1 Introduction: the prevalence of multidimensional phenomena

Think about the following well-known scenarios:

  1. A.

    The virus SARS-CoV-2 caused the COVID-19 pandemic. The size of the virus is much smaller than a cell. Despite this, its reproduction created a world crisis, triggering economic recessions, political turmoil, sociocultural changes, and mental breakdowns around the world. Due to the COVID pandemic, new social and cultural processes appeared; many were replicated and selected, while others were discarded or weakened. In turn, several of the processes and entities that socially and culturally survived experienced changes. Let us think, for instance, about the evolution of telecommuting or the use of sanitary masks since 2020. These changes represent novelties that triggered new processes of replication and selection, both of the virus and of social and cultural processes.

  2. B.

    Human activities, such as extracting and burning fossil fuels (coal, oil, and natural gas) since the Industrial Revolution, have led to an increase in greenhouse gases in the atmosphere. This has caused a radiative imbalance. As of 2019, the concentrations of CO2 and methane have increased by roughly 48% and 160%, respectively, since 1750. These levels of CO2 are higher than at any point in the last 2 million years, and methane concentrations are also much higher than they were in the last 800,000 years. In other words, advancements in human industrial technology have contributed to an increase in the Earth’s temperature, leading to environmental, political, economic, and sociocultural changes around the world. As a consequence of this phenomenon, new social and cultural processes have emerged, including political, economic, technological, and dietary changes, as well as new ideologies. Many of these processes and entities were replicated and selected, while others were discarded or weakened. In turn, several of the processes and entities that survived socially and culturally experienced changes. These changes represent novelties that triggered new processes of replication and selection.

  3. C.

    In cities like Beijing and Shanghai, pollution from factories has increased the levels of sulfur dioxide (SO2), nitrogen dioxide (NO2), and other dangerous pollutants. As a result, infertility has significantly risen in recent years. This change in fertility patterns due to pollutants has had political, economic, and sociocultural consequences. New social and cultural processes have emerged as a result, such as the use of pollution masks, air quality apps, restrictions on coal factories and traffic, and a shift towards indoor activities. Over time, some of these processes and entities have been replicated and selected, while others have been discarded or weakened. Consequently, the surviving social and cultural realities have also undergone changes, leading to the emergence of new processes of replication and selection.

What similarities can we find in these examples? We have tiny molecules triggering complex economic crises and human cultural activities changing the physical and chemical properties of the Earth. How can we find theoretical order in this chaos? Is there a way to not get completely lost in this epistemological and ontological maze? Let us consider a one-dimensional approach with the first example: COVID-19 → many people infected → social and economic turmoil → harsh political decisions. However, this apparent clarity is very illusory. The impact of the coronavirus in terms of the death toll and number of patients hospitalized did not correlate with the cultural perception of the virus, social upheaval, and political and economic measures. In fact, in several cases, the opposite of what common sense would dictate occurred: countries with a lower death toll were more culturally traumatized and imposed more measures than countries with a far higher impact in terms of death and hospitalizations (Alexander and Pérez-Jara 2021). We could use similar approaches to explain pollution-based infertility or climate change. Again, the causal lines *more pollution → more infertility*, and *more greenhouse gases → more temperature* although completely true, are crossed with many other causal arrows. For instance → more social pressure → more psychological stress → more infertility; inadequate diet → hormonal problems → higher infertility; advanced age searching for pregnancy → more difficulties in conception, and so on. The same effect can be caused by many different causes. And most effects we observe are the consequence of the ontological convergence of multiple causal lines.

This means that, more often than not, there are hidden variables that codetermine all kinds of phenomena around us. Monocausal, one-dimensional explanations, as attractive as they are, seldom (if ever) work to explain any dimension of the universe. Let us focus on the human domain for an example. In the phenomenon of depression, chemical factors sometimes have more weight than others (for instance, depression triggered by toxins or pollution). On other occasions, though, biological factors have more weight (for example, a strong genetic tendency for depression, a brain tumor, lack of sleep, disrupted circadian rhythms). While in other cases, social and cultural factors can be more prevalent behind depression—for instance, through pessimistic and dark religious or ideological narratives, bullying, and personal losses. Of course, on other occasions, depression, as a multidimensional phenomenon, is triggered by the combination of different factors with similar weight. This ontological thesis of dynamic asymmetrism also has an epistemological correlate: sometimes, social psychology is more important than neurobiology in explaining the origin of someone’s depression, but not on other occasions. We find similar situations with any complex sociocultural phenomena such as homicide, suicide, school failure, patterns of consumption, marriage rates, poverty, and economic crises. They are multidimensional, multicausal phenomena. This paper addresses the complexity of multidimensionality through an expanded coevolutionary theory.

2 The coevolutionary approach in economics and beyond

All sciences have philosophical assumptions, whether implicit or explicit, and when mistaken, they can mislead scientific theories into blind alleys. Economics is no exception. Wrong philosophical assumptions pose a double risk in the case of economics, as it is not only a descriptive science but also a normative one. In other words, wrong approaches to economics can lead to either disastrous consequences or the inability to solve social problems with a strong economic dimension. In 1977, Nelson (1977) wrote an essay titled “The Moon and the Ghetto: An Essay on Public Policy Analysis.” In this essay, he posed a question that remains relevant today: “If we can put a man on the moon, why can’t we solve the problem of the ghetto?” Nelson argued that the progress of public policy was often hindered by the incomplete and sometimes incorrect conceptualization of problems and solutions by the different parties involved in decision-making.

Nelson and Winter (1985) conducted a critical analysis of the philosophical assumptions underlying neoclassical economics. They argued that neoclassical economics was too simplistic and rigid, failing to explain many economic phenomena or address socio-economic issues. The term “neoclassical economics” was coined by Thorstein Veblen (1900) in his 1900 article “Preconceptions of Economic Science.” There is no universal consensus on the definition of neoclassical economics, leading to a diverse range of neoclassical theories applied to different problem areas and domains. These theories cover topics such as neoclassical labor and neoclassical demographic change. However, despite their diversity, some common threads can be found in these neoclassical theories. E. Roy Weintraub (2007) suggested that neoclassical economics is based on three fundamental assumptions: (1) rational choice theory; (2) individuals seek to maximize utility, while firms aim to maximize profits; and lastly, (3) people act independently based on full and relevant information. However, these three approaches are considered naive because they only capture a limited dimension of human nature and social interactions, including economic ones. For example, rational choice theory usually assumes a flawed philosophical and anthropological conception of human beings, which has been widely criticized by sociology, cognitive neuroscience, psychology, and philosophy (Pérez-Jara and Camprubí 2022).

The failure of traditional economic schools of thought to explain essential dimensions of the economic sphere has led to the development of Coevolution in Economic Systems (CES) in recent decades. The evolutionary approach in economics was primarily pioneered by Richard R. Nelson and Sidney G. Winter with their influential book “An Evolutionary Theory of Economic Change” (1982). Inspired by this approach, other economists have continued to develop the evolutionary perspective over the last four decades: see for instance Gowdy (1994); Metcalfe (1998); Dosi and Coriat (1998); Foster and Metcalfe (2001); Murmann (2003); Witt (2003); Dopfer and Potts (2008); Dosi and Roventini (2019); Freeman (2019); Shiozawa et al. (2019); Hanappi (2020); Sawyer (2020); Dow (2020); and Prochaska and Schiller (2021). Nelson has also further developed his approach (Nelson 2018). Currently, authors such as Isabel Almudi and Francisco Fatas-Villafranca are also delving into the evolutionary perspective. They are conducting in-depth studies on how evolutionary processes are always coevolutionary because they do not operate in isolation; rather, they (co)determine each other (Almudi and Fatas-Villafranca 2021, 2022).

Economic events are multidimensional, multicausal phenomena where causal arrows and the dynamic weight of factors involved are far more complex than the neoclassical economics approach. Different evolving systems shape each other and are dynamically codetermined following complex patterns. In order to capture this intricacy, Nelson and Winter applied the concept of evolution, originating from biology, to the analysis of the economic sphere. Authors who have emphasized the significance of (co)evolution in explaining the dynamics of the economic sphere have approached the coevolution phenomenon from various evolutionary perspectives, such as Hayekian, post-keynesian, Veblenian, institutionalism, neo-Schumpeterian, historicist school, evolutionary-naturalistic approach, and complexity economics. As Almudi and Fatas-Villafranca have emphasized (2021), “the diversity of these approaches underscores the importance of understanding how coevolution processes are linked to dynamic structural changes observed at various levels in social systems, including natural, societal, and cultural levels.”

In this paper, I will expand on the general ontology of coevolution beyond the economic sphere. I propose a non-flat process ontology that pays special attention to the interplay of continuities and discontinuities, symmetries and asymmetries, horizontal and vertical causation, and emergence and disappearance in the systems and subsystems that compose the universe. I will demonstrate that these notions are deeply interconnected into a coherent ontological framework that can help deepen the ontological assumptions of coevolutionary economics.

I will distinguish general and specific (co)evolution. General coevolution implies dynamic codetermination between processes that result in the emergence of new realities. That is, general coevolution implies ontological novelty through the interaction between interlinked processes that partially (co)shape each other. In turn, specific coevolution implies, aside from novelty, the ontological phenomena of replication and selection. More precisely, specific coevolutionary processes causally codetermine each other, triggering changes in some of the systems’ properties related to replication, selection, and/or variation/novelty. Because of structural or dynamic discontinuities, this codetermination is usually partial and can be reciprocal or asymmetrical.

I argue that general coevolution happens throughout the entire universe, while specific coevolution is only found in biological, social, and cultural processes. Moreover, as we will see, the processes of innovation, replication, and selection behind specific coevolution differ significantly in the biological, social, and cultural realms. Specific coevolution is limited to a small part of reality, but it is within this realm that humans exist, act, and perish. Therefore, it is significant.

With this understanding of coevolution, the following three sections will explore the concept of ontological novelty and its requirements. We will examine the existence of a plurality of layered entities and processes with complex causal interplays. This exploration will form the basis for our subsequent discussion on vertical and horizontal coevolution, which will help us justify the distinction between general and specific coevolution.

3 The importance of novelty in coevolution

Let us discuss the concept of novelty. It introduces us to neology, the study of what is new. This is a complex ontological phenomenon that has divided philosophers for millennia. I will distinguish three main kinds of ontological novelty:

  1. 1.

    Relational novelty: Every instant in the universe introduces some ontological novelty because even if the same entities and processes are present, their arrangement is different.

  2. 2.

    Ontic novelty: From atoms to galaxies and civilizations, new elements and processes are constantly coming into being. Even if they have the same fundamental properties, they are not the same in an ontological sense due to the Stoics’ and Leibniz’ Principle of the Indiscernibles. If two things or processes were exactly the same, they would not be two, but one. Challenges to this principle, such as Max Black’s famous “symmetric universe” with only two spheres, beg the question, as they introduce an imaginary reality whose ontological possibility has yet to be proved (Black 1999).

  3. 3.

    Qualitative novelty: Within these new entities or processes, some entail new properties that are qualitatively irreducible to previously existing entities or processes, for example, chemical compounds regarding quantum processes, living beings regarding chemical matter, and cultural achievements regarding biological processes. We could say that qualitative novelty is a strong or macro novelty, while ontic novelty is a soft or micro novelty. We can find these three kinds of ontological novelty in every known ontological dimension of the universe: physical, chemical, biological, social, and cultural-artificial. Novelty, therefore, is a transcendental attribute of the universe. This attribute is always linked to its nemesis: ontological disappearance or fading. In other words, the ontological novelty of the universe’s states, entities and processes, and qualitative novelty is always linked to the disappearance or fading of former states, entities, processes, or qualitative novelties.

The ontological phenomenon of novelty leads to several interesting philosophical issues, such as the ship of Theseus paradox regarding identity over time in physical, chemical, biological, social, and cultural systems and subsystems. National economies or big companies are studied over long periods as the same evolving entity, even when the populations that compose the nation or the company are replaced, while there are systems, such as family units, which are more sensible to the Ship of Theseus paradox. This paradox has an ontological side (populations as material collections of individuals) and an epistemological side (classes as abstract conceptualization of collections).

4 Flat ontologies miss the point

Exploring the reasons to reject flat ontologies will enable me to further ground the ontology of coevolution by proposing a rich interplay of asymmetrical dynamisms. My point here is clear: If we take the ontological phenomenon of novelty seriously (which is essential in coevolution), we must reject flat ontologies. Flat ontologies are models of reality that deny the asymmetrical, and therefore hierarchical, structure of reality, placing every entity at the same level. Physicalism, as a radical downward reductionism, is a clear example of flat ontology. Some authors like Bruno Latour (2005), though, advocate for a non-reductionist flat ontology that is very different from physicalism. In his approach, everything exists at the same level: atoms, empires, and fictional characters. However, by rejecting structural and dynamic asymmetries in reality, these kinds of ontologies end up with inconsistent metaphysics that cannot account for qualitative novelty and fading (Pérez-Jara 2022).

As we saw in the previous section, there are different kinds of novelty, and all of them are supported by the complex interplay of continuities and discontinuities that structure the universe, from quantum phenomena to sociocultural interactions. This essential interplay prevents us from petrifying any dimension of the universe, an idea I have referred to as the Medusa Effect elsewhere (Pérez-Jara 2022). According to Spanish philosopher Gustavo Bueno (1972), the Platonic principle of Symploké (Sophist 251e–259e) is the first historical defense of this interplay, stating that nothing is absolutely isolated, but not everything is connected to everything else. In the twentieth century, philosophers such as José Ortega y Gasset, A. J. Ayer, and Mario Bunge defended this ontological principle fruitfully (Romero et al. 2022).

While the Platonic principle of symploké focuses on connections, the rich interplay of continuities and discontinuities that structure the universe goes beyond it because there are entities and processes necessarily connected that, at the same time, are partially discontinuous. For that reason, I prefer to talk about an extended symploké to consider those phenomena. We can appreciate this complexity in the ontological phenomenon of emergence, which involves an interplay of continuities and discontinuities beyond connections. Emergence implies qualitative novelty, i.e., the appearance of properties missing in the system’s parts. The universe presents countless examples of emergent properties: the brightness of a star is missing in a star’s components; the capacity of cells to repair themselves is missing in cells’ chemical components; human agency is missing in the body’s organs, molecules, and atoms; and many features of human societies are missing in their individual members. Thus, all systems have qualitative properties that cannot be found in their parts—although those systems and processes cannot exist without them. This means, among other things, that in the ontological phenomenon of emergent complexity, continuity and discontinuity are the two sides of the same coin.

Systems’ emergent properties present a mixture of discontinuity (in the qualitatively irreducible novelty of these properties regarding the components of the system from which they arise) and continuity (since these properties do not arise ex nihilo, but from the system’s internal and external dynamics). For example, mental processes are partially discontinuous and continuous from the brain’s electrochemical activity (Pérez-Jara 2014; 2016; 2022; Ongay 2022).

Against classic Platonic, Neoplatonic, and Medieval pictures of reality, each known ontological level of the universe (physical, chemical, biological, social, and cultural) has the same degree of ontological reality. There is not any reason to hold that physical processes are more real than social systems or reciprocally. They are just different ontological dimensions of the universe. However, there is a hierarchy based on their asymmetrical relationships: physical processes can exist without social processes, but there are no social processes without physical events. The ontological irreducibility of each level is proof of the falsehood of downward reductionism (aggregationism) and upward reductionism (holism). From atoms to solar systems, living beings, and societies, most things surrounding us are more than aggregates or appearances of a metaphysical absolute whole: they are systems, i.e., complex entities whose contents, structure, mechanisms, and environment produce qualitatively irreducible properties lacking in their components. Although ontological continuity is as important as discontinuity, reductionistic approaches, which sadly can be found everywhere, downplay discontinuities. And, for that reason, and inspired in Gustavo Bueno (2016), I call discontinuous materialism my philosophical approach. I do so to vindicate the plurality and ontological novelty made possible by the principle of discontinuity (Pérez-Jara 2016, 2022; Pérez-Jara et al. 2022; Harman and Pérez-Jara 2022; Ongay and Pérez-Jara 2022; Avila-Castellanos 2024). How does causality fit into this plural and discontinuist picture?

5 The many faces of causality in coevolutionary processes

Contrary to popular opinion, causality is not a relation between things but between processes. Some philosophers, from Heraclitus and Buddha to Whitehead and Bertrand Russell, argue that processes are the fundamental elements of reality. Others contend that processes are just changes in the states of material things, like systems (Bunge 2010; Romero 2019). Since the things that compose the universe are constantly changing (i.e., they are always linked to processes), we can talk about a process ontology without taking sides in the metaphysical debate of whether entities are the primary ontological category and processes are derived or reciprocally.

Also with the purpose of dispelling archaic conceptions of causation, some philosophers argue that causal relations should be seen as triadic, not just two-sided (Bueno 1992). I agree that we need to move beyond the traditional cause-and-effect relationship by recognizing that a process acts as the causal determinant, affecting another process and changing it, resulting in an effect. However, the situation is even more complex than this basic triadic relationship. As I mentioned in the introduction of this paper, the processes around us are the result of many converging causes, not just one. I am not referring to an infinite regression; the processes we observe, from the death of organisms to divorce rates in a specific country, are the consequence of multiple simultaneous causes. These effects, in turn, act as causal determinants for new processes. Like emergent complexity, causality is linked to a rich, dynamic interplay of continuities and discontinuities. The plurality of processes implies discontinuity (otherwise, we would only talk about one single process). At the same time, their complex interactions entrain continuity (otherwise, they would be totally isolated, and causality would be impossible).

Since the universe is ontologically plural and layered, causation has two main dimensions: horizontal (i.e., given within an ontological level) and vertical (i.e., between processes belonging to different mereological levels of organization). Horizontal causation is linked to our ordinary, traditional notion of causality: countries cooperating or in conflict, individuals socially interacting, billiard balls colliding, molecular interactions, and so on. Traditionally, vertical causation seems harder to grasp despite its ubiquity. Vertical causation can be down-top (upward causation) and top-down (downward causation). Upward causation describes a situation we find everywhere: wholes ontologically constrained by their parts. In terms of process ontology, there is upward causation when a process causally influences another process that belongs to a higher ontological level. For instance, physical processes causally influence chemical and biological processes.

Upward causality is the essential concept to understanding what I have called elsewhere “reality muzzle(s)” against the excesses of radical constructivism and relativism (Pérez-Jara 2022; Pérez-Jara and Camprubí 2022). Human beings’ operations and constructed social and cultural artifacts (including scientific theories and models) are necessarily constrained, at least, by physical, chemical, and biological realities. Ontological transparency is a myth; our knowledge of the external world is far from immediate. Instead, knowledge always implies a transformation of mind-independent realities. This transformation can be purely cognitive (a butterfly’s cognitive system filtering external stimuli without modifying them externally). But it can also be external, such as scientists studying material systems by breaking them into their parts, whether they are physical, chemical, or biological. These independent realities, whether they are entities or processes, both enable and constrain human existence and knowledge (Pérez-Jara 2014, 2022; Camprubí 2022).

Given the plurality of ontological muzzles, I differentiate between, at least, physio-muzzles, chemo-muzzles, bio-muzzles, and socio-muzzles. It does not matter how flexible and powerful cultural narratives can cognitively, emotionally, and practically be. Any ideology that denies physical, chemical, or biological processes that powerfully alter human existence will crash headlong into harsh reality, no matter how strong their believers’ psycho-social mechanisms of suggestion and brainwashing are. Death is an obvious example of an ontological muzzle. More specifically, it is an example of a bio-muzzle. The ontological impact of this thesis in the economic sphere is obvious: environmental catastrophes or lethal viruses can make the most powerful economies collapse like a house of cards in a glimpse. Moral: Economic processes are ontologically muzzled by extra-economic processes. But this does not mean they are epiphenomenal (i.e., effects without causal power). They affect other sociocultural processes through horizontal causation but also affect infra-cultural realities. And this can be explained by downward causation. It happens when a process causally influences another process that belongs to a lower ontological level. We can distinguish micro-vertical causation within each level: within the social level, governments influence families, and families shape individuals reciprocally; within the biological level, organ systems determine cells and cells regulate their DNA sequences, combinations, and orders. And we can talk about macro-vertical causation when causal processes involve different ontological levels (such as physical, chemical, biological, social, and cultural). For instance, emotions can raise the body temperature, and typing on a computer redirects the flow of electrons in specific ways.

It is important to note that there is a structural asymmetry between upward and downward causation. Upward causal processes have more power of determination than downward ones. Life is only possible under very specific thermal, density, and gravitational conditions. A drastic physical or chemical change could wipe out all life on Earth in an instant. However, even with nuclear technology, living beings do not have the power to alter stars, galaxies, or the fundamental nature of the universe.

6 Vertical and horizontal coevolution

Based on the general idea of causality I explained earlier, we can categorize coevolution into horizontal and vertical types. Horizontal coevolution occurs at the same ontological level and can be further classified into physical, chemical, biological, social, and cultural coevolution. In turn, vertical coevolution takes place between processes at different levels of existence. For example, some biological and sociocultural processes exhibit vertical coevolution in the form of the Baldwin effect, niche construction, or epigenetic changes. This is referred to as macro-vertical coevolution. Additionally, micro-vertical coevolution occurs within the same level of existence, such as within biological matter.

In order to properly understand these relations of codetermination, we have to pay attention again to the ontological phenomenon of asymmetry. Vertical asymmetries are static because asymmetries between ontological levels do not change: social systems will always imply physical processes but not reciprocally. On the contrary, horizontal asymmetries are dynamic because they vary depending on different contexts: whether two national economies codetermine each other reciprocally or not depends on dynamic factors such as the volume and influence of each one of them.

Vertical coevolution is always asymmetrical, but horizontal coevolution can be or not, depending on the circumstance. Important to note that at each ontological level, there is also micro-vertical coevolution. For instance, there is micro-vertical coevolution between governments and families, families and individuals, organ systems and cells, cells and DNA sequences, and combinations and orders.

This ontological multidimensionality correlates epistemologically with the necessity of scientific interdisciplinarity against the above-mentioned epistemological reductionisms (physicalism, sociologism, biologism, psychologism, economicism). We have seen how the ontological dimensions of entities and processes are structurally asymmetrical, against the general symmetry principle held by some flat ontologies. This asymmetry is both general and specific. Macro-asymmetries can be found between physical matter, chemical matter, biological matter, social matter, and cultural-artificial matter in general: social systems imply physical, chemical, and biological processes, but physical processes do not have social properties.

On the other hand, micro-asymmetries are more variable and depend on more specific and dynamic contexts—let us remember the example of the causes of depression mentioned in the Introduction of this paper.

7 General coevolution in society and culture

The previous sections have outlined a coevolutionary ontology that applies to all processes, including social and cultural ones. We can identify general coevolutionary processes in society and culture when we recognize processes that mutually influence each other in a way that leads to ontological novelty. Numerous examples of new social and cultural realities that emerge as a result of dynamic mutual influence exist, such as the emergence of the family institution, trade, religion, writing systems, the state, and various technologies. The emergence of new cultural institutions and patterns plays a crucial role in shaping society, and vice versa. In other words, society and culture coevolve.

A brief clarification about the difference between society and culture: social systems are material entities composed of interacting sentient living beings. Every social system has a cultural dimension made of socially learned meanings in humans and other animal species. A family, for instance, is both a social system (composed of individuals) and a cultural system (composed of meanings such as family values, norms, and ideological discourses). The meanings that compose culture can be attached to natural things (such as stars or the forests) and to artificial things, from dances and language to the so-called artefactual culture (Alexander 2010; Camprubí and Pérez-Jara 2023).

Social systems are multidimensional realities: they have physical, chemical, biological, and specific cultural properties. Every society, as a global system, comprises sub-social systems composed of populations of individuals. Particular societies also interact coevolutively with social mega-systems such as imperialism, international institutions, conflicts, and cooperation.

Coevolution rejects what I call vertical and horizontal epiphenomenalism or predeterminismin. Macro-vertical coevolution shows vertical epiphenomenalism’s falsehood, be it in the philosophy of matter (biotic matter as an epiphenomenon of physical matter), the philosophy of mind (mental processes as epiphenomena of electrochemical processes in the brain), and the philosophy of culture (culture as an epiphenomenon of social structure or biology).

On the other hand, horizontal coevolution demonstrates horizontal epiphenomenalism’s error. Mutual codetermination within the same ontological level (for instance, in human societies) bestows a significant degree of causal power to all the systems and subsystems implied in the coevolutive process. For that reason, neither the economic sphere can predetermine the political or the religious sphere nor reciprocally.

This thesis is nonetheless rejected by many approaches to social and economic dynamics, from Confucianism and its attempt to explain everything from the family sphere to libertarians reducing every social dynamic to the thesis of homo economicus. Aside from these radical cases, we can also find partial reductions from some spheres to others: reactionary thinkers such as Juan Donoso Cortés reducing the political sphere to the religious sphere, or progressive thinkers such as many Marxists who reduce the religious sphere to the economic sphere. We can trace the thesis of the irreducibility of social and cultural spheres to Max Weber, who observed that every society comprises multiple, relatively independent but interacting spheres of life.

The thesis of the relative autonomy of the different cultural spheres I defend implies that no cultural sphere is an epiphenomenon, i.e., an effect without any causal power. These considerations about cultural spheres can be said about culture in general in relation to non-cultural entities and processes. Cultures are not epiphenomenal mirrors of social dynamics, environmental factors, or biological processes. There is an essential ontological difference between determining and predetermining. Predetermining implies epiphenomenalism and fatalism, while (co)determining allows space for contingency and, at least, bidirectional causality (although, as we have seen, most causal processes are, in truth, n-directional).

Cultural processes affect many non-cultural realities, from the biosphere to many chemical and physical processes from the Earth. The theses of global warming and the Anthropocene Age are clear indicators of human culture’s causal power over non-cultural things. Social systems are semi-autonomous because no social system is an epiphenomenon of other(s). At the same time, social systems co-influence each other, although in different degrees. The epistemological consequence is that to study a social system isolatedly or statically is as misleading as looking at a social system as the epiphenomenon of other(s) social system(s). The coevolutionary relationship between social systems is dynamic, semi-discontinuous, usually asymmetrical, and linked to the emergence and disappearance of properties.

8 Specific coevolution in society and culture

Until now, I have analyzed the main features of processes that codetermine each other horizontally and vertically involving ontological novelty. However, dynamic mutual influence alone is insufficient for discussing coevolutionary processes in human societies and cultures. This is because, in addition to novelty or variation, coevolution in biological, animal, and human spheres involves two essential ontological phenomena: replication and selection. This is what I called above specific coevolution. While we find novelty at every ontological level, replication and selection are absent at the physical and chemical levels. Scientists replicate physical and chemical processes in their experiments and with their technology. However, these processes cannot replicate themselves without human intervention. In other words, physical and chemical replication does not exist, and the same with selection.

Ontological novelty, and, more specifically, qualitative novelty, is the condition that makes replication and selection possible. Biological matter enables replicative and selective processes, which in turn increase the ontological complexity of biological matter (Ongay 2020). This ontological feedback can also be found in social and cultural levels, where we also find replication and selection. Let us briefly discuss replication. I will start with ontological replication. We can categorize replication into different types based on the ontological level at which it occurs and what is being replicated. Considering the first aspect, we can differentiate between biological replication, social replication, and artificial replication. Considering the second aspect, we can distinguish between the replication of entities, processes, or properties. As I mentioned when discussing ontic novelty, the Principle of Indiscernibles shows us that absolute replication is impossible. Every replica or copy is, at the very least, slightly different from its predecessor. This means that every replication process introduces some degree of novelty. Replication processes are mainly unconscious; they are totally unconscious at the biological level, but they can be partially conscious at the social and cultural-artificial level. Theories that analyze most social and cultural features through a Darwinian lens, known as Pan-Darwinist theories of society and culture, can lead to dangerous distortions. These theories fail to recognize that in the ontological phenomenon of emergence and the appearance of a new ontological level, some properties remain, while others are lost, and others are gained (the emergent properties). Thus, although it is obviously true that a society whose social and cultural features lead to total unsustainability will stop replicating itself in the long run, social and cultural features appear new and are replicated and selected by mechanisms that diverge from biological selection, despite some partial analogies. This notion can also be applied to animal cultures, which has proven to be much more complex than previously thought (Viciana 2021).

Even biological replication and selection is a far more complex phenomenon than biologists thought decades ago. Today, we know that there are inheritable epigenetic changes, non-genetic inheritance mechanisms (such as intestinal flora), along with a prevalence of maladaptive and vestigial features that are nonetheless replicated. Due to the emergent properties of the social and cultural-artificial levels, social and cultural replications and selections are even more complex and less linear than at the biological level. Rom Harré’s “Lamarkian logic” of ideologies (which “mutate” to adapt to social environments while they are still culturally “alive”) is an example of how the evolutionary logic of ideologies and philosophies do not follow Darwinian principles but in a very vague sense (Harré 1993; Pérez-Jara and Camprubí 2022).

In the economic sphere, I. Almudi and F. Fatas-Villafranca give the following definition:

Replication (...) is a specific type of process through which some traits (habits, routines, technologies, norms) of interacting intentional agents diffuse at different rates across the population of agents, under the conditions of causal implication, similarity and information transfer (…) The processes of learning, emulation and imitation that usually appear in economic models may be conceptualized as examples of replication in this sense. […] (Almudi and Fatas-Villafranca 2021).

Agential replication, a partially conscious process, cannot be found in the most basic forms of biological matter. Moreover, as I have already defended, there is no trace of replication in chemical and physical matter.

Finally, let us now talk about ontological selection. Similarly to replication, we can differentiate between (A) biological selection, social selection, and cultural-artificial selection and (B) selection of entities, processes, or properties. Biological selection has two interconnected sides: We can distinguish between non-agential selection and agential selection (such as trophic and sexual selection), both of which fall within the general scope of natural selection. However, it is important to note that this distinction should not be oversimplified. Biological selection typically occurs unconsciously. While sexual and trophic selections involve a degree of agency, their outcomes are unpredictable. Animals have no clue about the evolutionary implications of their sexual or trophic selection processes. Even when humans engage in artificial selection of animals, the results can be unforeseeable. For example, farmers or dog breeders may artificially select animals based on observable physical traits, but these traits are often linked to other invisible, sometimes undesirable, inheritable features (see Ongay’s paper in this volume). We cannot even equate agency with conscious experience. Recent research indicates that unconscious perception and cognition play a more prominent role in humans than previously understood (Reber and Allen 2022; Dennett 2018; Eagleman 2012). It is also possible that primitive animals possess perceptive and cognitive processes without conscious experience, as may be the case with insects (Adamo 2016a, 2016b).

Social selection has features that transcend beyond biological selection. There is, though, an intersection with biological selection through sexual selection. The concept of “sexual selection” was first articulated by Charles Darwin, who wrote about a “second agency” other than impersonal natural selection. On average, females sexually select males to mate, following patterns that Ronald Fisher started to study mathematically in the early twentieth century. Sexual selection leads males to make extreme efforts to demonstrate their fitness, resulting in sexual dimorphism in secondary sexual characteristics. The result is that, in most species, only a minority of males are selected. Even in the human species, we have more female ancestors than male ancestors.

However, even if we identified social selection with sexual selection, it is essential to emphasize that in human cultures, sexual selection sometimes converges and sometimes diverges with biological selection. Not only innate, hard-wired preferences but contingent cultural factors are behind shaping sexual preferences in human societies. The situation becomes even more complex when we consider other forms of social selection resulting from various sociocultural factors such as aesthetic, economic, and ideological influences. Discrimination based on religious, racial, or political dogma is an example of social and cultural selection that completely differs from the purposes and mechanisms of biological selection. On the other hand, intolerance of hate speech, racism, sexism, or defamation, as outlined in Karl Popper’s Paradox of Tolerance, is another well-known mechanism of social and cultural selection. This mechanism determines which social and cultural features will be replicated and which will not. And, again, such mechanism is very different from biological selection.

In the economic sphere, Almudi and Fatas-Villafranca define selection in the following way:

Selection in evolving economic systems involves a previous set of entities (firms, civil organizations, institutions) that are transformed into a posterior set, in such a way that all members of the posterior set are sufficiently similar to the members of the anterior set and the resulting shares or population frequencies of posterior entities are positively related with certain domain-specific competitive traits embodied in the agents under selection. These domain-specific traits condition the agents’ degree of adaptation to the environment and the fulfillment of their goals and influence the population structures that emerge from competition (these traits include firm technologies, prices, performances, individual skills and habits). (Almudi and Fatas-Villafranca 2021)

Once again, this definition could be seen as a type of selection within the cultural-artificial level. Biological, social, and cultural-artificial replication and selection processes partially influence certain chemical and physical processes through downward causation. A clear example of this is anthropogenic climate change. Conversely, some chemical and physical processes also partially influence biological, social, and cultural-artificial replication and selection processes. Social and cultural changes due to climate change provide an example of this. Furthermore, the political, economic, and social impact of specific viruses or rare earth scarcity also exemplifies this reciprocal influence. Even though human social and cultural selection processes suggest greater awareness and predictability, they are always susceptible to unforeseen consequences. Many social and cultural processes develop in unforeseeable ways, taking on a life of their own beyond humans’ initial calculations.

The dynamic interaction between novelty, replication, and selection allows for new forms of coevolution. In the economic sphere, Dosi and Nelson (2010) and Nelson et al. (2018) have provided numerous rich examples of this dynamic interplay: the medical devices sector coevolves with medical practice schools and advanced research hospitals; electronics, computer hardware, and software advance through interactions across corporations, intersectoral supply–demand dynamics, regulatory agencies, and public research programs, among others (Almudi and Fatas-Villafranca 2021). My approach goes further by incorporating coevolution beyond the human sphere: the Coronavirus coevolved with complex political, economic, social, and cultural dynamics (Rao 2020). Similar things can be said about climate change, terrestrial biomes, and so on. It is important to note that coevolution does not involve symmetry, as I have emphasized. Most coevolutionary processes are asymmetrical. When asymmetry is disproportionate, the codetermination itself almost fades away. For instance, the sun and its radiation determine life on Earth (including human politics and economy), but life on Earth has a minuscule effect on the Sun.

9 Conclusion: What do we do with Occam’s razor if reality is so complex?

We are now in a much better position to understand the initial examples that opened this paper. Despite being microscopic and lacking life in biological terms, a rapidly reproducing virus can cause entire economies to change and even collapse drastically. At the same time, social and cultural changes modify the patterns of the spread of the virus and even the virus’ structure itself. Some aspects of the current economic system can contribute to the rise in global temperature, which, in turn, affects economic processes. Air pollutants can lead to increased infertility, triggering demographic and economic changes, including shifts in economic consumption patterns related to babies and children. Some of these changes in sociocultural patterns can affect the air quality.

We could all provide many more examples. All would reflect the same concept: dynamic codetermination processes across various ontological domains. We can only comprehend this dynamics through a multidimensional epistemological approach, such as the coevolutionary perspective. This means that coevolutionary economics, like any other science, cannot exist in isolation. Economic phenomena are interconnected with other sociocultural processes, from biological to chemical and physical ones, in a complex and dynamic matrix of multiple causal relationships.

Philosophically, this involves constantly critiquing reductionistic and one-dimensional explanations of economic phenomena. From a scientific standpoint, it means embracing interdisciplinarity and breaking down the rigid barriers between academic departments and disciplines. The goal is not to arrive at a single unified science, as Leibniz or the Vienna Circle once envisioned, but rather to epistemologically consider the complex interplay of continuities and discontinuities that shape reality. Both holism and isolationism are not only incorrect ways of understanding existence but also lead to epistemological dead ends.

In this complex context, I have argued that we can discuss ontological (co)evolution in both a general and specific sense. General evolution can be observed in physical and chemical matter. Before the emergence of living beings, the universe itself evolved, gaining more ontological complexity and qualitative density. Due to the complex interplay of continuities and discontinuities within physical and chemical matter, these entities and processes are multifaceted and mutually influential. Inspired by Gustavo Bueno’s work, I have named this interplay an extended symploké, which also applies to the biological, social, and cultural levels.

We can observe general coevolutionary processes at every level of existence. However, specific coevolutionary processes, which involve novelty, replication, and selection, are only evident from the biological level and beyond. This is due to the fundamentally unique ontological novelty introduced by biological matter and later by social and cultural-artificial matter in the universe. There are variations in the ontological phenomena of novelty, replication, and selection in biological, social, and cultural-artificial matter. For example, in cultural spheres, qualitative properties associated with novelty, replication, and selection are not present at the basic biological and social levels. For instance, in the realm of cultural spheres, we encounter artifacts, values, and norms that partially conscious agents have intentionally designed. This introduces a level of emergent complexity that is absent in simple biological coevolutionary processes, which occur without any conscious design. Even more fascinating is that novelty, replication, and selection at biological, social, and cultural-artificial levels mutually influence each other through multidimensional causation. This mutual influence, and therefore coevolution, occurs both horizontally and vertically.

Dynamic, plural, and asymmetrical (and therefore hierarchical). This is how reality is. However, like Saint Anthony’s temptations in the desert, one-dimensional, monocausal, and reductionistic explanations are highly seductive. Like siren songs, they whisper to the ear that philosophers tend to complicate things too much, that the beauty of reality is that it is, in essence, simple and elegant. Proper knowledge must be guided by the principle of parsimony, often linked to Occam’s Razor. This principle suggests that we should prefer the simplest one if we have multiple explanations for a given data set. In other words, we should not make unnecessary assumptions nor over-complicate factors if a simpler explanation can suffice. In Occam’s words: Entia non sunt multiplicanda praeter necessitatem, i.e., entities are not to be multiplied beyond necessity [my emphasis]. Very true. But what happens when it is necessary to multiplicate entities? The history of science is precisely the clearest proof of the necessity of multiplicating entities (and hypotheses and variables). Empedocles’s theory of the four elements worked for millennia until the development of modern chemistry. From four simple elements, we passed to a table of 118 periodic elements. Hardly anyone would condemn this as an unnecessary over-complexation. About the same holds true for taxonomies and species in biology, ethnic groups, religions, types of social organization, and so on. Our ancestors believed that reality was much simpler than what came to be in truth. Nothing can make us believe that this process will not continue in the future. New biological mechanisms, species, galaxies, past societies, and physical entities will be discovered.

None of these examples violates the principle of parsimony. In good science, entities or processes that are not necessary to explain a phenomenon are discarded. Because of this, astrology is considered a pseudoscience, and the belief that we need aliens to explain ancient architectural artifacts is unfounded. However, has all this complexity made our understanding of reality more difficult? Not at all. Simplistic understandings of complex realities are not real understandings. Epistemological complexity is not an obstacle to our understanding of reality; it is the opposite. Theories that consider more real variables and factors allow us to understand reality. Increased complexity is an epistemological asset when it is correlated with ontological complexity.

The world is incredibly complex ontologically but not epistemologically unmanageable. There are millions of biological species, each with its own unique characteristics. However, they can all be categorized into 8 biogeographical realms and 14 biomes. There are billions of human beings, but only a few main psychological profiles. There are more than 200 countries, but a far smaller number of main political and economic systems. Similarly, in ontology, we can classify the myriad of ontological entities and processes that make up the universe into a few ontological realms: physical matter, chemical matter, biological matter, social matter, and cultural-artificial matter. Perhaps in the future, we will add new types of matter to our ontological framework, but it will still be a manageable diversity from an epistemological perspective. By studying part-whole relationships in systems, we can analyze what would otherwise be an unmanageable complexity. This, along with coherent theories that, as much as they can, eliminate ambiguity and vagueness, allows us to domesticate complexity.