Q & A for the 2021 version of the map of the complexity sciences

As each new version of the map of the complexity sciences is released, there are questions regularly asked of us. Social media is not the best place for having such discussions -- particularly when people get mean or aggressive or do not actually take the time to read the Map Legend or explore things before commenting. Also, even when there are great questions asked, others may miss our response. We therefore thought it usefult to try and answer the questions typically posed to us on social media or email.

Brian Castellani and Lasse Gerrits




What is this map? 


This map is an introduction to the complexity sciences – from physics and biology to sociology and psychology to computational modelling and policy evaluation. We purposely use the term ‘sciences’ in the plural because there is no one complexity science and no one boundary around it. The map was created as an educational tool. It is to be treated as an introduction, not an in-depth investigation into the field. Experts in the field will also find the map useful for exploring new areas and for teaching.


What about the Arts and Humanities?


Unfortunately, we cannot address the Humanities or Arts as the map would become unwieldy. Complex systems thinking, fractals, chaos theory and other areas of investigation have been used in the arts; and the Humanities have added key insights, for example, Buddhist meditation, deep ecology, fractal architecture, urban design, and assemblage art.


Is the map historical? 


It is roughly historical. The five lineages, running from left to right, are based on Fritjof Capra’s  The web of life: A new synthesis of mind and matter (1996), which organises the field into: (1) dynamical systems theory and complexity in mathematics (purple), (2) systems thinking/systems science (blue), (3) the core concepts of complexity (yellow), (4) cybernetics (grey), and (5) artificial intelligence/methods (orange). While not perfect, we’ve kept this basic framework as it provides a nice skeleton on which to assemble the map.


How should the map be read? 


It should be read left to right, moving from the early 1900s to the present. Topics are placed approximately at the point when they became a major area of study. For each topic, we have provided a handful of top scholars, including, when possible, the individual or team that was instrumental in advancing the topic.


How was the map compiled? 


Between the two of us, the map represents over forty years of combined research and reading, as well as in-depth discussions with colleagues across the various fields and around the world. Castellani launched the first version of the map in 2009. Since then, it has been revised every several years, as the field has massively expanded over the last decade. The current version, which is an update on the 2018 map, is rooted in our fellowship at the Institute for Advanced Study of the University ofAmsterdam.


Why didn’t you use a bibliometric analysis to make the map? 


Bibliometric analyses are all the rage, and they can be powerful tools. There are a couple of reasons why we don’t use those tools. First, bibliographic analyses struggle to construct a history of a topic. They are better at providing cross-sectional snapshots. For a good overview, see Thomas, J., & Zaytseva, A. (2016). Mapping complexity/Human knowledge as a complex adaptive system. Complexity, 21(S2), 207-234. Second, those tools depend heavily on articles (as opposed to books), recognised journals (which is an issue, as many complexity works are often in obscure journals or blogs, etc) and sources that are online as opposed to libraries, archives, conference letters, and so forth. They also do not capture historical impact beyond citations. Third, there is range of more minor technical problems that make us unsure if those tools can do a better job than we did manually.


Is this map complete? 


The map is not complete, and was not designed to be – in fact, we are not sure what would even entail. The complexity sciences represent a loose (but connected) and quickly evolving body of knowledge that intersects with almost every field in the sciences and the social sciences. A map created in 2019 will not look like a map in 2020. We focus on providing a reasonably comprehensive introduction to the field. The nice thing about the online version is that, by clicking on the links, users are taken into even more in-depth reviews that link to an even wider range of information.


How about diversity in terms of age, gender, ethnicity, and nationality?


We made a concerted effort to make sure the map highlighted the work of a wide variety of scholars around the world and up-and-coming researchers. We also sought gender and stage of career balance as well as ethnicity and nationality. We will continue to advance the work of everyone we can.


Why is author x not on the map? 


We often receive questions about why a certain author is not on the map. Sometimes the scholar is missing because of the limitations in our knowledge. Most of the time it is because the map is an educational tool, and of limited size and space, and can only include so many people.


Could you please include author x on the map? 


Unfortunately, we are unable to fulfil such requests.


Could you please include me on the map? 


Unfortunately, we are unable to fulfil such requests.


I know how to improve the map.


We invite everyone who believes that our map needs to be improved to make an alternative one themselves. We are looking forward to such initiatives.


This is not a good map / I don’t like your map / Your understanding of the field is incorrect.


As stated above, we are looking forward to alternatives. In the meantime, in the brave new world of social media, it is easy to be cruel. Please do not be mean.


Can I use your map? 


You can always use the map if you attribute correctly with this reference:

Castellani, B and Gerrits, L (2021). Map of the complexity sciences. Art and Science Factory, LLC.

While we encourage sharing of the map, we want to point out that the map may not be used for commercial purposes and / or without proper attribution.


CLICK HERE for downloads of the map as a jpg and pdf, which are located on the Map Legend page.



Under what license is the map released? 


The map is licensed on Wikipedia under the Creative Commons Attribution-Share Alike 4.0 International license.


Will you continue to update the map?


Yes, we plan to update the map regularly, as was done since the original version from 2009.


How do you know what is complexity and what is not? 


As others have said before us, and we agree with, the complexity sciences are not defined by clear boundaries. It is a sprawling and growing group of theories, methods, findings, and big and small ideas, that permeates in almost every field imaginable. The boundaries between what is about complexity and what is not are amorphous.


The map is skewered / biased towards…


Bias is real. For starters, we can only cover English / German / Dutch / French between us, as such we may overlook work published in other languages. There is also the fact that map-making is somewhat path-dependent, where those who are said to have made an impact may remain to be designated as such. Above all, it can be hard to trace the origin of ideas, especially going further back in time. It may very well be that someone took an idea from someone else without proper attribution. Sadly, some labs or research groups tended to build on the work of PhD’s and postdocs without giving much credit to their work. This continues to be an issue until this very day. We’ve tried our best to present a balanced overview of the people who have driven the study of complexity, and those who continue to do so.


I don’t think that all the names on the map are the big names…


This is correct. We purposely include names of upcoming scholars that we find worth following as they push the field into interesting directions. Map-making is not only about charting the terrain just crossed but also an attempt to charter the unknown terrains. That is why we include scholars that we believe have something novel to say.





The Limits of Evolution and the Modularity of Mind -- Can we please finally move on?

In preparation for writing the 'psychology and complexity' for my forthcoming book with Lasse Gerrits, The Atlas of Social Complexity, I have once again delved into the deep end of the pool on modularity of mind and its links to human evolution.

Modularity Ain't Rev-Evolutionary

Colombo, in an interesting article, Moving forward (and beyond) the modularity debate: A network perspective (2013), nicely sums the purported distinction between modularity and evolutionary psychology:

"At least since Fodor’s 1983 (The Modularity of Mind), the notion of modularity has been one of the most important concepts used to articulate an account of the human cognitive architecture, which provides us with an encompassing theory (a “blueprint”) of the nature, arrangement, and form of the structures and processes that are responsible for cognition and adaptive behavior. This should be distinguished from a theory of the origins of cognitive architectures, which is concerned with the evolutionary and developmental history of the structures and processes that are responsible for cognition and adaptive behavior."

An easy example that makes these differences clear is Chomsky's famous language acquisition device and his agnostic-to-antagonistic views of evolutionary psychology and sociobiology. Massive pages of ink devoted to the complexities of language acquisition aside, Chomsky's basic point is that our brains seem to come ready made, in certain ways, to engage in language -- call it a set of modules if you will. As to how much evolution drives this ability, the answer is probably not as much as we think. We only need so much evolution to survive. While the human brain is a product of evolution, the resulting mind, in all of its massive complexity, is so much more than what we need to evolve and survive. Trying to retrospectively link each little thing we do, from painting a picture to nibbling our nails, to some evolutionary necessity in our past is pointless -- and now it seems, given the current state of the literature, mostly untenable.    

Finding a middle ground

For me, modularity as a strong programme has always been underdetermined by the evidence, with its widespread acceptance appearing to be driven more by academic celebrity and its theoretical simplicity. Modularity as a weak theory has always held my interest, as complex systems are generally comprised of subsystems and lower orders of complexity, working top-down and bottom-up simultaneously, with command centres of limited ability and so forth. The embodied mind being comprised of modules -- that is, some complex network of functional subsystems (from our brains to our gut microbiota), with fuzzy organisational closure, for the purposes of completing a task outside conscious awareness -- just seems to make reasonable sense. Language acquisition, speech, motor function, paleomammalian emotions, microbiota-gut-brain axis, there is an endless list of things needing constant care and attention that our conscious, command-centre selves cannot effectively manage while writing sentences, eating toast and drinking our coffee in the morning.

Evolutionary psychology and sociobiology, however, and their just-so, reverse engineered usge of modularity -- in particular the massive modularity thesis -- have always seemed baffling. That is not to say that they don't have useful things to say, as they do. They just as often say things that are not useful. It is therefore exciting to see, fingers crossed, that the over-usage of modularity to support evolutionary psychology, as well as modularity in its strong form, are finally going the way of the dinosaurs. Or, at least, that is the sense I got from several recent articles. Here is a quick list:

Pietraszewski, D., & Wertz, A. (2021). Why evolutionary psychology should abandon modularity. Perspectives on Psychological Science, 2.

Palecek, M. (2017). Modularity of mind: Is it time to abandon this ship?. Philosophy of the Social Sciences, 47(2), 132-144.

Colombo, M. (2013). Moving forward (and beyond) the modularity debate: A network perspective. Philosophy of Science, 80(3), 356-377.

Bertolero, M. A., & Bassett, D. S. (2020). On the nature of explanations offered by network science: A perspective from and for practicing neuroscientists. Topics in Cognitive Science, 12(4), 1272-1293.

Sosis, R., & Kiper, J. (2014). Why religion is better conceived as a complex system than a norm-enforcing institution. Behavioral and Brain Sciences, 37(3), 275.

The basic story I get from these articles is that, despite the propaganda, modularity as a basis for our evolutionary psychological development has failed to produce the necessary evidence for its assertions and has made little advance in the past decade or so. In short, it is mostly wrong. Colombo sums it up rather nicely:
"Because of mere terminological disputes, because of vagueness surrounding putative central features of modularity such as functional specialization, domain speci´Čücity, and informational encapsulation, and especially because of little agreement about the proper empirical methods for discovering and justifying the existence of candidate modules (cf. the controversy around the cheater-detection module: e.g., Fodor 2000, 2008; Sperber and Girotto 2003; Cosmides and Tooby 2008a, 2008b), the modularity debate in the cognitive sciences and philosophy of psychology has often been frustratingly fruitless." (See Moving forward (and beyond) the modularity debate: A network perspective)
In turn, as the theoretical foundation for evolutionary psychology, this has come at a cost. In Why evolutionary psychology should abandon modularity (2021), Pietraszewski and Wertz make this point clear:  
"The upshot of all of this back-and-forth is that both sides in this modularity debate feel as if the other is patently absurd in its convictions. Evolutionary psychologists cannot imagine what else could exist but functional specialization in the mind. Meanwhile, critics on the other side feel as if the bottom has been pulled out from their understanding of evolutionary psychology if it does not intend the attributes of modularity that it now seems to be backing away from. Both sides are left, understandably, exasperated and at a seeming impasse.
The cost of this state of affairs cannot be over-stated. It has misled an entire generation of scientists about how to think about relationship between evolution and the mind, and it actively hinders progress in understanding how the mind works."
An example of this problem is a recent publication in the Proceedings of the National Academy of Science. The article, "The pandemic exposes human nature: 10 evolutionary insights" is of value not only because its authors include some of the key names in the field of evolutionary psychology, but also because it represents the problematic over-valuing of these ideas into the social realm, where simpler and more empirically viable explanations from fields such as public health, social pscyhology, sociology, political science and cultural anthropology already exist. This is not to say that the article does not make some excellent points, such as how quarantine will negatively impact the gut microbiota of kids, which will impact their long-term health. The problem is its reductionist centering of evolutionary theory and its modular understanding of social life as the basis for making sense of human life. 
The first line of the article says it all:
"Nothing in biology makes sense except in the light of evolution, and nothing about the human response to COVID-19 will either (p. 27768). Nothing!

Evolving forward -- Network and more networks!

Image by Bertolero, M. A., & Bassett, D. S. (2020).
I cannot remember where, exactly, Richard Rorty stated it, but his points was this: philosophical arguments generally don't resolve with a winner, instead they get disgarded once they are no longer useful. "Modularity or not" seems to be losing steam as an argument -- and hopefully the antagonistic clash between opposing views will eventually lose sway over evolutionary psychology too. 

What is replacing the "modularity or not" debate? That is not entirely clear, but one productive avenue is complex network modularity and, more generally, a complex systems modularity -- both of which have been around for a while and seem to be gaining ground. I cannot survey this research presently, except to offer a few articles to explore:

Favela, L. H., Amon, M. J., & van Rooij, M. M. (2018). The incommensurability of emergence and modularity in complex systems: A comment on Wastell (2014). Theory & Psychology, 28(4), 559-567.

Wastell, C. A., Purcell, Z., Howarth, S., Paterson, W., & Slocombe, B. (2018). The development of Complex Emergent Modularity: A reply to Favela, Amon, & van Rooij (2018). Theory & Psychology, 28(4), 568-571.
Colombo, M. (2013). Moving forward (and beyond) the modularity debate: A network perspective. Philosophy of Science, 80(3), 356-377.

Zerilli, J. (2019). Neural reuse and the modularity of mind: where to next for modularity?. Biological Theory, 14(1), 1-20. 

Newman, M. E. (2006). Modularity and community structure in networks. Proceedings of the national academy of sciences, 103(23), 8577-8582.

The other avenue would be to finally recognise that cognitive science is complexity science and to search for ways to integrate the various ideas and empirical evidence using a complex systems framework. Examples of this view (of which network modality is a part) include embodied cognition, connectionism and ecological psychology, as well as as recasting cognition as complexity and challenging theoretical and methodlogical reductionism. Embodied mind resarch is very well known. The latter is a more general argument. See for example:

Favela, L. H. (2020). Cognitive science as complexity science. Wiley Interdisciplinary Reviews: Cognitive Science, 11(4), e1525. 

Van Orden, G., & Stephen, D. G. (2012). Is cognitive science usefully cast as complexity science?. Topics in cognitive science, 4(1), 3-6.

The utility of these alternative framings come with their own intellectual traps. All scientific ideas do. They are at least more interesting then telling me:

"A key insight of evolutionary thinking is that—in contrast to the metaphor of the invisible hand—the pursuit of lower-level interests, such as short-term individual, corporate, partisan, or nationalistic interests, is far more likely to undermine than contribute to the global common good. . . . [As such,] treat differences in nation/state responses to COVID-19 as natural experiments in evolutionary processes by documenting different phylogenies of responses, measuring the efficacy of each, and then replicating successful approaches in necessary areas and future pandemics" (Seitz et al, p. 27772). 

If this is a major observation from a field some fifty years in the making, I think it is time to move on.