Metacognition

Metacognition as a concept suffers from the same lack of categorical precision as critical thinking, so it may be a risky strategy to promote it from its role on the periphery to the center of attention as a critical thinking attribute. Let me therefore provide a brief overview of how metacognition may be understood, along with how it is said to be beneficial as an educational strategy.

Understanding metacognition is not just an issue of constitution (i.e., what makes it up); it is an issue of framing. Are we to understand it as a feeling, as a cognitive skill set, as a clutch of strategies, or as intellectual self-governance? It is also debatable as to whether some concepts, such as self-regulation, are subordinate to, or inclusive of, metacognition (Veenman, Van Hout-Wolters, and Afflerbach 2006). These are questions generally bound up in cognitive science, and the hierarchy or relationship linking these ideas is far from clear.

Despite this confusion, metacognition makes regular and frequent appearances in the literature on critical thinking (see, for example, Chaffee, McMahon, and Stout 2004; Elder and Paul 2001; Mulnix 2010; Petress 2004; Scriven and Paul; van Gelder 2005 for a range of conceptualizations and instantiations). This may be because, while there is no suggestion that metacognition is synonymous with intelligence, it does seem to be the case that metacognition, or the skills of self-awareness and self-regulation one might associate with it, improves learning outcomes. Veenman (2006, 6), for example, claims that "on the average intellectual ability uniquely accounts for 10 percent of variance in learning, metacognitive skills uniquely account for 17 percent of variance in learning."

That metacognitive skills or strategies may contribute to academic success is not difficult to imagine if we accept that "a person who is metacognitive knows how to learn because he/she is aware of what he/she knows and what he/she must do in order to gain new knowledge" (Wilson and Bai 2010, 270). How this might work would presumably be a function of discipline or situational context, and involve a grasp of metacognitive strategies and skills, including knowing how and when to apply these.

What does seem consistent in discussions regarding metacognition is that the "meta" part of the word means that we create representations of our thinking, specifically our "first-order mental states" (Fletcher and Carruthers 2012, 12). These states are those desires and beliefs, including beliefs about the truth of factual knowledge, that we use to move through an inferential process to reach conclusions or direct actions. It is also the case, however, that we make higher-order representations, which may be about how these states or representations interact, including the dynamics of the learning and reasoning process. For example, when we perceive a chair we create a mental representation of it. We might also have a mental representation of the fact that our legs are sore, perhaps from a long day standing. Working with these two representations we could create a further representation of the chair being used to relieve the pressure on our legs, and so on until we take appropriate action. What's more, we may work with several representations of chairs and call into use an existing representation of an evaluation process to determine which chair would best suit our immediate needs. More formally, we might construct a mental representation of a valid deductive argument.

It is not the case, however, that higher-order representations are, simply by virtue of being higher order, consciously attended. We may drive a car using a remarkably complex suite of higher-order representations, including some very impressive future matching and evaluation processes, while chugging comfortably along in cognitive neutral, giving these representations little or no conscious attention. The use of the term "metacognition" for simply having both first-order and higher-order representations, while perhaps cognitively descriptive, does not seem able to account for the rich educational concept of consciously modifying or accommodating our thinking toward a specified end with a view to optimizing how we get there, let alone evaluating the process as one possible path among many. I suggest that we need to both have and be consciously aware of these representations to be metacognitive. We in education might find a definition of metacognition that goes along the following lines more informative and productive: metacognition is attending to mental representations such that the representations themselves, and their interactions, become objects of study.

I do not claim that this is ontologically the case, nor that a broader definition is not more descriptive of findings in cognitive science, but I will attempt to show that using this definition (or perhaps focusing on this aspect) has pedagogical implications that deliver insights into creating better critical thinkers.

This sounds, and is, a complex way of thinking about metacognition, but we can achieve much the same end using an idea grounded more in philosophy than in cognitive science. This is not to ignore the science of cognition, but rather to frame the concept of metacognition at a different functional level. While the cognitive science underpinning our understanding of metacognition continues to develop, how it is eventually understood may remain coherent with this higher function. In the same way as our idea of a car allows us to plan for transport needs even as the specific nature of cars changes significantly over time, progress may be made in the use of metacognition as an educational concept grounded in a functional understanding while our scientific understanding of its nature continues to develop.

Dennett's (1983) development and use of the intentional stance provides a useful conceptualization of the issue that focuses on the functional aspects of metacognition rather than the underlying cognitive processes studied by cognitive scientists. The intentional stance is one that is adopted to explain the behavior of complex systems (biological or otherwise) through the attribution to them of states of desires, needs, goals and ambitions—that is, to consider them as agents (Dennett 1988, 496).

A point to consider with regard to the intentional stance is that these states, such as we might ascribe them, are a consequence of an agent's place and purpose in the world. It also assumes the agent holds, and is subject to, a rationality that prescribes for it courses of action to take and courses of action to avoid. This is to say that the dynamic interactions of representations unfold in a way that is goal driven and geared toward a specified end. We automatically take the intentional stance toward others because we are naturally vigilant as to the manner in which other agents may benefit, inform, or deceive us, and the intentional stance is useful in predicting their behavior. It is very much to our advantage to understand what they may do in the future, based on their intentions.

Metacognition can be achieved by the deliberate and explicit adoption of the intentional stance toward oneself—an internal rather than external application. In considering our own drives, beliefs, desires, thoughts, and processes, we become the object rather than the subject, just as considering another agent would make them the object, for the purposes of anticipating or planning possible future events. Our own mental representations, or some of them, become explicit, thereby making them objects of study to better understand and direct the intentional systems that we individually are. There is an asymmetry

Figure 24.1 Mental representations and their interactions—a tool for examining thinking.

Note: There is no defined starting point.

between the external and internal intentional stance, as we ascribe states to others but experience our own; however, as we consider ourselves as agents, we can use our inwardly directed intentional stance to interrogate ourselves as to what these states might be. We adopt the intentional stance when we ask questions such as "what are the agent's beliefs in this matter?" or "how do these beliefs influence the agent's thinking?" and "what are the agent's goals in this situation and how can they best be achieved given the resources and opportunities available?" To replace "the agent" in these questions with "my," and to then engage with the question, is to become metacognitive.

In contrast to the asymmetry mentioned above, another result of this dual external and internal adoption of the intentional stance is the emergence of symmetry between how we think critically in processing and evaluating the beliefs and arguments of others and those of ourselves. It means the set of skills and abilities developed for external parsing can be applied with equal efficiency internally.

To gain a better understanding of this concept of representing mental states, figure 24.1 shows a visualization of thinking processes that I have trialed in critical thinking courses to promote metacognition and assist students to critically examining their thinking. I call this an MRI diagram (Mental Representations and their Interactions) in the hope, I suspect forlorn, that the analogy with the medical MRI (Magnetic Resonance Imaging) of making things clearer will stay with students.

This tool allows students to visualize their thinking. It captures in a diagram the nonlinear, temporally recursive processes that make up thinking about an issue. It is not suggested that this particular MRI diagram is definitive, or that it is comprehensive, it is simply presented as a type of diagram that assists critical analysis of thinking. Any number of similar diagrams could be constructed to accommodate different educational contexts.

The specifics of this diagram go beyond what might be considered necessary for critical thinking (the inclusion, for example, of "creativity" as a consideration), but it does show that such a representation is consistent with the intentional stance in that it incorporates states such as beliefs, goals, and desires, as well as how these may contribute to the reasoning process. It helps students to render explicit that which might otherwise have remained implicit in their thinking.

Visualizations of thinking processes are common, and they have significant educational benefits. Argument mapping, in which evidentiary links, premises, reasons, and conclusions are explicitly stated and diagramed to support or explain reasoning, have been shown to be efficacious in developing critical thinking skills (van Gelder 2005, 45). While argument maps focus on cognitive reasoning skills, figure 24.1 broadens the approach to include aspects that affect arguments and that are implied in the intentional stance.