Skip to main content

The Foundations of Creativity: Human Inquiry Explained Through the Neuro-Multimodality of Abduction

Abstract

This chapter offers arguments in favor of a morphological characterization of situated abductive processes in perception, considering them as adaptation mechanisms to the varieties of experience. The mechanism that has been analyzed in this essay is the creativity. The thesis defended in this chapter is that the human being maintains a constant hypothetical openness to adapt to the uncertainty of the future. Characterizing creative processes using abduction means analyzing this phenomenon from morphological bases. Therefore, the inevitability of creativity defended in this chapter is situated in the neurochemical and morphological bases of perception to characterize the adaptive dimension of creativity through abduction. The starting point of this proposal is based on the EC-Model of abduction: a contextualized interpretation of classical pragmatism from the naturalization of cognitive processes. Abduction is proposed as the simple mechanism of hypothesis generation and its selection. This mechanism can be considered present in all degrees of human experience: the generation of epistemic content is grounded in natural biologically based adaptive processes. Therefore, there is a coupling between embodied mechanism of cognition and sociocultural values. Because of its fundamental value, the implementation of abductive mechanisms in Machine Learning systems is explored, especially in most recent Deep Learning models.

Keywords

  • Determinate indeterminism
  • Abductive cognition
  • Neuro-multimodality
  • Inevitable creativity
  • Human inquiry

Informi esseri il mare vomita

Sospinti a cumuli su spiagge putride

I branchi torbidi la terra ospita

Strisciando salgono sui loro simili

E il tempo cambierà i corpi flaccidi

In forme utili a sopravvivere

(Banco del Mutuo Soccorso, 1972, L’evoluzione, 04m26s)

Introduction: Morphological Determination and Cognitive Inevitability

During the second half of the twentieth century, the revolution held in cognitive sciences implied a new way of understanding cognition. It implied the naturalization of the field, looking for the natural mechanisms that could explain human thinking. The resulting field is also summarized as 4E Cognition (for Embodied, Embedded, Enacted, and Extended), which reintroduced body and environment into the cognitive equation (Newen et al., 2018). Classic top-down symbolic approaches to cognition were heavily affected and criticized by these new perspectives. This situation changed with technological revolutions in neuroscience such as the scanning technologies. With the possibility of checking in vivo brain functioning, new data about brain performance were made accessible, paving the way for new revolutions. The most salient of these data offered the possibility of understanding the fundamental role of emotions in cognition (Damasio, 1994). Such work opened the understanding of cognitive processes as bodily processes, as interactions of a soup of hormones and neurotransmitters created to manage social interactions. That revolution also had an impact on robotics and Artificial Intelligence fields (Kurzweil, 2000), opening the path to the area of affective computing (Picard, 2000) and social robotics (Breazeal et al., 2016).

In any case, the neuroscientist successes breasted a bias: an anthropomorphization of the field, as well as a neurocentrism. In addition, the area experienced a lot of hype because funny (but serious) experimental biases such as the Dead Salmon fMRI (Lyon, 2017) could not be diminished. Nevertheless, despite the influences of both anthropocentrism and braincentrism, the field has evolved thanks to the research on minimal and plant cognition (Calvo Garzón, 2007), opening new ways to understand cognition as both natural and social processes (Vallverdú et al., 2018).

The analysis of the evolutionary and functional basis of cognition, from minimal to highly complex systems, helped to understand the role of morphological constraints in the process of knowing. The primary elicitors of this revolution were the requests of the robotics researchers, who were trying to design a scenario that allowed a new step toward the creation of intelligent machines (Pfeifer & Bongard, 2006).

Therefore, the role of the morphologies was crucial for understanding the properties of a cognitive system. This statement had another reading: the morphology afforded specific ways of processing information and, therefore, creating cognitive mechanisms. In this naturalistic and evolutionary perspective, the authors affirm that cognition is inevitable (although dynamically evolvable, in some cases). Furthermore, according to local environmental variations, such morphologies can act differently as an ecological constraint that shapes cultures (Nisbett, 2004).

Therefore, how living or artificial cognitive systems can select, process, and react to information is morphologically constrained. That morphological design makes social interaction and intentional behavior possible, which is not a “learned mechanism.” Throughout human history, the hot ethical and religious debate about freedom vs. determinism has been discussed from an ethical perspective. The only existing determinism is the cognitive horizon mediated by morphological constraints by which human beings feel and understand the world. It is essential not to separate the processes of knowing and feeling because both processes are intrinsically connected in experiencing the world and ourselves (Koch, 2019). This chapter offers an approach to biological determinism to have an indeterminate cognitive predisposition based on abduction. Such a predisposition is abductively articulated because a weak but necessary relation between knowing and feeling is embedded in perception as a constant openness experience system.

Abduction is traditionally understood as a form of reasoning: a complex inferential cognitive process. One of the traditional forms of this reasoning is scientific discovery. As it is known, abductive reasoning was recovered by Peirce from the Aristotelian apagōgē (ἀπαγωγή) (Aristotle 1957: An. Pr., II, 25, 69 to 20–35) as the third syllogism (Peirce, 1958, CP: 5.14–40): the hypothetical or possible inference. As the authors will explain in the second section, abduction is the cornerstone of pragmatism: a philosophical theory that assumes an ontological perspective of the unfinished world. Therefore, it is a philosophical system to face the uncertainty of becoming. One of the main characteristics of Peirce’s abduction is that it is oriented to complete and to characterize perceptual judgment (Peirce, 1958, CP: 5.348). This characteristic has notoriously influenced current inquiries on abduction. For example, this form of inference was recovered to disentangle the debate in the philosophy of science on the context of discovery. Nevertheless, an analysis of pragmatism may offer clues to reinterpret abduction as a mechanism of hypothetization inherent in perception.

This pragmatic perspective is approached from an epistemological holism. The main idea is to show how the most complex forms of hypothesizing depend on the most basic ones and vice versa. In this way, it can be argued that there are micro-processes of hypothesizing in perception. This view is defended from different perspectives of classical pragmatism. However, Peirce and Mead are the authors who treated it in more depth. The reason for using Peirce’s theory for the question of hypothesizing processes in perception is based on his characterization of the abductive process situated in firstness: a non-cognitive, inexplicable, and immediate qualitative state of feeling or sensation. Likewise, Mead emerges as an indispensable author to situate the category of firstness in perception. The primary motivation for relaunching this genealogy of classical pragmatism toward a foundation of the abductive processes of perception lies in the need to broaden the philosophical tradition from which the current debate on abduction is articulated.

The perspective of recent abduction from which our proposal is articulated is the EC-Model of abduction: an enactive cognitive proposal based on contextualization and anthropomorphization of the agents who know as they interact with the environment. The critical point is that the anthropomorphization of cognition is based on the naturalization of logic: abductive management of heuristic information. The proposal to apply this model of abduction to perception implies extending the influence of factors such as emotions, feelings, and narratives to characterize hypothetical manipulation as an unreflective process prior to cognitive ones – section three attempts to characterize many abduction-based cognitive hypothesizing strategies as creative processes. Two forms of creativity are distinguished: genuine generation and divergent adaptation to surprising situations. The second type of creativity can be characterized abductively in phases of complex reasoning as divergent hypothesizing in the face of methodologically surprising cases. It is also possible to understand divergent hypothesizing processes in perceptual phases, for example, unreflective resources that arise during manipulation. In the same section, triggers and constraints are characterized: both articulate abductive processes. They are logical and cognitive mechanisms that represent the operational bivalence of morphological and cultural aspects in defining the margins of our reality (constraints) and the process of change (triggers).

What has been said so far allows us to argue in section four that a naturalistic and evolutionary perspective needs to incorporate abduction as a morphologically embedded mechanism. Thus, abduction appears as an embodied mechanism for explaining creativity. Consequently, because of the fundamental role of creativity in the epistemic process of knowledge generation, it has to be undertaken that abduction must be assumed to play a fundamental role in creative processes. A proof of this lies in the attempt to capture this epistemic dimension from evolutionary and cultural perspectives for implementation in IA systems. As shown at the end of section four, bioinspired mechanisms of abductive reasoning are a fundamental key to the phylogenetic understanding of cognition, which must include abduction.

The Key Role of Abduction as Cognitive Switch

From Complex to Minimal Abduction: A Genealogy

As previous section has shown, there is a very particular form of determinism: living beings are determined to maintain an indeterminate state of adaptation. The two biases presented, neurocentrism and anthropocentrism, aim to point out the problem of embodying cognition from a single definition. In other words, there are many bodies and many ways of being concerning the environment. This statement goes a step beyond adaptability in speciesist terms and focuses attention on the individual. In this sense, biological, social, and accidental circumstances imply substantial environmental differences. Genetic circumstances are the conditions ascribed in the DNA (genotype) and their phenotypic expression modulated with the environment. There are also losses of faculties that the genotype may determine. For example, it is necessary to have at least one of the celiac disease alleles (HLA-DQ2 and/or HLA-DQ8) for the disease to manifest. However, the disease is not active when it is not phenotypically expressed. On the other hand, accidental circumstances are the loss of psychological and physical faculties. The facultative loss can be natural and progressive caused by oxidation (old age) or abrupt, such as accidents (injuries, contusions, etc.), intoxications, and diseases. Finally, social circumstances are the cosmovision in which the human being lives: the natural environment measured and transformed by the actions of people who are part of the same context. This situation highlights the need to address the anthropomorphic and neural issue of cognition in a way that considers this degree of divergence between living beings of a given species.

Thus, the challenge is to characterize a human’s embodiment of cognition (anthropomorphic and neural) that does not fall into the biases mentioned above. This chapter takes abduction as the genuine mechanism for managing cognition’s neural and anthropomorphic connection with living beings’ open and constant relationship with the environment. In particular, the authors are interested here in the role of abduction in what has been defined as “minimal cognition”: all living systems’ characteristics cover a vast cognitive spectrum that fills the gap between the mindful and the mindless about interacting actively with the world, which requires an embodiment consisting of a sensorimotor coupling mechanism that subsumes an autopoietic organization.

Open door with the dimension of minimal cognition compels a kind of genealogy of abductive cognition. This is because the contemporary debate on abduction has been directed toward complex cognitive processes, such as discovery, hypothesizing, and guessing (Gabbay & Woods, 2005; Aliseda, 2006). Typically, situations that trigger abduction are understood as surprising, puzzling, etc. Classic examples would be Le Verrier’s discovery of the planet Neptune (Sans Pinillos, 2017) and Kepler’s extrapolation of the elliptical orbit of Mars to the Solar System (Hanson, 1972: 72–85).

Likewise, other more common situations are cases in which complex abductions are generated. For example, the way a person dresses can lead us to think about his or her profession (Thagard, 1988: 54–56). Similarly, factors such as skin color, hairstyle, way of dressing, moving, and acting (smoking, speaking loudly, etc.) can trigger hypotheses that determine a course of action. Examples are “not crossing a street,” “not passing through an alley,” “not renting an apartment,” or “accusing someone of a crime.” An autobiographical episode in which Peirce had his watch stolen shows how some of these factors can be –abductively – crucial in determining who the thief was (Sebeok & Umiker-Sebeok, 1983). Of course, these factors and many others can come from prejudices (racism, aporophobia, homophobia, etc.), assumptions, emotions, and feelings. However, it is essential to see that the final result is a complex process of hypothesizing in which abduction is its cornerstone. For this reason, this inference is often referred to as abductive reasoning: the mechanism by that surprising situation (they cannot be reacted to in a usual way) can be tentatively managed from a non-classical epistemic virtue (in short, which does not participate in the classical processes of verification, justification, falsification, contrasting, etc.). From this perspective, abduction is concerned with managing the experience of novel facts through a hypothesis generation process.

The role of abduction can be better understood with a summary of its development in the contemporary debate. Abduction was one of the concepts that were inherited along with the aporetic debate on the context of discovery. This problem stated that it was impossible to represent discovery because it was not logical but psychological. Therefore, only those results that could be justified were considered scientifically relevant (verifiable): scientific language was the best to represent true knowledge because of its descriptive capacity (Niiniluoto, 2014: 378). This situation changed with the possibility of formalizing synthetic (ampliative) reasoning brought about by the computational paradigm and AI’s emergence. One example is formalizing sensible knowledge by describing and automating heuristic processes (Simon, 1985). The programs BACON, GLAUBER, STAHL, DALTON (Simon et al., 1997), and the EURISKO system (Lenat & Brown, 1984) were models of human discovery based on heuristic relations built on ampliative inductive inference.

The motivation for bringing heuristics into creativity is its functional connection to abduction. These two concepts have been intimately related to the discovery debate. The reason for this lies in need to rationalize abduction: to find a way to represent hypothetical inference. For this reason, heuristics is characterized as “the method -neither totally rational nor blind- of discovery to characterize the selective search with reliable results” (Aliseda, 2006: 16). Research in computational science and AI has led to the conceptualization of the role of heuristics in the discovery process. The relationship between discovery and heuristics stems from the classical relationship formulation between – universal – axioms (analysis) and sensible knowledge (synthesis) (Hintikka & Remes, 1974). In other words, it aims at establishing a kind of dialectical relationship that allows explaining how the general rule (heuristic) is extracted and experimentally confirmed. Likewise, once confirmed by experience, the regularities are understood as universal rules (Sans Pinillos, 2021). From this perspective, abduction is understood as the inference proper to the hypothesis (Thagard, 1988: 51–52): the mechanism that can relate the regularities extracted from experience (induction) to universal axioms (deduction) and vice versa.

This first reinterpretation of the discovery debate made from inquiries on computation and AI has allowed us to consider heuristics as more than a way of relating information. Thanks to the unification of AI and cognitive science, the possibility has also arisen to understand some cognitive processes as heuristics. This unification has made it possible to address a fundamental challenge: a formalization of ampliative inferences (induction, inference to the best explanation, abduction, etc.) that consider the cognitive richness of complex processes such as discovery. This challenge has forced the incorporation of additional elements hitherto been considered spurious because of their lack of descriptive content (Putnam, 2001: ch. 1). The generation of novelty and creativity are two cases that cannot be explained through the accumulation and combination of information (heuristic or diagrammatic) (Boden, 2004: ch. 8). At this point, the emergence of a new approach to “being there” (Dasein) in which the classical characterization of abduction plays a predominant role converges. This final consequence is directly linked to the abduction under the actual Pragmatist project.

Classical Pragmatist Holism

Pragmatism is a philosophical theory of managing a relationship with a world conceived as unfinished. Pragmatism assumes the ontological thesis that the world is known from a continuous flow of information. For this reason, Pragmatism is essentially fallibilist: continuous experience incites us to maintain an attitude of provisionality toward the knowledge that humans possess so that it is possible to revise and adapt our beliefs to the new circumstances experienced (Sans Pinillos, 2021). Thus, abduction (as the mechanism of Pragmatism, cf., Peirce, 1958, CP: 5.180-212) is prima facie, an open-ended inference of hypothetical generation and selection process (Kapitan, 1997) focused on managing the variations of contingency. As mentioned above, abduction is the mechanism of hypothetical inference by which surprising situations are managed from a non-classical epistemic virtue. As it is well known, Peirce places abduction at the heart of his pragmatic theory precisely in order to ground sensible knowledge (Peirce, 1958, CP: 5.348). His characterization of abduction is that of the continuous mechanism of generating hypotheses to infer a reasonable conjecture (Peirce, 1958, CP: 2.619–644) that allows an open, in turn, controlled approach to what is still unknown.

To properly understand the pragmatist project, it is crucial to understand that hypotheses have a relevant epistemic role precisely because they arise in situations where no other type of more specific knowledge can be expected. As it has been said, these situations are those of surprise, bewilderment, etc., because the classical epistemic process cannot reach a satisfactory resolution. Abduction generates knowledge to complement percipient moments (Peirce, 1958, CP 5.41–56). Thus, knowledge is not a passive act-reflection but an activity that all human beings develop in our interaction with the world. This does not annul the possibility that there are other simpler processes at work in the complex epistemic processes, which the authors call active act-reflection: unconscious intentional reactions. For example, emotions or feelings predispose the agent to interact from different affective modes. The reason for highlighting Peirce here from among the pragmatists is to emphasize that adaptivity is found in the early formulations of abduction.

Peirce’s adaptivity is manifested through language, which maintains an iconic semiotic relation of resemblance between signs and meaning (Dingemanse et al., 2020: sec. 2.2). One of the attractions of resemblance is its potential to relate things that humans do not yet know about their icon. Therefore, it can be affirmed from Peirce’s Pragmatism that all languages are iconic in the sense that they function, assuming their potential capacity for resemblance. However, such resemblance also introduces some biases, such as the belief in supranaturalistic agents, due to the cognitive mechanisms of anthropomorphization of reality. Consider, for example, pareidolia or transfer of analysis from social to natural events (Willard et al., 2022): both are examples of innate social cognitive mechanisms used by human agents to transfer and find sense of the external world. Likewise, inquiry in robotics has opened debates on anthropomorphism, situational relevance, and interaction with external entities and the environment (Müller & Hoffmann, 2017).

One way to push the limits of these biases is to take the more general perspective of pragmatist holism. The intention is to show that the generation of each agent’s cosmovision is closely linked to both the worldview and the general community’s cosmovision to which one belongs (cosmovision is understood as the unified image of the world, the product of the sociocultural and biological interpretation of the environment, and worldview as the image generated solely from perception) (Magnani et al., 2021). Therefore, mediation with the environment is highly influenced by biological and social factors that the agent cannot control, of which he or she often does not take into consideration or is simply unaware. Putnam picks up pragmatic holism:

  1. 1.

    Knowledge of particulars (facts) presupposes knowledge of theories.

  2. 2.

    Knowledge of theories presupposes knowledge of (particular) facts.

  3. 3.

    Knowledge of facts presupposes knowledge of values.

  4. 4.

    Knowledge of values presupposes knowledge of facts.

  5. 5.

    Knowledge of facts presupposes knowledge of interpretations.

  6. 6.

    Knowledge of interpretations presupposes knowledge of facts (From point 1–4, Putnam, 2001: 136–137; from point 5–6, Putnam, 2006: 33).

As can be seen, knowledge of facts depends on theories, values, and interpretations. In other words, a “fact” is interpreted, so that: 1) every fact is experienced from a theory, and 2) every theory presupposes and assumes some facts. 3) Every theory presupposes values that allow it to justify, falsify, corroborate, etc., the facts. 4) Therefore, every value about something presupposes a fact. 5) All knowledge of facts is based on interpretation, and 6) all interpretation presupposes a fact. In other words, perception is full of theories, values, and interpretations that determine what humans know as facts. The critical point is that the “presupposition” in factual knowledge is given in experience and can be characterized abductively. This statement allows us to anticipate complex discovery processes and assume that there are forms that the authors call simple hypothesizing: adaptive conjectures that are not necessarily linked to a systematized intellectual activity to increase knowledge of a given topic.

Simple Hypothesizing and Adaptive Conjectures

It is possible to ground this perspective from classical approaches to Pragmatism. As noted above, Peirce characterizes abduction as the process of experimentation. For this reason, abduction is a mechanism of continuous and open-ended generation of reasonable hypotheses. In this way, facts suggest hypotheses (Peirce, 1958, CP: 7.202), which propose lines of action to direct the theoretical, evaluative, and interpretative background. Mead made profound reflections on the action from a pragmatic point of view. He assumed an ontology of a world of events, in which the present was perceived as the becoming and disappearing of an event (Mead, 1932). For Mead, the experience of events is action-based perception and manipulation. Briefly stated: “there can be no society without selves, no selves without minds, and no minds without embodied social interaction” (McVeigh, 2020). To situate the body as the central axis of biological interaction with the social environment, Mead distinguished two sensory experiences related during experimentation: distance sensations (smell, sight, hearing) and contact sensations (Lewis, 1981).

Perceived objects arise from experience-oriented purposes of use that guide action. It is interesting to highlight Mead’s theory because it encompasses the whole dimension of interaction with the social and natural environment, down to the very basis of experience: perception. The basic process of extracting information is defined as a highly hypothetical system even in temporal perception: the “present facts” are the conjugation of the past (the instant that disappears) with the present experience, which configures a projection toward a new line of action (becoming). Although there is no time to develop it in this chapter, it may be relevant to the following. In Dewey’s pragmatic approach on human nature and behavior to understand better the role of values for factual knowledge, causal perception oriented to future consequences are moral because they claim how things should be (appealing to justice, order, the good, etc.) (Dewey, 1930: 18). Thus, the theories and models with which human beings know the world depend on their workableness to achieve the hypothetically stated end (James, 1987: 826). Nevertheless, it is essential to consider all these factors that determine both the ways of knowing facts and the facts themselves as social and cultural constructs that participate in how humans experience the environment on a biological scale.

The reconstruction made here from different perspectives of classical Pragmatism allows us to apply the debate on abduction to the discussion of minimal cognition. Similarly, it is possible to link the perspective defended in this chapter to contemporary abduction theories. This is possible because both perspectives have the same pragmatic roots. Therefore, it is plausible that the same theoretical foundation found genealogically to ground minimal cognition from classical Pragmatism is also in contemporary abduction theories. Moreover, considering that the contemporary debate on abduction is grounded on the ideas of classical Pragmatism, it is also plausible to think that the same elements will be found in it, allowing us to raise relative questions about minimal cognition in classical Pragmatism. An excellent starting point for this chapter’s project is the eco-cognitive model of abduction (aka EC-Model). As presented below, it is a model that allows us to approach the question of cognition from a situated and embodied perspective that emphasizes contextualization. Thus, it allows addressing the question of hypothesizing and discovery from the interaction with the environment. Therefore, it is also an excellent theoretical point to deepen determinism to maintain an indeterminate perspective at the scale of simple hypotheses and, therefore, to implement minimal cognition.

The Eco-cognitive Perspective on Minimal Cognition

The proposal of this chapter is articulated with the EC-Model of abduction: cognition is [contextually] embodied and the interactions between brains, bodies, and external environment are its central aspects (Magnani, 2017: 207). The pragmatic viewpoint assumed in the EC-Model is a naturalistic perspective of cognitive processes such as reasoning, which allows conceptualizing the generation and selection of hypotheses as mechanisms of adaptation to varieties of experience. This model emphasizes the contextualization and anthropomorphization of reasoning. In a general sense, the main idea that defines the EC-Model is that both the agent and the environment are modified during experimentation. In this perspective, it is assumed that part of human cognition is structured to adapt to the various ways of experiencing contingency. Therefore, the varieties in perception do not allow us to conceive reasoning as a definite and structured form of inference. On the contrary, these varieties allow us to conceive reasoning as deterministically open (aka indeterminate): a cognitive system that tends to make the most of the resources at its disposal.

As has been said, the EC-Model allows us to synthesize the essential features of Classical Pragmatism from a naturalized philosophical perspective. This is crucial for this work. The main reason is that the EC-Model allows us to update Pragmatist reflections on the most basic layers of perception. In this chapter, perception is characterized as a cognitive process of information extraction that is deterministically open-ended: the active extraction of information and elaboration of representations not only tends to make the best use of the resources at its disposal but is also sensitive to adapt to the changes that may be experienced. Thus, new forms of perception may occur. For example, introducing the notion of number helps the children to discriminate, select, temporize, etc. (c.f., Pirahã language as a counterexample) and create ethical relations (consider when a child understands that they has more candy than their friend). Then, the pragmatic maxim of considering that knowledge is defined from the practical effects of objects (Peirce, 1958, CP: 5.402) can be approached from a cognitive perspective of perception. Furthermore, the EC-Model of abduction translates the iconic relation-based resemblance property of Peirce’s theory into the form of hypothetical reasoning. From this perspective, abduction can be conceived as a process of adaptation to an environment composed of a constant flow of information in which experiences proliferate.

As is known, Peirce incorporates abduction into perceptual judgments through the reasoning (syllogism) of experimentation. Although Peirce’s primary concern is structurally cognitive (the relation between abductive-perceptual process and judgment) (Tibbetts, 1975: 229), his analysis of perception allows the introduction of non-cognitive and immediate elements that participate in the subsequent abductive process. Phenomenologically, Peirce recognizes this gnoseological stage in firstness (the information given immediately in experience). The category of firstness is characterized as non-cognitive: it is an inexplicable and immediate qualitative state of feeling or sensation (ibid.: 223). Likewise, Mead goes further into how the contextualization of biological needs determines perception. This reflection is situated at a non-cognitive stage of perception: perceiving is an act that is not thought (ibid.: 227). On the contrary, immediate perception predisposes (anticipates) the way objects will be experienced and, therefore, the form of perceptual judgment.

In other words, both the agents’ perception and the factual context determine the hypothetical generation and selection of lines of action. The abductive relationship between the practical dimension and the hypothesized product’s truth (Magnani, 2017: 15) can be explained through pragmatic holism: any reasoning that predominates in each circumstance is supported by the rest of the inferential cognitive resources. On the one hand, the EC-Model emphasizes the pragmatist idea that the diversity of inquiry forms requires a reasoning effort on the agent’s part. On the other hand, it is possible to incorporate the non-cognitive elements integrated into perception (action) that trigger subsequent cognitive mechanisms. In both cases, the specificity of the context and the moment in which agents inquire are essential. However, the second case deals with how the biological dimension of the agent determines the hypothetical approach to, for example, perceive novel cases.

Naturalization of Logic as Anthropomorphization of Cognition

The EC-Model of abduction is situated within the project of naturalization of logic (Magnani, 2017: 140), in the sense of fixing attention on the cognitive dimension of the agent to define logical systems. It is possible to identify the naturalistic project of logic to translate the epistemic issues of scientific investigation into its cognitive dimension (ibid.: 1-1n). For this reason, elements such as heuristics are introduced. It is common to refer to Polya’s notion of heuristic reasoning to discuss this concept.

An example is a discovery (Polya, 1971: 113). From a cognitive point of view, heuristics can be considered a sophisticated type of inference that occurs as an auxiliary mechanism when someone cannot reach specific knowledge. It is common to characterize the process of abductive hypothesizing as to applying heuristic strategies (c.f., Magnani, 2017: 57–60; Amra et al., 1992). This perspective understands hypothesizing as relating information in the face of a novel scenario. Therefore, it is an inferential resource for problem-solving situations management.

Likewise, if attention is paid to the primary dimension of action, another aspect of heuristics emerges. It is possible to understand heuristics as a minimal cognitive strategy that manifests itself through interaction with the environment. In other words, it is an immediate and permanent resource of perception that determines the indeterminate way of experiencing the constant flow of information. The perspective offered here makes much more sense within the program of naturalized abduction offered by the EC-Model. As mentioned, one of the strengths of this theory is the contextualization of cognition: contemplating the conditions under which experimentation (perception and perceptual judgments) takes place. The EC-Model can be equated with Peirce and Mead’s process of perception because it allows a) analyzing the complex processes of judgment (abduction as reasoning), and b) analyzing the processes inherent in the act of perceiving. Demonstrating this last point implies an extension of the EC-Model.

It can be considered that the agent becomes part of the context because it is also part of the environment. After all, agents are modified by their actions in a particular context. In other words, every agent is always part of the context of others. For this reason, it can be stated that the naturalization implied by the EC-Model is based on an anthropomorphization. This is a systematic integration of human capacities within an embodied conception of cognition that contemplates α) the general and concrete characteristics of bodies and β) the possibilities and limits that arise from interacting with the environment with this (type of) body.

For this reason, the naturalization of logic also means assuming the pragmatist postulate that relates action to reasoning. This relation follows those phenomena such as attention not excluding other (heuristic) information, transformed while the agent adapts during experimentation. Indirect information is incorporated into direct one through actions. In other words, there are indirect elements involved (unconscious, automatic, reflexes, etc.), entailing variations of different types on different scales that allow agents to adapt to the contingency of experience. In this sense, the naturalization of logic also means giving more weight to the (still) non-formalizable aspects of human reasoning. Presumably, the different types of information involved are not inferred and related using a single type of reasoning or a single pattern of inference. Therefore, the pragmatic postulate that operates at the epistemic scale to accommodate variations in perception (Shanahan, 2005) could be considered to be articulated from minimal cognitive mechanisms that integrate the basic cognitive processes that relate the action of perceiving to the action of thinking about perception.

Let us take, for instance, an archaeological investigation in which it is necessary to hypothesize a plausible scenario to make sense of some remains found. Archaeologists need to use resources such as imagination, creativity, etc., to generate hypotheses to develop their investigations. In particular, the type of hypotheses that archaeologists must come up with must offer an explanation that makes sense of the objects they study and their distribution in the place where they have been discovered. In this hypothesizing process, a series of mental templates in recognition patterns let archeologists perceive and observe phenomena during the archaeological investigation (Shelley, 1996). It is important to note that Shelley’s characterization of abduction using recognition patterns is based on the process of conscious inference.

The example of archaeology allows the authors to discuss both processes regarding abduction in perception and abductive reasoning. Usually, abductive reasoning is characterized as an under-coded process (Meyer, 2010): a strategy to find a hypothesis to a fact that has surprised us. There are, on the other hand, over-coded abductions (Eco, 1983): the unconscious (Schurz, 2008: 207) and instantaneous (Magnani, 2001) process that occurs in perception. Just as under-coded abduction is triggered by a situation that classical epistemic processes cannot answer, over-coded abduction is triggered because cognitive resources manifest themselves in the form of interaction, which humans adopt in the face of suggestions offered by an object or fact. Factors such as memory, sensations, emotions, narratives, and feelings predispose how something unknown will be perceived because they will influence the final form of the interaction.

Returning to the case of archaeology, the artifact that archaeologists want information about triggers a series of under-coded abductive processes: the hypotheses are aimed at solving a defined problem of ignorance. Likewise, during the manipulation of that artifact, over-coded abductions occur. In this sense, the characteristics of the object suggest affordances that propose ways of interacting with it (Withagen & Costall, 2021; Sans Pinillos & Magnani, 2022). For example, a fissure predisposes the archaeologist to trigger the hypothesis of the object’s fragility, which predisposes the agent to act carefully. The authors call this type of action hypothetical-irreflexive. This means that human perceptions are biologically determined to maintain an open interaction with the environment. As seen, it is possible to manage this indeterminacy in perception through the EC-Model of abduction (Magnani, 2017: 15). However, it is essential to differentiate between indeterminacy and unlimitedness. While indeterminacy is an inherent biological condition in perception, unlimitedness is a circumstance subject to the context of the perceived fact. In other words, it can be said that there is no definite number of patterns for interacting reality, although a limited number of them usually apply.

Through abduction, it can be shown that the situation is quite the opposite. Using the idea that perception is based on agents’ actions, infinite combinations arise between the different elements that make up the environment. The reason is that this process involves the different cognitive dispositions of the agents interacting with their context. In other words, the way agents may consider responding, and the form of these responses are part of the world because they define the ways of perceiving and knowing. Section four shows that this active predisposition to manage contingencies of varying degrees and magnitude perceived and conceptualized may be related to creative processes.

Hybridization of Cognitive Processes: Multimodality

Therefore, abduction can be conceived as something that integrates the agents’ interactive predispositions with their environment. In this way, the context is closely related to experiencing the facts. Therefore, its possible change also comes from how agents deal with different situations. There are four requirements that abductive reasoning must meet in order to be conceptualized from the EC-Model proposal:

  1. 1.

    It must optimize the different situations we live in (optimization of situatedness).

  2. 2.

    There is a mutability (permutation) between the roles (input/output) of the elements (maximization of changeability).

  3. 3.

    Abduction is sensitive to absorbing information from what is presented to us continuously (information-sensitive).

  4. 4.

    It is necessary to enrich the inferential system we have to acquire all these requirements (Magnani, 2017: 138–139).

Therefore, a multimodal and non-monotonic system is necessary, which considers how the information is presented to us, the means that the agents have to apprehend it, and, in addition, to consider the possibly new information generated through inferences (ibid.: 139). Hybridity is a correct way of referring to this system. In a strict sense, the proposal defended in the EC-Model of abduction is a sophisticated version of Hintikka’s (2007) selective and creative abduction. Also, this model contains many of the elements with which Thagard defined abduction. For example, abduction is a component in discovering a hypothesis and an essential element for justification (Thagard, 1988: 52). In addition to selective and creative abduction, the EC-Model includes manipulative and theoretical abduction, and sentential and model-based abduction (Fig. 1).

Fig. 1
figure 1

The pattern of the EC-Model of abduction, inspired by Park (2017: ch. 2)

Theoretical abduction is dominated by an internal relationship between our knowledge and the cognitive schemas and strategies to acquire new information. Therefore, theoretical abduction is the reasoning in which creativity has a more significant presence. On the contrary, manipulative abduction is dominated by the tacit application of knowledge. However, this perspective can be broadened to include situations where creative cognitive processes determine hypothetical manipulation. For example, the relationship between emotions, feelings, and memory may suggest hypotheses about the environment that manifest themselves in the form of thoughtless behavior. Similarly, the predisposition to act in one way may determine the perception and, finally, the more complex modes of hypothesizing, such as abductive reasoning.

The Cognition of Creativity

Creativity is undoubtedly the most elusive concept affecting all human disciplines. To paraphrase St. Augustine: “What then is creativity? If no one asks me, I know what it is. If I want to explain it to anyone who asks me, I don’t know.” In many ways, explaining this phenomenon is the holy grail of current AI inquiries, especially those related to Machine Learning and Deep Learning (Nguyen et al., 2015). However, surprisingly, the study of such ability lacks systematicity, as no simple formula for reproducing or obtaining creative skills has been obtained either (Sans Pinillos & Vallverdú, 2021). The best empirical, evolutionary, and comprehensive approaches to creativity (Csikszentmihalyi, 1997) have shown that no single heuristic explains the rules followed by creative people in a wide range of human activities. At the same time, there seems to be bad news: most people are not creative in their everyday activities. Consequently, only a few individuals will create new avenues of knowledge or human practices in their lifetime. Moreover, analysis of creative agents in very different specialized fields, such as the sciences or the arts, do not show significantly different or similar action patterns to be creative. In short: creativity is a capacity that has no direct way of being achieved.

This reality forces a hypothesis: there is more than one meaning of creativity. In this chapter, two meanings are explored: 1) creativity understood as the production of theoretical or physical artifacts or ideas that imply a radical novelty for humanity (a solution to a crucial problem, an artistic work, etc.) and 2) divergent ways of managing the environment. The authors’ interest lies in the second type. Both types of creativity are primarily social and cultural phenomena. This is because human beings are essentially gregarious and live primarily in communities. In this sense, it is crucial to keep in mind that creativity arises from the interaction of agents in their sociocultural cosmovision. Therefore, the statement that few creative people are directed at the first type of creativity means that few results are perceived as creative. However, both the processes by which a person achieves a creative outcome and the appreciation of that outcome as creative by the rest of society could be placed under the second type of creativity. In other words, creativity is not usually an individual phenomenon (Feyerabend, 1987). The conceptual elements with which humans build our systems always have enough erosions to break them, combine them, etc., to use them as convenient.

Abduction has a fundamental role in eliciting creative behaviors of the second type of creativity. On the one hand, creativity is closely related to the contextual conditions that agents manage with abductive mechanisms. In this sense, those mechanisms will differ from one person to another, but in the backbone, they would remain the same biological basis: abduction as a switch activator to decide positively toward innovative possibilities. The authors assume that there is also a combination of attitudinal characteristics: stubbornness, dedication to work, focus, self-confidence, and confidence in the success of existing ideas. Again, it must be differentiated between what is considered creativity as a complete cognitive ability at the animal level (Kaufman & Kaufman, 2004) and the specific human ability to be creative. This distinction is fundamental to confront cognitive approaches to creativity with algorithmic and statistical ones. It is necessary to clarify some neural mechanisms that explain abduction, which the authors will discuss in the next section to understand how this process might be possible. It is necessary to clarify some neural mechanisms that explain abduction, which the authors will discuss in the next section to understand how this process might be possible.

Sometimes, the creative process is reduced to combining information in different ways to get out of the impasse. In those cases, the creative response is a new line of inquiry. However, it is essential to note that not every new course of action is perceived as creative. For example, the creative response is often the “icing on the cake” when faced with a defined problem. Although it does not matter whether the “icing” is placed at the beginning, during, or at the end of the process, it must usually be an element that will be decisive for an outcome that was not foreseeable. However, the possibility of introducing this crucial element will be determined by many external factors. This chapter hypothesizes that the abductive mechanisms of triggering and constraining critical information for hypothesis generation and selection are sometimes related to creative processes. For example, Łukasiewicz claimed that all types of reasoning involve a degree of creativity because they include interpretations of facts using laws, generalizations, etc. This Łukasiewicz’s holistic approach of complex hypothesizing processes is structured abductively (through a reduction) because it is fundamentally constructive: information is added to make sense of something unknown (Łukasiewicz, 1970: 7).

Likewise, it is possible to identify this relationship between abduction and creativity in perception. Both types of abduction will be creative if the puzzling fact triggers a divergent hypothesis. The authors use divergent to capture the novelty-never-seen: the novelty that at first may even be challenging to comprehend. Socially, something will be considered genuinely creative if it is original. However, any divergent hypothesis never experienced before will be experienced as genuine for an agent (Boden, 2013). This chapter is interested in the second type of creative product. It may be that a personal experience best exemplifies these cases. For example, one of the ways to interrupt an immobilization in Judo is for the one receiving the technique (uke, 受け) to lock the legs of the one applying the technique (nage, 投げ). It is expected that when learning immobilizations such as Kuzure-Kesa-Gatame (崩袈裟固) (nage wraps an arm around uke’s waist), nage develops a creative strategy so that uke does not manage to lock his legs: turning like a clockwork with the help of the legs. This “technique” is not taught but is a resource that emerges unreflexively in almost all novices in the first fights they participate in and can be understood as a creative resource.

As a second example, the authors introduce the Deep Learning system created by Google’s Deep Mind company: AlphaGo, which is considered the most complex game. In March 2016, such AI beat the best-known player Lee Sedol (Wang et al., 2016). In Game Two, Move 37, AlphaGo moved completely unexpected for any expert human in the last 3000 years of the game’s existence. Of course, such movement changes how humans approach it, but it did not take years of analysis: just two Games later, in Game Four, Sedol made an unexpected move again, “Move 78,” also called “God’s touch.” Despite losing the match, the lessons were formidable: again, machines dominated humans even in abstract games, but at the same time, the human player was able to react instinctively to the pressure and abductively chose a new strategy that gave him some advantage. Furthermore, too, Deep Learning started to demonstrate its overwhelming power.

Cognitive Triggers: Surprise, Ignorance, and Interest

Triggers and constraints articulate the abductive processes. Both are logical and cognitive mechanisms that represent the operational bivalence of morphological and cultural aspects in defining the margins of our reality (constraints) and in the process of change (triggers). Introducing these mechanisms allows us to address different degrees of the materiality of the interactions that humans perceive and with which they conceptualize reality. Reference is made to agents’ interaction with themselves, the interaction between other agents, artifacts, and technological devices.

As already stated, the relationship between triggers and creativity can be traced back to Csikszentmihalyi’s (1997) inquiries in psychology. In a general way, triggers can be conceptualized as the set of mechanisms that operate bivalently with the constraints, whose function is to break or expand the margins with which humans configure reality. Cognitive triggers are those emotions and sensations that generate different strategies, including abduction. They are complicated to conceptualize, mainly because there are no developed representation tools, but also because they are concepts that have more than one meaning.

A classic example is a surprise (Peirce, 1958, CP: 5.188–189), which can occur in many ways, intensities, contexts, etc. One way to interpret surprise is in terms of an event that violates a preexisting belief (Gabbay & Woods, 2005: 82). In this sense, surprise initiates a highly original and creative doxastic process, the purpose of which is to begin to devise something new. Examples could be when a judo player realizes how simple it can be for his legs to be blocked and when in an experiment, incomprehensible data emerges from the theoretical framework on which the inquiry is based. The attractiveness of this proposal is that this process is understood in a gradual sense as follows:

  1. 1.

    Showing for the first time that some element, however vaguely characterized, is an element and must be recognized as distinct from others.

  2. 2.

    To show that this or that element is not needed.

  3. 3.

    Giving distinctness – workable, pragmatic, distinctness – to concepts already recognized.

  4. 4.

    Illuminative and original criticism of the works of others (Gabbay & Woods, 2005: 82).

Another feature of this scheme is that it assumes Anderson’s distinction (from Peirce’s one) between reordering and concept creation (Anderson, 1987: 47). While reordering deals with the structure of experience, concept creation is concerned with generating new ideas and with the concept of novelty itself. To assert that only a surprise can interrupt the ordinary course of reasoning implies accepting that humans are not usually surprised. There is every reason to reject this idea. For example, the act-reflection processes (unconscious intentional reactions) introduced in this paper can be characterized as abductive processes embedded in perception to provide an adaptive mechanism capable of overcoming the surprise of experiencing new information. In other words, just as people do not always ignore in the same way, there are different ways of being surprised, for instance, at the perceptual level. For example, the surprise of noticing something cold when it “should” be hot (a pot that was heating in the oven) is not of the same intensity as the surprise that it does not fall when it “should” fall (e.g., the step that has broken has a surface immediately below it). Another very creative case is the surprise of being sure that someone has survived something that “should” have been fatal.

From this point of view, it is possible to consider ignorance as a second-order trigger: the experimentation and possibility of obtaining empirical knowledge are mediated by a degree of ignorance (Sans Pinillos & Magnani, 2022). Likewise, ignorance of something is one of the possible triggers of surprise. Then, ignorance generates a genuine situation that provides inquiry (action) (Arfini, 2019). From the EC-Model of abduction perspective, ignorance is produced by cognitive interaction with the environment. Another critical but underworked trigger is the interest. There is a crucial relationship between interest and ignorance:

  1. 1.

    Having a particular interest

  2. 2.

    Understanding the meaning of an impending phenomenon as a chance

  3. 3.

    Putting a scenario based on a selected chance into a concrete shape

  4. 4.

    Running a simulation or taking action based on the scenario

  5. 5.

    Acquiring a new interest (Maeno & Ohsawa, 2007)

It is possible to deal with the trigger of interest in the discursive context. The question of how the particularity of the method is configured through the shared observations that generate the appreciative discursive basis can be addressed by abduction to apprehend sensibility. From the Habermasian perspective, a distinction can be made between description (Beschreibung) and narration (Erzählung). While the former refers to observed objects, the latter understands the observed objects under a concrete discourse (aka method) (Habermas, 1992: 395). This distinction allows Habermas to formulate the possible experience. Likewise, the discourse of possible facts is constructed based on the distributive property that this form of experience acquires when situated in a discursive framework (ibid.: 396). This schematization of sensibility is grounded in action: all descriptions are based on the action of knowing the object of inquiry. Therefore, description and the action of describing define the subsequent narration (understanding of the observed objects under a concrete discourse) (ibidem).

As it has been seen, everything said so far is part of the discursive realm, where facts are transformed into parts of it in order to separate themselves from the world and, in turn, maintain a specific connection through their use. In other words, with use (action) within a narrative, descriptions are thematized to give meaning to something that may vary, depending on the context through which it is viewed. The connection is maintained by the logic of shared discourse and also by pre-scientific logic. Finally, this makes sense in the theory of interest, which motivates the desire to know something. The point of common discursive understanding ensures that interests can be shared. This point is critical for how extra-theoretical information is abductively used to advance our work method. The first point of understanding is crucial for recognizing what the other is inquiring into and then learning it as a regulative guide to our actions.

Cognitive Constraints

Constraints are to be understood as the bivalent counterpart of triggers. These elements are postulated to show that our capacity to generate hypotheses is controlled by different morphological and cultural mechanisms that determine our reality. The term has been used in the abduction debate, especially from logical and computational theories, where these mechanisms are understood as protocols to avoid the massive proliferation of generated information. Cognitive constraints structurally operate similarly to logical ones but are enriched by many factors. This circumstance is so because, on the one hand, the logical/algorithmic reduction is based on an attempt to capture our psychological life and, on the other hand, because, as already seen, computation has very dense material limits. Indeed, some of these constraints, as has also been seen with triggers, are inherently human and, in this sense, are as rigid as the computational ones, with the difference that humans are hardwired to adapt. Ergo most of these strategies are also hardwired to change. Nevertheless, on the other hand, there are also cultural and social strategies where change should also be contemplated in the use of humor, trivialization, power hierarchies, advertising, or the economic factor, among others.

It is important to note that, although everything can be generally understood as a trigger from an affordance point of view (Estany & Martínez, 2013), this does not reflect the kind of disposition that an agent must have to consider it as such. On the contrary, from the constraint perspective, it is evident that our disposition often has little or nothing to do. Examples are circumstantial conditionals that allow us to be aware of some anomaly or unexpected event (Roberts, 1989) (such as the weather). The tools that agents use must also be considered. Languages are tools that open a field of possibility (Magnani et al., 2021), but they are also the limits of what can be said and thought. Moreover, devices are designed precisely to change the world through their operation. Language is composed of an infinite number of resources that allow us to modify its tone and offer poetic, sarcastic, metaphorical visions, etc., which have served to understand the world in a certain way. An example that authors love is Pedersen’s analysis of the Norse sagas, which shaped a culture’s cosmovision (Pedersen, 2013). The way things are related opens the doors to new interpretations and, more interestingly, when mixed in different domains, superb results can be given. Successful theories are known that have started with a visual metaphor or using an idea taken from a dream, an extrasensory experience, or using religious beliefs (Boden, 2004: ch. 1). However, at the same time, these possibilities are also an indeterminate limit. In other words, it conditions the way reality is understood and our role in it (Huang & Jaszczolt, 2018). To the extent that the possibility of signification is determined, the eco-cognitive environment is also configured. In turn, the environment that the authors refer to as the socially shared cosmovision is traversed and shaped by culture.

The Neuro-cognitive Basis of Abduction and Related Bioinspired Computation

From a naturalistic and evolutionary perspective, abductive processes should be elicited by some morphological mechanism. All humans perform innately abductive processes, and, therefore, some embodied system must be the cause of such a process. Neurochemical processes are the best explanation of such a mechanism (Thagard, 2007; Seddon, 2021), at least for humans (it is assumed the necessity of exploring abductive processes in non-neural cognitive organisms, but it is beyond our scope in this chapter to justify or give solid support to it). An excellent way to check the neuroanatomical basis of abductive processes is to check their malfunction due to some psychological disorders (Coltheart et al., 2010) and the study of single isolated pieces of such mechanisms, like the dopaminergic mechanisms (Dasgupta et al., 2018). Cognition can then be understood as a heuristical process that uses several sources of data and integration strategies as a blending process (Vallverdú, 2019). It is possible to find connections between abduction and other cognitive processes for those reasons (Calzavarini & Cevolani, 2022). On the other hand, although there are ways to consider the naturalistic connections between abductive reasoning and Bayesian statistics (Vallverdú, 2016), this aspect will also be neglected in this paper because it is out of the limits of our current research. Nevertheless, it is an interesting debate that can help understand the different perspectives about information analysis (Psillos, 2004).

Despite the sound arguments in favor of the existence of abductive practices in animal cognition (Vitti-Rodrigues & Emmeche, 2017), more studies elucidating in detail these mechanisms and their variabilities (as a taxonomical approach to abductive cognition) are needed. In any case, there are enough empirical shreds of evidence that support the hypothesis of this paper: embodied cognition is morphologically mediated, and this functional framework determines that agents are inevitably forced to use abductive reasoning in our day-to-day activities. Consequently, abduction must have a fundamental role in creative processes (Nubiola, 2005).

Mirroring (Artificial) Abduction: Machine Learning

Because of the fundamental role of creativity in knowledge acquisition, this process is being tried to capture and implement into AI systems from an evolutionary and cultural point of view. Machine Learning and Deep Learning fields (henceforth, Ml and Dl, respectively) are working intensively on achieving artificial creativity using algorithmic and statistical processes. The extraordinary recent successes of Deep Learning methods applied by Google’s DeepMind company has provided new insights (in fact, a revolution) into classic games like Chess or Go, as well as on very complex scientific problems, like the protein folding structures, with their AI systems AlphaGo, AlphaZero, and AlphaFold (Callaway, 2020). The funny side of this revolution is that it started with Demis Hassabis’ (founder of DeepMind) omnivorous curiosity and his interests in neurosciences and videogames. So again, the paths of creativity are beyond the expected roads of sound and official thinking.

There is a fundamental aspect for analyzing abductive models in Machine Learning: considering the internal trends and debates inside AI communities. Classic GOFAI created an AI-based on logical rules, while the Second Wave AI experts tried to create a system from a bottom-up view, introducing morphologies into the intelligence equation. In any case, the successes came from the creation of bioinspired techniques, the Neural Networks, which matured and increased in complexity in the machine learning subfield, being the same statistical techniques necessary for processing the new huge sets of Big Data, the called Data-Tsunami, at the beginnings of the twenty-first century. Deep Learning was so successfully applied to solve a broad range of socio-technical and scientific problems that it has even been affirmed in 2020 by one of its creators, Geoff Hinton (Hao, 2020): “Deep learning is going to be able to do everything.” Nevertheless, the real thing is that DL is entering into an epistemic bottleneck despite some wonderful achievements, for several reasons: opacity due to black-boxes, general lack of explainability, and, among all the others, the lack of causal understanding (Schölkopf et al., 2021; Vallverdú, 2020). For these reasons, current studies are trying to introduce meaning into statistical approaches to AI systems, that is, combing symbolic with statistical knowledge in Machine Learning (Liu et al., 2019). Even non-conventional logics is fundamental for non-conventional computing (Schumann & Zenil, 2020).

Anyhow, it will be soon noticed that the grounded basis of symbolic emergence, a combination of embodied plus social aspects of cognition, is still missing. So again, the bioinspired mechanisms of abductive reasoning are a fundamental key for the phylogenetical understanding of cognition, which must include abduction. On the other hand, an essential question concerning our research is: AI researchers have tried to implement abduction in ML or DL fields. But, why? The reasons are because they aim at improving AI’s creative and innovative properties (if any). After some early attempts (Shanahan, 1989; Marquis, 1991), the conceptualization of abductive reasoning done by computer scientists working on Machine Learning has appeared to be soon-systematic. Let us see some examples of it:

  1. I.

    Bergadano et al. (2000) defined abduction as a form of defeasible reasoning, usually implemented into Machine Learning as (a) an aside technology and used as reasoning in explanation-based learning systems (as a heuristic to guide search in top-down specialization), or as (b) a way to generate missing examples in relational learning. The authors considered that abduction was not implemented as a general reasoning way but, in contrast, as a way to solve every tiny and specific problem.

  2. II.

    Ignatiev et al. (2019) bring an example and a recent analysis of abduction in ML. They used abductive reasoning to allow a constraint-agnostic solution for computing explanations for any ML model. With this tool, they tried to exploit the best properties of logic-based and heuristic-based approaches because they represented the ML model as a set of constraints in some theory (e.g., a decidable theory of first-order logic).

  3. III.

    Mooney and Shavlik (2021) applied abduction to the “Theory refinement” (theory revision, knowledge-based refinement) as the Machine Learning task of modifying an existing imperfect domain theory. Then, it can be made consistent with a set of data from an accurate abduction definition as “the process of inferring cause from effect or constructing explanations for observed events and is central to tasks such as diagnosis and plan recognition.”

  4. IV.

    (Dasgupta et al., 2018): “Abduction refers to inferring the premises (causes) from the observations (effects) of any rule of the form: if-then. Although the logic of propositions/predicates does not support abduction, it has importance in many real-world situations. The logic of fuzzy sets offers a solution to abductive reasoning problems. Several techniques of abductive reasoning are available in the literature. For example, for a given rule: if x is A, then y is B, where x and y are linguistic variables and A and B are fuzzy sets, given y is B/, where B/ is an observed MF, it can be inferred A/ by computing the implication relation R(x, y), and then by computing the MF for x is A/ by using max-min composition (o) of B/ oR-1(y, x), where R-1(y, x) is the inverse fuzzy implication relation, such that R-1o R = I, the identity matrix.” Perhaps it is too reductionist and misleading to reduce abduction to any if-then computational rule description.

Other authors have explored the connections of introducing abduction in neural networks (Ray & d’Avila Garcez, 2006) or modeling ML with abductive reasoning from an early wave at the end of the twentieth century (O’Rorke, 1988; Hirata, 1993), to the second one at the beginning of 21st one (Chakraborty et al., 2009; Kakas & Michael, 2020; Huang et al., 2021). In some cases, adjustments on the classic view of abduction have also been used in these technical fields, like “weighted abduction” (Appelt & Pollack, 1992).

Among plenty of explorations using abduction, the authors want to highlight Vladimir Vapnik and his colleagues’ research. They attempt to introduce statistical tools to describe abductive practices to improve ML methods. For example, in Vapnik and Izmailov (2019), the authors introduce a new type of inference based on statistical invariants (see formula 13). Since it is valid for any predicate (any function $\psi(x)\in L_2$), one can construct as many statistical invariants as one wants (by defining properties of class). In philosophy, sometimes it is called “The Duck Test$” and refers to abductive inference. Vapnik suggests that learning using statistical invariants is the most effective way of learning (Vapnik & Izmailov, 2015).

From these examples, the authors can defend the interest and necessity of introducing abductive mechanisms in artificial reasoning, something especially requested in the case of causal connections between events. At the same time, the lack of universal formalization and amateurism in these topics held by computer scientists has created a fuzzy implementation and understanding of such fundamental cognitive mechanisms.

Conclusions

This chapter has given arguments favoring morphologically situated abductive processes in perception. This perspective extends the naturalistic proposal defended in the EC-Model of abduction. As has been shown, naturalized abduction is based on a contextualized view of the principles of Classical Pragmatism. On the other hand, the pragmatic viewpoint assumed in the EC-Model is a naturalistic perspective of cognitive processes such as reasoning, which allows conceptualizing the generation and selection of hypotheses as mechanisms of adaptation to varieties of experience. The mechanism that has been analyzed in this essay is creativity. Characterizing creative processes using abduction means analyzing this phenomenon from morphological bases. In doing so, the authors have tried to defend the thesis that human beings are determined to maintain a constant openness of hypothesizing in the face of the contingencies experienced. Although this has been raised in this essay only as conjecture, this thesis could gradually be applied to the rest of living beings.

For this reason, the inevitability of creativity defended in this chapter goes beyond a methodological attitude or a theory. Because of the morphological bases of perception (neurochemical, in human beings) abduction appears as a key mechanism for the adaptive dimension of creativity. As it has been argued, this dimension is present in all degrees of human experience, always a fundamental part of the generation of epistemic content in the socially shared cosmovision using natural processes of biological basis. Bodies naturally evolved accompanied by a social pressure to be creative: humans must be creative to solve plenty of challenges throughout their life. Abduction is the ground mechanism by which (naturalized and socialized) cognition adapts to evolving scenarios and therefore is at the bottom of all cognitive procedures: from intuition to high symbolic data processing. Causal knowledge is at the horizon of such artificial abductive systems, finding the same meaning explanations that humans have searched throughout history. The excellent news for cognition designers is that the fundamental mechanism which allows the ladder of other cognitive mechanisms is identified and explained here: abduction.

References

  • Aliseda, A. (2006). Abductive reasoning: Logical investigations into discovery and explanation. Springer.

    MATH  Google Scholar 

  • Amra, N. K., Smith, J. W., Jr., Johnson, K. A., & Johnson, T. R. (1992). An approach to evaluating heuristics in abduction: A case study using RedSoar – An abductive system for red blood cell antibody identification. In Proceedings of the annual symposium on computer application in medical care (p. 690). American Medical Informatics Association.

    Google Scholar 

  • Anderson, D. R. (1987). Creativity and the philosophy of C. S. Peirce. Springer-Science+Business Media, B. V.

    CrossRef  Google Scholar 

  • Appelt, D. E., & Pollack, M. E. (1992). Weighted abduction for plan ascription. User Modeling and User-Adapted Interaction, 2(1), 1–25.

    CrossRef  Google Scholar 

  • Arfini, S. (2019). Ignorant cognition. A philosophical investigation of the cognitive features of non-knowing. Springer.

    CrossRef  Google Scholar 

  • Aristotle. (1957). Analytica Priora et Posteriora. In W. D. Ross (ed.). Oxford University Press.

    Google Scholar 

  • Bergadano, F., Cutello, V., & Gunetti, D. (2000). Abduction in machine learning. In Abductive reasoning and learning (pp. 197–229). Springer.

    MATH  CrossRef  Google Scholar 

  • Boden, M. (2004). The creative mind: Myths and mechanisms. Routledge.

    CrossRef  Google Scholar 

  • Boden, M. A. (2013). Creativity as a neuroscientific mystery. In O. Vartanian, A. Bristol, & J. C. Kaufman (Eds.), The neuroscience of creativity (pp. 3–18). MIT Press.

    CrossRef  Google Scholar 

  • Breazeal, C., Dautenhahn, K., & Kanda, T. (2016). Social robotics. In: Siciliano, B., Khatib, O. (eds) Springer Handbook of Robotics. Springer Handbooks. Springer, Cham, 1935–1972.

    Google Scholar 

  • Callaway, E. (2020). ‘It will change everything’: DeepMind’s AI makes gigantic leap in solving protein structures. Nature, 588(7837), 203–205.

    CrossRef  Google Scholar 

  • Calvo Garzón, F. (2007). The quest for cognition in plant neurobiology. Plant Signaling & Behavior, 2(4), 208–211.

    CrossRef  Google Scholar 

  • Calzavarini, F., & Cevolani, G. (2022). Abductive reasoning in cognitive neuroscience: weak and strong reverse inference. Synthese 200, 70.

    Google Scholar 

  • Chakraborty, S., Konar, A., & Jain, L. C. (2009). An efficient algorithm to computing max–min inverse fuzzy relation for abductive reasoning. IEEE Transactions on Systems, Man, and Cybernetics-Part A: Systems and Humans, 40(1), 158–169.

    CrossRef  Google Scholar 

  • Coltheart, M., Menzies, P., & Sutton, J. (2010). Abductive inference and delusional belief. Cognitive Neuropsychiatry, 15(1-3), 261–287.

    CrossRef  Google Scholar 

  • Csikszentmihalyi, M. (1997). Flow and the psychology of discovery and invention (p. 39). HarperPerennial.

    Google Scholar 

  • Damasio, A. (1994). Descartes’ error: Emotion, reason, and the human brain. Putnam.

    Google Scholar 

  • Dasgupta, M., Konar, A., & Nagar, A. K. (2018, November). Online prediction of dopamine concentration using EEG-induced type-2 fuzzy abduction. In 2018 IEEE symposium series on computational intelligence (SSCI) (pp. 212–218). IEEE.

    Google Scholar 

  • Dewey, J. (1930). Human nature and conduct. An introduction to social psychology. The Modern Library.

    Google Scholar 

  • Dingemanse, M., Perlman, M., & Perniss, P. (2020). Construals of iconicity: Experimental approaches to form–meaning resemblances in language. Language and Cognition, 12(1), 1–14. https://doi.org/10.1017/langcog.2019.48

    CrossRef  Google Scholar 

  • Eco, U. (1983). Horns, hooves, insteps. In U. Eco, U. Eco, & T. A. Sebeok (Eds.), The sign of the three. Dupin, Holmes, Peirce (pp. 198–220). Indiana UP.

    Google Scholar 

  • Estany, A., & Martínez, S. (2013). “Scaffolding” and “affordance” as integrative concepts in the cognitive sciences. A Philosophy of Psychology, 27(1), 98–111.

    CrossRef  Google Scholar 

  • Feyerabend, P. (1987). Creativity: A dangerous myth. Critical Inquiry, 13(4), 700–711. http://www.jstor.org/stable/1343525

    CrossRef  Google Scholar 

  • Gabbay, M., & Wood, J. (2005). A practical logic of cognitive systems: The reach of abduction, insight and trial (Vol. 2). Elsevier.

    Google Scholar 

  • Habermas, J. (1992). Erkenntnis und Interesse. Suhrkamp.

    Google Scholar 

  • Hanson, N. R. (1972). Patterns of discovery. Cambridge University Press.

    Google Scholar 

  • Hao, K. (2020, November 3). AI pioneer Geoff Hinton: “Deep learning is going to be able to do everything”. MIT Technology Review.

    Google Scholar 

  • Hintikka, J. (2007). Socratic Espistemology. Explorations of Nowledge-seeking by ques-tioning. Cambridge University Press.

    CrossRef  Google Scholar 

  • Hintikka, J., & Remes, U. (1974). The method of analysis: Its geometrical origin and its general. Springer-Science+Business Media, B. V.

    MATH  CrossRef  Google Scholar 

  • Hirata, K. (1993). A classification of abduction: Abduction for logic programming. Machine Intelligence, 14, 405.

    Google Scholar 

  • Huang, M., & Jaszczolt, K. M. (2018). Expressing the self: Cultural diversity and cognitive universals. Oxford University Press.

    CrossRef  Google Scholar 

  • Huang, Y. X., Dai, W. Z., Cai, L. W., Muggleton, S., & Jiang, Y. (2021). Fast abductive learning by similarity-based consistency optimization. Advances in Neural Information Processing , 26574–26584.

    Google Scholar 

  • Ignatiev, A., Narodytska, N., & Marques-Silva, J. (2019, July). Abduction-based explanations for machine learning models. In Proceedings of the AAAI conference on artificial intelligence (Vol. 33, No. 01, pp. 1511–1519).

    Google Scholar 

  • James, W. (1987). The meaning of truth. William James. Writings 1902–1919. The varieties of religious experience; pragmatism; a pluralistic universe; the meaning of truth; some problems of philosophy; essays (pp. 821–978). Literary Classics of the United States, Inc..

    Google Scholar 

  • Kakas, A., & Michael, L. (2020). Abduction and argumentation for explainable machine learning: A position survey. arXiv preprint arXiv:2010.12896.

    Google Scholar 

  • Kapitan, T. (1997). Peirce and the structure of abductive inference. In N. Houser, D. D. Roberts, & J. van Evra (Eds.), Studies in the logic of Charles Sanders Peirce (pp. 477–496). Indiana University Press.

    Google Scholar 

  • Kaufman, J. C., & Kaufman, A. B. (2004). Applying a creativity framework to animal cognition. New Ideas in Psychology, 22(2), 143–155.

    CrossRef  Google Scholar 

  • Koch, C. (2019). The feeling of life itself: Why consciousness is widespread but can’t be computed. MIT Press.

    CrossRef  Google Scholar 

  • Kurzweil, R. (2000). The age of spiritual machines: When computers exceed human intelligence. Penguin.

    Google Scholar 

  • Lenat, D. D., & Brown, J. S. (1984). Why am and eurisko appear to work. Artificial Intelligence, 23(3), 269–294. https://doi.org/10.1016/0004-3702(84)90016-X

    CrossRef  Google Scholar 

  • Lewis, J. D. (1981). G.H. Mead’s contact theory of reality: The Manipulatory phase of the act in the constitution of mundane, scientific, aesthetic, and evaluative objects. Symbolic Interaction, 4, 129–141. https://doi.org/10.1525/si.1981.4.2.129

    CrossRef  Google Scholar 

  • Liu, J., Patwary, M. J., Sun, X., & Tao, K. (2019). An experimental study on symbolic extreme learning machine. International Journal of Machine Learning and Cybernetics, 10(4), 787–797.

    CrossRef  Google Scholar 

  • Łukasiewicz, J. (1970). Creative elements in science. Selected works (pp. 1–15). North-Holland Publishing Company.

    Google Scholar 

  • Lyon, L. (2017). Dead salmon and voodoo correlations: Should we be skeptical about functional MRI? Brain, 140(8), e53–e53.

    CrossRef  Google Scholar 

  • Maeno, Y., & Ohsawa, Y. (2007). Human-computer inter-active annealing for discove-ring invisible dark events. IEEE Transactions on Industrial Electronics, 54(2), 1184–1192.

    CrossRef  Google Scholar 

  • Magnani, L. (2001). Abduction, reason, and science: Processes of discovery and explanation. Kluwer.

    CrossRef  Google Scholar 

  • Magnani, L. (2017). The abductive structure of scientific creativity. Springer.

    CrossRef  Google Scholar 

  • Magnani, L., Sans Pinillos, A., & Arfini, S. (2021). Language: The “ultimate artifact” to build, develop, and update worldviews. Topoi. https://doi.org/10.1007/s11245-021-09742-5

  • Marquis, P. (1991, September). Extending abduction from propositional to first-order logic. In International workshop on fundamentals of artificial intelligence research (pp. 141–155). Springer.

    Google Scholar 

  • McVeigh, R. (2020). The body in mind: Mead’s embodied cognition. Symbolic Interaction, 43, 493–513. https://doi.org/10.1002/symb.476

    CrossRef  Google Scholar 

  • Mead, G. H. (1932). The philosophy of the present. The Open Court Company.

    Google Scholar 

  • Meyer, M. (2010). Abduction – A logical view for investigating and initiating processes of discovering mathematical coherences. Educational Studies in Mathematics, 74, 185–205. https://doi.org/10.1007/s10649-010-9233-x

    CrossRef  Google Scholar 

  • Mooney, R. J., & Shavlik, J. W. (2021). A recap of early work on theory and knowledge refinement. In A. Martin, K. Hinkelmann, H.-G. Fill, A. Gerber, D. Lenat, R. Stolle, & F. van Harmelen (Eds.), Proceedings of the AAAI 2021 spring symposium on combining Machine Learning and Knowledge Engineering (AAAI-MAKE 2021). Stanford University, March 22–24, 2021.

    Google Scholar 

  • Müller, V. C., & Hoffmann, M. (2017). What is morphological computation? On how the body contributes to cognition and control. Artificial Life, 23(1), 1–24. https://doi.org/10.1162/ARTL_a_00219

    CrossRef  Google Scholar 

  • Newen, A., De Bruin, L., & Gallagher, S. (Eds.). (2018). The Oxford handbook of 4E cognition. Oxford University Press.

    Google Scholar 

  • Nguyen, A. M., Yosinski, J., & Clune, J. (2015, July). Innovation engines: Automated creativity and improved stochastic optimization via deep learning. In Proceedings of the 2015 annual conference on genetic and evolutionary computation, Association for Computing Machinery, (pp. 959–966).

    Google Scholar 

  • Niiniluoto, I. (2014). Representation and truthlikeness. Foundations of Science, 19(4), 375–379.

    CrossRef  Google Scholar 

  • Nisbett, R. (2004). The geography of thought: How Asians and westerners think differently... And why. Simon and Schuster.

    Google Scholar 

  • Nubiola, J. (2005). Abduction or the logic of surprise. Semiotica, 2005(153), 117–130.

    Google Scholar 

  • O’Rorke, P. (1988, March). Automated abduction and machine learning. In Proceedings of AAAI symposium on explanation-based learning. USA, Stanford, AAAI PRESS, (pp. 170–174).

    Google Scholar 

  • Park, W. (2017). On classifying abduction. In Abduction in context. Studies in applied philosophy, epistemology and rational ethics (Vol. 32). Springer. https://doi.org/10.1007/978-3-319-48956-8_2

    CrossRef  Google Scholar 

  • Pedersen, A. (2013). The last conference. A pragmatist Saga. Akademika Publishing.

    Google Scholar 

  • Peirce, C. S. (1958). In C. Hartshorne & P. Weiss (Eds.), Collected papers of Charles Sanders Peirce (Vol. 1–6). Cambridge: Harvard University Press, 1931–1935; (Vol. 7–8) (A. W. Burks, Ed.). Harvard University Press.

    Google Scholar 

  • Pfeifer, R., & Bongard, J. (2006). How the body shapes the way we think: A new view of intelligence. MIT Press.

    CrossRef  Google Scholar 

  • Picard, R. W. (2000). Affective computing. MIT Press.

    CrossRef  Google Scholar 

  • Polya, G. (1971). How to solve it; a new aspect of mathematical method. Princeton University Press.

    MATH  Google Scholar 

  • Psillos, S. (2004). Inference to the best explanation and Bayesianism. In Induction and deduction in the sciences (pp. 83–91). Springer.

    CrossRef  Google Scholar 

  • Putnam, H. (2001). The collapse of the fact/value dichotomy. Harvard University Press.

    Google Scholar 

  • Putnam, H. (2006). El pragmatismo. Un debate abierto, Editorial Gedisa Sevilla.

    Google Scholar 

  • Ray, O., & d’Avila Garcez, A. S. (2006, August). Towards the integration of abduction and induction in artificial neural networks. In Proceedings of 2nd international workshop on neuralsymbolic learning and reasoning, Riva del Garda.

    Google Scholar 

  • Roberts, R. M. (1989). Serendipity: Accidental discoveries in science. Serendipity. Wiley.

    Google Scholar 

  • Sans Pinillos, A. (2017). El lado epistemológico de las abducciones: La creatividad en las verdades-proyectadas. Revista Iberoamericana De Argumentación, 15, 77–91. Recuperado a partir de https://revistas.uam.es/ria/article/view/8573

    Google Scholar 

  • Sans Pinillos, A. (2021). Neglected pragmatism: Discussing abduction to dissolute classical dichotomies. Foundations of Science. https://doi.org/10.1007/s10699-021-09817-x

  • Sans Pinillos, A., & Magnani, L. (2022). How do we think about the unknown? The self-awareness of ignorance as a tool for managing the anguish of not knowing. In S. Arfini & L. Magnani (Eds.), Embodied, extended, ignorant minds: New studies on the nature of not-knowing. Elsevier. Forthcoming.

    Google Scholar 

  • Sans Pinillos, A., & Vallverdú, J. (2021). What the #®¥§≠$@ is creativity? Debats. Revista De Cultura, Poder I Societat, 135–147. https://doi.org/10.28939/iam.debats-en.2021-9

  • Schölkopf, B., Locatello, F., Bauer, S., Ke, N. R., Kalchbrenner, N., Goyal, A., & Bengio, Y. (2021). Toward causal representation learning. Proceedings of the IEEE, 109(5), 612–634.

    CrossRef  Google Scholar 

  • Schumann, A., & Zenil, H. (2020). Non-classical logics in unconventional computing. International Journal of Unconventional Computing, 15(4), 237–244.

    Google Scholar 

  • Schurz, G. (2008). Patterns of abduction. Synthese, 164, 201–234. https://doi.org/10.1007/s11229-007-9223-4

    MathSciNet  CrossRef  MATH  Google Scholar 

  • Sebeok, T. A., & Umiker-Sebeok, J. (1983). You know my method. In U. Eco & T. A. Sebeok (Eds.), Dupin, Holmes, Peirce. The sign of three (pp. 11–54). Indiana University Press.

    Google Scholar 

  • Seddon, P. B. (2021). Nature chose abduction: Support from brain research for Lipton’s Theory of inference to the best explanation. Foundations of Science, 1–17.

    Google Scholar 

  • Shanahan, M. (1989, August). Prediction is deduction but explanation is abduction. In IJCAI (Vol. 89, pp. 1055–1060).

    Google Scholar 

  • Shanahan, M. (2005). Perception as abduction: Turing sensor data into meaningful re-presentation. A Cognitive Science, 29, 103–134.

    CrossRef  Google Scholar 

  • Shelley, C. (1996). Abductive reasoning in archaeology. A Philosophy of Science, 63(2), 278–301.

    CrossRef  Google Scholar 

  • Simon, H. A. (1985). Psychology of scientific discovery. Paper presented at the 93rd Annual APA Meeting, Los Angeles, CA.

    Google Scholar 

  • Simon, H. A., Valdés-Pérez, R. E., & Sleeman, D. H. (1997). Scientific discovery and simplicity of method. Artificial Intelligence, 91, 177–181.

    CrossRef  Google Scholar 

  • Thagard, P. (1988). Computational philosophy of science. MIT Press.

    CrossRef  Google Scholar 

  • Thagard, P. (2007). Abductive inference: From philosophical analysis to neural mechanisms. Inductive reasoning: Experimental, developmental, and computational approaches, 226–247.

    Google Scholar 

  • Tibbetts, P. (1975). Peirce and Mead on perceptual immediacy and human action. Philosophy and Phenomenological Research, 36(2), 222–232. https://doi.org/10.2307/2107055

    CrossRef  Google Scholar 

  • Vallverdú, J. (2016). Bayesians versus frequentists: A philosophical debate on statistical reasoning. Springer.

    MATH  CrossRef  Google Scholar 

  • Vallverdú, J. (2019). Blended cognition: The robotic challenge. In Blended cognition (pp. 3–21). Springer.

    CrossRef  Google Scholar 

  • Vallverdú, J. (2020). Approximate and situated causality in deep learning. Philosophies, 5(1), 2.

    CrossRef  Google Scholar 

  • Vallverdú, J., Castro, O., Mayne, R., Talanov, M., Levin, M., Baluška, F., & Adamatzky, A. (2018). Slime mould: The fundamental mechanisms of biological cognition. Biosystems, 165, 57–70.

    CrossRef  Google Scholar 

  • Vapnik, V., & Izmailov, R. (2015). Learning using privileged information: Similarity control and knowledge transfer. Journal of Machine Learning Research, 16(1), 2023–2049.

    MathSciNet  MATH  Google Scholar 

  • Vapnik, V., & Izmailov, R. (2019). Rethinking statistical learning theory: Learning using statistical invariants. Machine Learning, 108(3), 381–423.

    MathSciNet  MATH  CrossRef  Google Scholar 

  • Vitti-Rodrigues, M., & Emmeche, C. (2017). Abduction: Can non-human animals make discoveries? Biosemiotics, 10(2), 295–313.

    CrossRef  Google Scholar 

  • Wang, F. Y., Zhang, J. J., Zheng, X., Wang, X., Yuan, Y., Dai, X., ... Yang, L. (2016). Where does AlphaGo go: From church-turing thesis to AlphaGo thesis and beyond. IEEE/CAA Journal of Automatica Sinica, 3(2), 113–120.

    Google Scholar 

  • Willard, A. K., Turpin, H., & Baimel, A. (2022). Maximally Intuitive, Minimally Evidenced: Universal cognitive biases as the basis for supernatural beliefs. Retrieved from psyarxiv.com/aubem.

    Google Scholar 

  • Withagen, R., & Costall, A. (2021). What does the concept of affordances afford? Adaptive Behavior. https://doi.org/10.1177/1059712320982683

Download references

Acknowledgments

Research for this article was supported by the “Innovacion epistemológica: el caso de las ciencias biomédicas” (FFI2017-85711-P), and ICREA Acadèmia Grant, and the PRIN 2017 Research 20173YP4N3-MIUR, Ministry of University and Research, Rome, Italy.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Jordi Vallverdú .

Editor information

Editors and Affiliations

Section Editor information

Rights and permissions

Reprints and Permissions

Copyright information

© 2022 Springer Nature Switzerland AG

About this entry

Verify currency and authenticity via CrossMark

Cite this entry

Vallverdú, J., Sans Pinillos, A. (2022). The Foundations of Creativity: Human Inquiry Explained Through the Neuro-Multimodality of Abduction. In: Magnani, L. (eds) Handbook of Abductive Cognition. Springer, Cham. https://doi.org/10.1007/978-3-030-68436-5_71-1

Download citation

  • DOI: https://doi.org/10.1007/978-3-030-68436-5_71-1

  • Received:

  • Accepted:

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-030-68436-5

  • Online ISBN: 978-3-030-68436-5

  • eBook Packages: Springer Reference Intelligent Technologies and RoboticsReference Module Computer Science and Engineering