, the 
children confront each other’s ideas in order to collectively evaluate arguments while building their 

                                                 
28 P4C’s curriculum provides logical tools, among others, Lipman’s novel Harry Stottlemeier’s discovery, for example, tells the 
story of children who discover the rules of logic by themselves by learning to avoid generalizations and other common 
logical mistakes. Also, exercises are included in the facilitator’s guidebook in order for children to practice logical notions 
seen in the story, such as deduction and avoiding fallacies. 
29 Robert & al, 2009, p. 21. 
30 A difference of more than 10% was noted among children who practiced P4C, compared to children who did not. 
31 Sasseville et Gagnon, 2012, p. 117. 
32 Sasseville & Gagnon, 2012, p. 85. 



ANALYTIC TEACHING AND PHILOSOPHICAL PRAXIS VOLUME 41, ISSUE 2 (2021) 

  
 

40 
 

own knowledge.33 Peer interactions raise objections and divergent interpretations about reality, thus 
engaging children in multiple reflections and negotiations, encouraging them to modify their own 
conceptions of knowledge.34 In time, children participating in a CPI become more aware of their 
active role in knowledge construction, notably the importance of collaborative thinking and 
responsibility for justifying philosophical positions.  

 
P4C and Metacognition 

 
Peer Interactions 

 
In a P4C setting, epistemic prudence favours metacognitive thinking in children since they must 

give reasons for the claims they make. This justification process contributes to metalogical thinking, 
requiring them to review their inferences and argumentation in order to evaluate them. When making 
their inferences explicit during a CPI dialogue, children gradually become aware of what an argument 
needs to be valid while group dynamics allow them to combine a diversity of inferences that improve 
one another, leading to more sophisticated forms of reasoning. As Moshman writes, 
 

In social contexts we may find ourselves challenged to justify our conclusions, and thus to 
recognize and justify our inferences. We may also be challenged to understand the 
inferential paths that led others to alternative views, and to coordinate those inferences and 
conclusions with ours.35 

 
This is precisely what children are doing in a CPI: they acknowledge their own reflection 

processes and strategies to improve the inquiry, motivating them to excel from a cognitive point of 
view. The goal of this practice is not to reach consensus but to cultivate autonomous thinking—
children are not only working to find thoughtful answers, but also to determine ways to achieve such 
answers.36 In the words of P4C co-founder Lipman, 
 

The improvement of thinking involves reflection. […] Reflective thinking is thinking that is 
aware of its own assumptions and implications as well as being conscious of the reasons 
and evidence that support this or that conclusion. Reflective thinking takes into account its 
own methodology, its own procedures, its own perspective and point of view.37 

 
A CPI practice requires children to think about their thinking, therefore activating a 

metacognitive process of self-correction toward the mindful application of thinking strategies. If an 

                                                 
33 Prevalent knowledge would be objectivist epistemology, where it consists in directly and empirically observable facts, like 
a large proportion of what children learn at school, for example. However, subjectivist epistemology has an important role 
to lay in the construction of knowledge since individuals acquire information according to their cultural and social 
specificities. Nevertheless, a radical subjectivism could lead to epistemological relativism. Rationalist epistemology requires 
to evaluate, criticize and justify different standpoint without asserting one single truth, therefore is more complete and 
support better inferences than objectivist and subjectivist taken separately. It links one to the other, building a more 
adequate theory of knowledge. (Moshman, 2004, p.224) 
34 Sasseville & Gagnon, 2012, p. 86. 
35 Moshman, 2004, p. 233. 
36 Sasseville, 2009, p. 5. 
37 Lipman, 2003, p. 26. 



ANALYTIC TEACHING AND PHILOSOPHICAL PRAXIS VOLUME 41, ISSUE 2 (2021) 

  
 

41 
 

argument is not valid, they must work to understand where it fails in order to correct it. When CPI 
dialogues emphasize awareness of reasoning processes, they therefore offer children the opportunity to 
develop their rationality, thereby collectively achieving more rigorous and coherent multidimensional 
thinking. The group dynamics enable children to think out loud without being afraid of making 
mistakes. Unlike traditional education that implores children to think before speaking, in P4C, 
speaking is viewed as a thinking-aloud strategy: children help each other detect mistakes and 
determine how to correct them. This approach challenges the dominant educational paradigms, 
depicting mistakes not as a source of shame but as useful tools for learning and for accepting human 
imperfection.  

 
Metasubjective Objectivity 

 

We often associate objectivity with proven facts and rigorous justification, whereas subjectivity is 
usually linked to emotions, individual perceptions and unjustified thoughts. Generally, subjectivity is 
opposed as objectivity. However, Moshman’s conception of rationality is a systematic interaction of 
both. Rationality is a form of objectivity that emerges from the reflective reconstruction of our 
subjectivity. Since our perceptions come from our senses, and we interpret them with our personal 
conceptions, they are in part subjective. On the other hand, the objects we perceive are parts of an 
independent reality regardless of our perceptions. Thus, for a more objective conception of reality, we 
must be aware of our different ways of perceiving. Since our perceptions reflect our subjectivity, but 
are supported by actual reality, conscious reflection about our subjectivity helps to increase its 
objectivity. As a result, our conceptions are built from subjective perceptions, but supported by real 
facts. Metasubjectivity consists in having reflective thoughts about our subjectivity. Therefore, 
interactions between metasubjective thoughts and objective facts create meta-subjective objectivity.38  

 
Metasubjective objectivity represents our self-reflection on our interactions with our 

environment, whether between subjects and objects or between subjects. The plurality of perspectives 
generates conflicts that motivate metasubjective reflections, thus increasing metasubjective 
objectivity.39 The collision of ideas in a CPI is important as it constitutes a conflict that triggers an 
active process of dynamic reflection that is then deepened by collaboration through multiple cognitive 
acts.  

 
A CPI setting is the perfect environment to practice metasubjectivity as it invites children to 

exchange ideas about topics that matter to them and justify their positions by making their arguments 
explicit. The CPI curriculum is also full of cognitive tools that help children in their argumentation, 
such as giving reasons, examples and counterexamples, formulating comparisons and analogies, etc. 
These mental acts help children improve their thinking so as to reach more complex thinking levels as 
they co-construct their own knowledge. 40 

 

                                                 
38 Moshman, 1994. 
39 This meets Piaget’s theory on cognitive conflict and the need to find balance in solving contradictions. 
40 The curriculum in P4C contains guidebooks that help teachers and facilitators to recognize these mental acts as they 
occur in discussions. They also provide philosophical guidelines to make sure that dialogues reach concepts that involve 
thinking rather than maintaining episodic discussions. 



ANALYTIC TEACHING AND PHILOSOPHICAL PRAXIS VOLUME 41, ISSUE 2 (2021) 

  
 

42 
 

Mental Acts 

 

In P4C dialogues, children learn to detect the various mental acts that are at play rather than 
have them imposed by adults.41 This bottom-up educational approach teaches children how to think 
rather than what to think, helping them become aware of how they can improve. By labelling mental 
acts as they occur during CPI dialogues, facilitators help to make arguments and strategies explicit to 
children.42 

 
In the dialogue excerpt below, we underline three different mental acts that also consist in metalogical 
strategies: 
 

• Pointing out assumptions enables a better understanding of assertions as they examine their 
foundations.43 Making these implicit premises explicit leads to a better understanding of what 
works in the argumentation and what does not. Therefore, it makes it easier to correct logical 
mistakes while also reinforcing beliefs in correct claims. 
 

• Identifying consequences requires thinking about the implications of a claim, thereby 
developing a better understanding of causality. As a metalogical strategy, understanding that 
there is a distinction between implication and equivalence is crucial to avoiding the 
affirmation of consequents and the negation of antecedent fallacies. The practice of 
identifying consequences allows for a better understanding of what an implication is. 
 

• Finding counterexamples is a metalogical strategy that prevents the induction fallacy, which is 
one of the logical fallacies that occurs the most and almost inevitably leads to prejudice. 
Counterexamples consist in generating alternatives, thus engaging deeper reflections before 
concluding. Giving children the opportunity to challenge their own presuppositions with 
counterexamples helps them understand the incoherence of prejudices. 
 

In this dialogue excerpt, Participant 2 points out Participant 1’s assumption, but in doing so, he 
denies the antecedent fallacy: 
- Participant 1: “A wealthy parent will provide a good future for his children.” 
- Participant 2: “I disagree! You’re saying that a poor parent won’t take care of his kids 
as well as a rich parent will?” 

Following the dialogue, Participant 1 prevents Participant 2 from committing the denial of antecedent 
Fallacy by identifying the true consequences of his thoughts: 

- Participant 1: “No, that’s not what I’m saying. But when you are wealthy, you can 
provide a house, food and education to your children.” 
 

                                                 
41 Tens of mental acts and behaviours have been listed in Sasseville & Gagnon, 2012.  
42 A recent study on expert dialogic facilitation suggests that facilitator moves like paraphrasing, locating, distilling, probing 
reasoning and identifying warrants help to clarify the inquiry process by underlining the argumentative features of 
children’s philosophical contributions.  Accordingly, it seems that facilitation moves that emphasize metacognitive are 
most effective at improving children’s reasoning. (See, Oyler, 2015.) 
43 Gagnon, 2005, p. 41. 



ANALYTIC TEACHING AND PHILOSOPHICAL PRAXIS VOLUME 41, ISSUE 2 (2021) 

  
 

43 
 

Then, Participant 2 gives a counter-example pointing out Participant 1’s assumptions about wealth: 
- Participant 2: “But a person can also be wealthy and not take good care of his 
children at all.” 
- Participant 3: “I am agreeing with both of you. Being wealthy does not mean that you 
will necessarily take good care of your children, but at least you could afford it.” 

 
When asked to give reasons, participants had to work on making the arguments explicit and 
accentuate their logical structure. For that to happen, they tried to review every step of their thought 
process and make the steps clear enough so that they could be explained to other participants. The 
reconstruction of an argument is a metalogic strategy that helps to gradually understand the 
distinction between premises and conclusions—and that the latter results from the former.  

- Participant 1: “Cars are a bad invention.” 
- Facilitator: “Why do you think that?” 
- Participant 1: “Well, they cause accidents and they pollute.” 
- Participant 2: “But they also help people to travel easily.” 
- Participant 1: “I agree, but if traveling easily involves people getting hurt and the 
environment getting damaged, then it does more damage than good. And if it makes 
more damage than good, then it’s a bad invention.” 

 
Making analogies is a metacognitive strategy that enables the development of abstract thinking by 
comparing actual situations to other situations, and in extracting the concept underlying them to 
apply them to both. Explaining abstract concepts in a more accessible way can help children to 
understand them better. 

- Facilitator: “What is intelligence?” 
- Participant 1: “Having good grades at school, I guess.” 
- Participant 2: “It could also be making new discoveries, like Darwin and Galileo did!” 
- Participant 3: “That’s true. And they did not make their discoveries in school.”  
- Participant 4: “But at their time, intelligence was not the same as today. Galileo was 
put in jail and we had proof of his intelligence with facts only years later.” 
- Participant 5: “I think that intelligence is like cars: models can always improve with 
new technology.” 
- Facilitator: “Are you saying that new technologies are to a car what proof and facts are 
to intelligence?” 
- Participant 5: “Yes, that’s it… they both improve and continue to move forward with 
time.” 

 
The Fishbowl  

 

These mental acts are only some of what we can observe in P4C. Participants put them into 
practice and regularly identify them as they move through the dialogue. However, identifying mental 
acts as they get used is not an easy thing to do. Lipman noted that we teach young children to 
recognize animals and colours by showing them pictures. In the same way, it should be possible to 
teach children—as well as adolescents and adults—to recognize a variety of mental acts by clearly 



ANALYTIC TEACHING AND PHILOSOPHICAL PRAXIS VOLUME 41, ISSUE 2 (2021) 

  
 

44 
 

identifying them when they occur.44 Thus, when participants activate a metacognitive process, it is not 
necessarily explicit. The fishbowl practice in P4C is an effective way to make the process explicitly 
metacognitive. 

 
Fishbowls consist in dividing a class of students in two subgroups: the observers and the 

participants. Observers have to keep silent during the whole dialogue. They observe if mental acts are 
indeed made or not, and note whenever they are, linking them to their context. Here is an example of 
the metacognitive processes involved in the fishbowl practice: 

Participants during the CPI: 
-Mateo: “Dogs are dangerous; a dog bit my sister.” 
-John: “Well, my neighbour’s dog is a good dog; he never bites anyone.” 

 
Lucy, as an observer, takes note of John’s counter-example. When the CPI dialogue is over, 
observers are asked to share their observations with the rest of the group: 

 -Lucy: “John gave a counter-example when he mentioned his neighbour’s dog.” 
 

By explaining the counter-example, Lucy is making it explicit to herself and to others.  
 

Meanwhile, John thinks about the way he was thinking earlier and realizes that when he was 
giving his neighbour’s dog as an example of a good dog, he was, in fact, giving a counter-example so 
Mateo could understand that not all dogs are dangerous. Here again an explicit metacognitive process 
is triggered: by listening to observations of the dialogue in which they participated, children are 
invited to think about their argumentation strategies. 
 

Then, in the next CPI dialogue, the roles are reversed: participants become observers and vice 
versa. Participants who have already been observers tend to argue differently: their observation helped 
to make them more conscious of their own intellectual moves. In the same way, participants who 
become observers consciously realize what they have been asked to do in the CPI. Both Lucy and John 
will be more aware of what a counter-example is and will even be able to name it explicitly as they 
discuss:  

 –Lucy: “I have a counter-example for what Mary just said about…” 
 

Why Should Rationality Be a Goal in Education? 

 

P4C contains a broad range of metalogical and metacognitive strategies that contribute to the 
development of logical reasoning. We have seen that doubting is an important feature of inquiry, 
helping to develop epistemic prudence and that, when coupled with many other metacognitive factors 
and the collective nature of CPI dialogue, it can foster dialectic thinking and logical reasoning. This 
makes it useful across the disciplines from science and mathematics to language acquisition. 

 
Studies in reasoning from the fields of psychology and neuroscience offer insights into possible 

avenues to improve our imperfect minds—hypotheses that may lead to a paradigm shift in our 

                                                 
44 Lipman, 2003, p. 142. 



ANALYTIC TEACHING AND PHILOSOPHICAL PRAXIS VOLUME 41, ISSUE 2 (2021) 

  
 

45 
 

relationship to knowledge and education. A society that only promotes education for technical ends 
and labour skills misses out on the pedagogical value of rationality. As Moshman writes,  
 

Democratic self-government operates best if citizens are capable of making rational political 
decisions. Promotion of rationality, then, serves a variety of personal, community, and 
governmental purposes and is thus an important goal of education.45 

 
As citizens, if we do not possess the tools for reflective thinking, we risk maintaining a subjective 

position towards our knowledge, which reduces our potential for civic contribution beyond workforce 
integration. It is therefore as crucial that children learn reasoning skills as literacy and numeracy skills. 
Otherwise, society may retain power dynamics that privilege the reasoning of a small elite, whose 
decision making might not benefit the entire community. A democratic society should be composed 
of individuals capable of autonomous political choices. To encourage rationality will not only lead to 
better social choices, but also to genuine political engagement in a democratic system. By promoting 
rationality, P4C is primed to contribute to this change. 

 
 

Bibliography 

 
Beaulac, G. et Robert, S. (2011). « Les théories de l’éducation à l’ère des sciences cognitives : le cas de 
     l’enseignement de la pensée critique et de la logique ». Les Ateliers de l’éthique. 5(2). 
Cosmides, L. (1989). « The logic of social exchange: Has natural selection shaped how humans reason?  
     Studies with the Wason selection task ». Cognition. 31, 187-295. 
Evans, J.St.B.T. et K. Frankish, dir. (2009). In Two Minds. Dual Processes and Beyond. New York : 
     Oxford University Press. 
Gagnon, M. (2005). Guide pratique pour l’animation d’une communauté de recherche 
     philosophique. Québec : PUL. 
Gigerenzer, G., Goldstein, D. G. (1996) « Reasoning the Fast and Frugal Way: Models of Bounded  
     Rationality ». Psychological Review, 103(4), 650-669. 
Hawkins, J. & Pea, R.D., Glick, J. and Scribner, S. (1984) « "Merds That Laugh Don't Like  
     Mushrooms": Evidence for Deductive Reasoning by Preschoolers ». Developmental Psychology.  
     20(4), 584-594. 
Houdé, O. et S. Moutier. (1996) «Deductive Reasoning and Experimental Inhibition Training: The  
     Case of the Matching Bias». Current Psychology of Cognition. 15. 409-434. 
Lipman, M. (2003). Thinking in education, NewYork: Cambridge University Press. 
Lipman, M., Sharp, A. M., Oscanyan, F., S. (1982), Guide d’accompagnement de la découverte de  
     Harry, trad: Marie-Marthe Ménard-Markiza AQPE, ULaval. 
Lipman, M. (1994). La découverte de Harry, trad. de Michel Haguette, CECM. 
Markovits, H., Schleifer, M., & Fortier, L. (1989). « Development of Elementary Deductive Reasoning 
     in Young Children ». Developmental Psychology, 25(5), 787-793. 
Moshman, D. (2004). « From inference to reasoning: the construction of rationality ». Thinking and  
     Reasoning, 10, 221-239. 

                                                 
45 Moshman, 1990, p. 343. 



ANALYTIC TEACHING AND PHILOSOPHICAL PRAXIS VOLUME 41, ISSUE 2 (2021) 

  
 

46 
 

Moshman, D. (1998-a). « Cognitive development beyond childhood ». Handbook of child psychology:  
     Cognition, perception, and language, (5e éd). 2, 947-978. 
Moshman, D., & Geil, M. (1998-b). « Collaborative reasoning: Evidence for collective rationality ».  
     Thinking & Reasoning. 4, 231–248. 
Moshman, D. (1994). « Reason, reasons, and reasoning: A constructivist account of human  
      rationality ». Theory & Psychology, 4, 245–260. 
Moshman, D. (1990). Rationality as a goal in education in « Educational Psychology Review ». 2(4),  
      335-364. 
Oyler, J. (2015). Expert teacher contributions to argumentation quality during dialogue. Dissertation,  
      Monclair State University, NJ. 
Piaget, J. (1974) Contradiction et conservations des quantités dans « Recherches sur la contradiction »,  
     Paris: PUF, 51-65. 
Robert, S., Roussin D., Ratte, M. et T. Guèye. (2009) Évaluation des effets du programme de  
     « Prévention de la violence et philosophie pour enfants » sur le développement du raisonnement  
     moral, rapport présenté à La Traversée. Montréal. 
Sasseville, M. & Gagnon M. (2012). Penser ensemble à l’école, Des outils pour l’observation d’une 
     communauté de recherche philosophique en action. Québec : PUL. 
Sasseville, M. (2009) La pratique de la philosophie avec les enfants. 3e éd. Québec : PUL. 
Schraw, G. & Moshman, D. (1995). « Metacognitive theories ». Educational Psychology Review. 7(4),  
     351-371. 
Stanovich, Keith E., West, Richard F. (2008) « On the Relative Independence of Thinking Biases and  
     Cognitive Ability ». Journal of Personality and Social Psychology, 94, (4), 672–695. 
Wason, P. C. (1969) « Regression in reasoning? ». British Journal of Psychology, 60(4) 471-480. 
 
 
Address Correspondences to:  
Anne Brel Cloutier 
Doctoral Candidate, Department of Philosophy, Université du Québec à Montréal,  
Email: annebrelcloutier@gmail.com  
 

mailto:annebrelcloutier@gmail.com

