Metacognition as a prerequisite for interdisciplinary integration

By Machiel Keestra

Machiel Keestra (biography)

What’s needed to enable the integration of concepts, theories, methods, and results across disciplines? Why is communication among experts important, but not sufficient? Interdisciplinary experts must also meta-cognize: both individually and as a team they must monitor, evaluate and regulate their cognitive processes and mental representations. Without this, expertise will function suboptimally both for individuals and teams. Metacognition is not an easy task, though, and deserves more attention in both training and collaboration processes than it usually gets. Why is metacognition so challenging and how can it be facilitated?

Understanding cognitive processes and representations

Whenever we engage with any cognitive or behavioral tasks, our brain employs a mental representation or knowledge structure that corresponds to a word, image, or other information pertaining to that task. Experience contributes to further enrichment and structuring of that representation. A beginner’s mental representation of a new word (‘space shuttle’, eg.) may thus contain just its letters and an image of the object, yet additional experience and information is automatically cognitively integrated with that representation, providing associations with mental representations of shuttle parts, of launching and landing actions, images and so on.

The superior performance of experts relies upon their having assembled many more, and more complex, mental representations pertaining to their field of expertise. Examples range from accurate and fast recognition, recall and processing by chess masters of a large number of complex chess positions, the complex sequences of movements of musicians playing by sight, fast strategic decisions by sport champions, and immediate detection of a mistake in a function by mathematicians.

Importantly, we do not need to learn those mental representations explicitly, nor are we often able to make them explicit once we have learned them. Indeed, we usually acquire and employ them automatically and implicitly in our cognition and behavior: for example, young children don’t explicitly learn grammatical rules yet can use them well.

Problems with expertise

Handy as such automatic and implicit handling may be, it also contributes to ‘brittleness’ and other flaws of individual expertise. In particular, experts:

  • are overconfident in exceptional situations
  • often demonstrate a bias or fixedness towards habitual responses
  • tend to rely upon their expertise in neighbouring yet different domains
  • often display a lack of creativity compared to beginners.

These are direct consequences of the cognitive processes upon which expertise rests and the knowledge structures or mental representations that are involved in those processes.

Particularly relevant for interdisciplinarity is the disappointing fact that expertise can make it more difficult for us to recognize how the insights from an expert in another domain can be added to our knowledge or performance. Making our cognitive processes and representations explicit by metacognizing is a requisite for recognizing implicit assumptions and for acknowledging gaps in knowledge and methods that another expert might help to fill, as shown in the first figure below.

An expert engages in meta-cognition about their thinking and knowing. Here, the expert reflects specifically about their learning process (in the cloud) and the set of representations it has yielded. Such reflections also prepare them for the integration of an additional insight from another expert, such as the green square added here (from Keestra 2017: 142).

Team collaboration and metacognition

Adding another level of complexity to this situation in the case of team work is the fact that experts automatically develop mental representations related to their team work, in addition and connected to those pertaining to their individual expertise. These representations concern the ‘who, what, why, when, and how’ of the team, containing information about the team itself, its task, process-related information and representation of its overarching goal.

Team mental representations bring along similar risks to those mentioned above. For example, overlapping representational contents that are shared by all team members are quickly recognized and tend to dominate joint cognition and actions. Analogous to an individual expert’s bias, a team risks slipping into groupthink when it does not engage in team metacognition.

For team members to recognize and make effective use of each other’s non-overlapping representations and skills, in contrast, requires extra time and effort devoted to team metacognition in addition to individual metacognition, as represented in the figure below.

An interdisciplinary team of experts together develops a more comprehensive understanding of a phenomenon – represented by the three-dimensional cube composed of different elements each of them contributes. Their joint or team meta-cognition upon their interdisciplinary collaboration facilitates the process of their development of an interdisciplinary integration of their distinct mental representations of the phenomenon (from Keestra 2017: 156).

Working in a team also has metacognitive benefits. Being confronted with the metacognitive self-reflections and the self-regulatory strategies of others and receiving feedback from them can help in recognizing and articulating individual metacognition. In addition, individuals can feel more motivated to metacognize when functioning in teams.

However, to have such results team metacognition must be adequately guided in order to avoid unnecessary confusion, which can be exacerbated when status and cultural differences are insufficiently addressed. Adequately planning research phase-specific rounds of metacognition is important. Team leaders should formulate relevant prompts or questions to elicit the required individual and team metacognitive reflections and discussions.

Useful prompts

A few prompts are listed here; more can be found in an appendix to Keestra (2017):

  • What added value could research from the humanities/social sciences/sciences have for your research?
  • Why do you think that the task you propose to perform is optimal for solving the problem at stake? Would an alternative route be possible? Do you have doubts about the routes proposed by others?
  • Of all the features of the problem under scrutiny (as perhaps represented in a figure), what does and what does not make sense from your disciplinary perspective? Or what is especially difficult to understand? What would you like to know more about?
  • What goals have you as a team determined and what is the plan for reaching those goals? What are the issues that might arise with the current plan? Do your answers to these questions vary among individual team members?
  • Did new insights emerge or a new situation present itself to the team, that makes you feel you should revisit some of your previous personal contributions to the team work?


What has your experience been with metacognition challenges in interdisciplinary work? Have you ever seen or experienced groupthink? Are there other prompts that you have found useful? How have you incorporated metacognition in an interdisciplinary team project?

To find out more:
Keestra, M. (2017). Meta-cognition and reflection by interdisciplinary experts: Insights from cognitive science and philosophy. (With an appendix of prompts or questions for metacognition). Issues in Interdisciplinary Studies, 35, 121-169. Online:

Further reading:
Wiltshire, T. J., Rosch, K., Fiorella, L. and Fiore, S. M. (2014). Training for collaborative problem solving: Improving team process and performance through metacognitive prompting. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 58, 1: 1154-1158. Online (DOI):

Biography: Machiel Keestra PhD is a tenured assistant professor of philosophy at the Institute for Interdisciplinary Studies at the University of Amsterdam, the Netherlands. He teaches philosophy of science and interdisciplinary research in the Natural and Social Sciences bachelor and in the Brain and Cognitive Science master programs. He is a researcher at the Institute for Logic, Language and Computation, focusing on the philosophy of cognitive neuroscience. He is past-president of the international Association for Interdisciplinary Studies (AIS) and co-chairs the upcoming AIS conference on ‘Interdisciplinarity in Global Contexts’, October 24-26, 2019, in Amsterdam.

15 thoughts on “Metacognition as a prerequisite for interdisciplinary integration”

  1. Hi Femke, glad to see you appreciated this post. I typically ask graduate/undergraduate students to respond individually (via a google form, for example) to a list of such metacognitive prompts – tailored to their context and project – and then project the results for a plenary discussion of them. Sometimes I also let research teams (generally 4 persons) fill those in individually and then have them identify potential challenges for their team work. Depending upon the level, I combine it with a seminar on philosophy of (interdisciplinary) science, particularly on implicit and auxiliary assumptions. Thanks also for the link to the metaphor-focused clean language technique, which I wasn’t familiar with. What is the aim for you when using that technique and discovering relevant metaphors with it?

  2. Dear Machiel,

    I read the blog about “metacognition” with much interest. In her book entitled Cultural Evolution (CUP, 2011), Kate Distin uses the word “metarepresentation” which she defined as follows (at p. 86):

    A metarepresentation is a representation of another representation. Its content is that other representation, and crucially this includes information about both form and content. The ability to metarepresent is the ability to recognise the distinction between the two: to reflect on the connection between a representation and the informa¬tion that it represents. The information that evolves, when we metarepresent, is information about how we represent. To put this another way, once we start comparing the representational features of different lan¬guages, the two systems effectively begin to compete with each other, under representational pressure.

    As the briefest glance at modern culture makes clear, our cognitive escape route from the restrictions of our native language has not been restricted to other natural languages. Limited as it is by the length of the critical period and by the human capacity for learning, natural language has become, over time, inadequate to the representational task that it was originally set. If language is to account for cultural evolution then we need to look beyond natural language to the artefactual languages that have evolved in its wake.

    An advantage of metarepresentations above metacognition may be that representations can be expected to circulate, as can discursive knowledge. Metarepresentations can be compared with Luhmann’s (e.g., 1986) codes of the communication which span “horizons of meaning” (Husserl, 1929).

    When metarepresentations operate upon one another, redundancy may be generated; that is, new options are added to the communication because of synergy in the circulation at the supra-individual level (Leydesdorff et al., 2017). Growth of the number of options is crucial for a viable innovation system (Petersen et al., 2016).

    Loet Leydesdorff
    University of Amsterdam
    Amsterdam School of Communication Research (ASCoR) ;


    Distin, K. (2010). Cultural Evolution. New York, NY: Cambridge University Press.

    Husserl, E. ([1929] 1960). Cartesianische Meditationen und Pariser Vorträge [Cartesian Meditations and the Paris Lectures, translated by Dorion Cairns] (Vol. at ). The Hague: Martinus Nijhoff, 1960.

    Leydesdorff, L., Petersen, A. M., & Ivanova, I. (2017). Self-organization of meaning and the reflexive communication of information. Social Science Information, 56(1), 4-27. doi: 10.1177/0539018416675074

    Luhmann, N. (1986). Love as Passion: The Codification of Intimacy. Stanford, CA: Stanford University Press.

    Petersen, A., Rotolo, D., & Leydesdorff, L. (2016). A Triple Helix Model of Medical Innovations: Supply, Demand, and Technological Capabilities in Terms of Medical Subject Headings. Research Policy, 45(3), 666-681. doi: 10.1016/j.respol.2015.12.004

    • You’re absolutely right, Loet, that metarepresentation should be part of the metacognitive process – thanks for drawing attention to it. In the article underlying this blog I’ve mentioned how crucial such second-order reflection is for a team, e.g. for articulating doubts or convictions about its own knowledge and about the knowledge acquired from others. I think it is recommendable that the IPCC reports now include such ‘2nd order science’ into its finding, articulating the uncertainties inherent in its models, for example (cf. my p. 148, with reference to Sperber’s ‘epidemiology of representations, I’ll check out your interesting reference to Distin)
      As for metarepresentations operating upon each other: I’m not sure I understand how that would work – would that imply a 3rd order science, leading to a novel shared framework to guide further metacognitive and metarepresentational reflections? In any case, I certainly recognize how metarepresentational reflection should be part of metacognition, as they can help to further support metacognitive self-regulation which is the important goal of the task.

      • Dear Machiel,

        Whereas metacognitions are attributes to cogitantes (human minds), meta-representations are attributes of representations (cogitata). The dynamics of our minds are historically coupled to and constrainted by our life-cycles, whereas the dynamics of representations are second-order and potentially restructuring the historical conditions from which they emerge.

        The (re)construction by agency remains bottom-up and historical. The emerging feedback is top-down and can develop an evolutionary dynamic; counter-clockwise. The unit of analysis is no longer individual agency, but intersubjective intentionality. Differentiation prevails whereas metacognitions tend towards integration (in order to prevent cognitive dissonances).

        The bottom-up process is generative and operating in terms of knowledge claims. The top-down arrow of validation on the basis of codes is incursive on the historical conditions and can thus drive a techno-cultural restructuring.


        • Thanks for the clarification, Loet. I appreciate the dynamics you’re referring to. Yet we don’t need to restrict metacognition as only focusing on the cognizers and not their cogitations (representations). Flavell’s early definition of metacognition as being “knowledge and cognition about cognitive phenomena” (Flavell, 1979, p. 906) is also more comprehensive, taking into account that when talking about cognizers we need to include their cogitations being part of that reflection. In my article I argue that for experts it is important to be aware of the implicit cognitive process of ‘Representational Redescription’ that facilitates their growing expertise yet brings along the brittleness, biases etc. mentioned in the blog.

          This brings me to your nice account of the bottom-up and top-down processes – you perhaps agree that some form of interaction (not to mention dialectic) between those processes would be optimal: if intersubjective deliberation would include adjustment of the knowledge claims (and associated representations) and v.v. The end of my article briefly touches upon the Reflective Equilibrium process (which is different, of course) while making a more normative claim, but I do like your articulation of the historical dynamics that might be involved in such a process.

          Thanks again and best, Machiel

  3. If I understand your argument correctly, then the role of metacognitive activity you describe, Machiel, is not so much that to simplify the process of communicating across disciplines as that using explicit metacognitive techniques can help ID researchers to avoid obstacles and pitfalls that threaten to stymie progress.
    I can think of two common scenarios where this comes into play–and can have strong epistemic effects. One is when the norms or practices of our home discipline are significantly different than a collaborator’s discipline. Often, those are ethically or socially charged. For instance, someone in the humanities or natural sciences might consider doing research that would affect people, while a social science collaborator might think “Are you lazy or are you unethical? You need to get IRB approval for that!” When working with colleagues, our first thought should be “this could be a disciplinary difference” rather than going straight to the lazy/unethical dichotomy! The metacognitive rule (to consider disciplinary differences first) does not make conversations about how to handle the disciplinary difference easier, but it does help to maintain respectful conversation.
    The second scenario is similar but addresses standards of evidence and argument. We frequently use shortcuts to judge the quality of others’ work, and our own disciplinary standards of evidence and rhetoric provide the heuristic. Metacognitive rules should encourage us to trust our collaborators–but the process of negotiating the appropriate standard in an ID context could still be long and difficult.
    Thank you for this post-it’s interesting to think of rules for developing trust in ID practice as being metacognitive.

    • Fascinating scenarios, Ebrister: thanks! I appreciate how you aim to present a pragmatic ‘metacognitive rule’ that could work in those. I would suggest to expand the rule from ‘this could be a disciplinary difference’ to something like ‘your disciplinary perspective suggests implications that someone with another perspective might not see – or, alternatively, your disciplinary perspective obscures content and implications that your colleague notes’. It is not only about judging, biased by your own perspective, but really also about expertise-dependent information processing & representations that differ from field to field.
      As for trust: that is indeed absolutely crucial. In my article I’ve referred to the space shuttle Challenger disaster from 1986 when the engineers did not trust their own judgments, nor did the launch management team trust the engineers’ doubts about the ‘unknowns’, leading to a fatal launch decision. Insight in how colleagues think helps to build trust, indeed.

    • Thanks to the interesting reference, Jon. That article offers a quite specific diagnosis of how the problem of fragmentation and specialization has impacted upon our universities – referring to i.a. the pragmatist distinction between denotation and connotation that explains why often the difficulty of integration is underestimated as some overlook the fact that our utterances are not only denotative but also connotative (carrying values, etc.). My plea for metacognition sits well with the article’s diagnosis, even though I think we do need other diagnostical tools as well – like attention to the expertise-dependent cognition that I’ve mentioned above.

  4. “What has your experience been with metacognition challenges in interdisciplinary work?” I’ve noticed it goes a lot better when there is a facilitator involved, someone with experience connecting team members who doesn’t necessarily have a role in the work of the team. Good facilitators seem to know when to encourage self-reflection and when to encourage other behaviors.

    • Great point, exaptly-Jill (?)! Depending upon the composition of the team and the familiarity with these issues, facilitation can be very helpful. Although the metacognitive prompts that I’ve added to the article ( ) can be responded to individually, my argument is that discussion of these prompts in a diverse /interdisciplinary team actually help this process of metacognition. Yet facilitators can certainly help, e.g. to avoid unproductive debates about the expertise-dependent differences. The role of facilitative leadership of interdisciplinary teams is elaborated upon in another article that I included in the special section on ‘Interdisciplinary Collaboration’ to which my piece belongs, viz. the article by Lash-Marshall, Hirsch e.a.


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: