Abstract
This article presents an epistemological multilevel analysis of the embodied cognition studies’ programme. It is proposed that within the cognitive-embodied type it is possible to find at least four distinct hypotheses regarding the role of the body in human cognition: (a) body-in-action hypothesis, (b) extended body hypothesis, (c) ecological body hypothesis, and (d) body-as-a-physical-datum hypothesis. The foundations of these hypotheses and some philosophical debates underlying embodiment are discussed in-depth. After briefly addressing the key contributions of social embodied theories, the article presents a hierarchical model that allows for the analysis of the epistemological consequences derived from each embodiment conception. Finally, a prospective epistemological criticism is introduced to provide a comprehensive and contemporary overview of the issue.
Within cognitive sciences, the computational-representational model of cognition (C-R henceforth)—whose roots can be traced to Alan Turing’s (1950, 1948/2004) studies in informatics carried out between 1940 and 1960, or to Ulric Neisser’s (1967/2014) groundbreaking developments—makes up one of the central and hegemonic hypotheses on human cognition. It proposes to understand—in an analogical sense—our mind as a computer, based on the supposed existence of a group of fixed computational algorithms that act upon mental representations.
For computational models, symbolic operations at the basis of human cognition are no different from those carried out by a calculator, for example, and the role of the body in cognition is relegated to that of a conductor of environmental stimuli to the mind (now considered an abstract dimension with no physical location; Piccinini, 2007; Shapiro, 2007). Therefore, knowledge is represented in symbolic terms and concepts are abstract models of thought unrelated to corporeal activity, lived experience, and areas of the brain that control how the body works in the world (Calvo Garzón, 2007).
The symbols proposed by the C-R model as units of thought represent a group of properties formalised and systematised by Newell and Simon (1976) under the physical symbol system hypothesis (see also Newell, 1980). First, symbols exhibit representational properties; that is, they stand for certain entities that exist in the world, such as colours, tastes, emotions, and shapes, amongst others. Also, symbols are bearers of a trans-situational identity—the reason for which they are independent of their context and of the use they are given, remaining static and unalterable. In addition, symbols are manipulated by the mind with a group of fixed algorithms (i.e., rules), which, generally, follow the structure of a conditional proposition (p→q).
In recent years, in opposition to the abstract nature of the mind proposed by the C-R model, embodiment has emerged as an alternative paradigm. It proposes that cognitive processes and knowledge itself are inseparable from our corporeal experience in the world (Varela, Thompson, & Rosch, 1991/1993), or, in other words, that the body plays a primordial role in human cognition. Cognitive phenomena, including those traditionally deemed as being archetypes of amodal cognitions (such as numeric reasoning and language comprehension), should be considered, from an embodied point of view, as constructions grounded in a variety of corporeal processes (Barsalou, 2008).
Nonetheless, the multiple interpretations of what the concept “embodied” entails in conjunction with diverse conceptions of other associated concepts—such as “experience” or “context”—paved the way for the emergence of dissimilar epistemological positions and different explanations about the development and the characterisation of cognitive processes. Over the last few years, some authors have dedicated themselves to systematising some of these differences, showing that the term “embodiment” can encompass arguments that are not only different but even opposing (Borghi & Cimatti, 2010; van Dijk, van der Lugt, & Hummels, 2013; A. D. Wilson & Golonka, 2013).
This article proposes that within the cognitive-embodied type it is possible to find at least four distinct hypotheses regarding the role of the body in human cognition: (a) body-in-action hypothesis, (b) extended body hypothesis, (c) ecological body hypothesis, and (d) body-as-a-physical-datum hypothesis. The first section of this paper will present the foundations of the first three hypotheses.
The second section will present the fourth hypothesis and will analyse Esther Thelen’s developments within Dynamic Systems Theory (DST henceforward; e.g., Thelen & Smith, 1994) in greater depth, with particular emphasis placed on the concepts of self-organization and phase shifts. Anthony Chemero’s (2009) philosophical contributions are considered as well, specifically those related to his proposed radical embodied cognitive science (RECS henceforward). Both perspectives are congruent with the body-as-a-physical-datum hypothesis.
The third section of this paper will present the philosophical foundations underlying embodiment taking into particular account the contributions of Gabriel Marcel (1935/1949a, 1950/1964) and Maurice Merleau-Ponty (1945/2005, 1964/2010). It is proposed that developments in DST and RECS both constitute forms of epistemological reductionisms that contradict the foundations and objectives of an embodied cognitive science different from the traditional C-R model. Additionally, a hierarchical model that allows for the analysis of the epistemological consequences derived from each embodiment conception will be presented.
Embodied hypotheses’ repertoire
Body-in-action hypothesis
The first hypothesis states that the way we humans conceive and conceptualise the world is based on—and restricted by—the nature and functioning of our bodies (M. Wilson, 2002). So, we have a repertoire of cognitive processes consistent with the body we have and the actions we enact with it. If we were to be different, we would exhibit these differences in the group of characteristics that define our psychological functions. Consequently, it is possible to affirm that a human body is needed to conceive a human world (Shapiro, 2011). Those who defend this hypothesis propose that daily, embodied performances mould our conceptual system: “Imagine a spherical being living outside any gravitational field, with no knowledge or imagination of any other kind of experience. What could UP possibly mean to such a being?” (Lakoff & Johnson, 1980/2003, p. 57).
This hypothesis claims cognition emerges from the actions carried out by our bodies in the world and, consequently, complex mental processes have an embodied base that defines its nature. This proposal contradicts that our representations are independent of the processes that act upon them (e.g., action) and elevates the explicative status of our bodies within cognition. Physical experiences provide the pre-reflective foundations of meaning that allow us to think in an abstract way and execute satisfactorily in what relates to areas such as interaction, expression, and human symbolic communication (Johnson, 2007, 2008).
An interesting element of this hypothesis is worth mentioning: our bodies’ performances in the world and complex cognitive processes (i.e., language or metaphoric thinking) are considered phenomena of a different order, but intimately linked by a supervenience-continuity relationship. Supervenience in that if Ψ is the set of properties of human actions in the world and ø the set of properties of complex cognitive processes, it cannot be the case that Ψ differs and ø does not. Continuity because the embodied, subject–world interaction is the condition of possibility for the constructive process that allows complex cognitive processes to emerge, this being an epistemological characteristic whose roots can be traced to the works of Jean Piaget (1936/1981).
In line with this conception, recent developments within the cognitive semantics field have proposed that linguistic meaning is not achieved via the arbitrary manipulation of symbolic units, but instead is based on the external environment’s corporeal experiences (Bergen, 2012). Thus, embodied simulation theory proposes that we understand language by “simulating in our minds what it would be like to experience the things that the language describes” (Bergen, 2012, p. 13). Simulation is characterised here as the imaginative process of creation of virtual mental experiences of perception and action in the absence of its external manifestation. These developments are congruent with contributions that affirm that imagination is the creation of motor commands that inhibit efferent signals before muscles move (Grush, 2004) and that imagination and action have a neural substrate in common that links them directly and makes them functionally equivalent (Ferri, Frassinetti, Constantini, & Gallese, 2011; Gallese & Lakoff, 2005).
Conceptual Metaphor Theory (CMT) proposed by Lakoff and Johnson (1999, 1980/2003) is also a good example of the hypothesis analysed in this section. It states that metaphorical linguistic or written expressions are surface realisations of cognitive metaphorical projections that allow us to comprehend abstract areas of knowledge (target domains) via analogical relationships between said areas and others of a more concrete or embodied nature (source domains). Therefore, metaphor is not an accessory to language or rhetoric anymore, but a tool for understanding (for a discussion, see Alessandroni, 2017). For example, while classic linguistic analysis would not consider the phrase “the time will come when” as an example of figurative language, for Lakoff and Johnson it makes up a metaphorical expression, so long as the time cannot physically (i.e., literally) move in space. Since we do not have sensorial receptors associated with the passing of time, we are only able to conceptualise it using characteristics of the physical domain (TIME IS MOTION metaphor). In this way, we allot different temporal events with the ability to move in space (time-moving perspective, e.g., “The exam date is getting closer”), or we spatialise and immobilise them in order to move towards them (ego-moving perspective, e.g., “We are approaching the end of the congress”; Boroditsky & Ramscar, 2002; Richmond, Wilson, & Zinken, 2012).
As in the example, all metaphorical projections involve the importation of information taken from the source domain to the target domain. This process has been deemed cross-domain mapping. These mappings are possible thanks to the existence of image-schemas (Johnson, 1987/1990) that allow us to build our experience on the levels of physical and movement perceptions, understand abstract aspects of reality, confer meaning, and guide our reasoning of the world. They are “recurring, dynamic patterns of our perceptual interactions and motor programs” (Johnson, 1987/1990, p. xiv) that emerge due to our recurrent kinaesthetic experience in the world (for a critical review of image-schemas, see Alessandroni & Rodríguez, 2017). For example, the BALANCE schema emerges from activities such as being moved, swinging oneself, walking without falling, and dancing, by a process of conflation and abstraction of experience, images, and perceptions related to activities that involve two opposing forces (see Figure 1).

Embodied construction of the “balance” image-schema.
Since metaphorical thought is an extension of image-schemas (Lakoff, 1990), and that the latter are embodied structures, metaphorical thinking becomes a psychological function associated with and dependent on our corporeality.
Extended body hypothesis
Apart from considering the influence that an individual’s actions exercises over the genesis of cognitive processes, to state that cognition is embodied could imply conferring some relevance to the environment within which a subject acts. In this way, some argue that, for example, the objects surrounding us (such as cellular phones) play a fundamental role in our cognition (Clark, 1997, 2008). In this model, cognition emerges from a group of relationships within the brain-body-environment continuum, or as a succession of feedback loops that arise from the intimate integration between physical action and environmental elements (Clark, 1997, p. xii). The body should no longer be considered as a means for transmitting data, such as in the C-R model, but, instead—literally—as a part of the mind. In other words, while within the C-R model the body is a casual influence, and in the body-in-action hypothesis the body emerges as a causal influence, in the extended body hypothesis, it is no longer an influence but a constituent, internal part of the mind itself.
That which differentiates the theorists grouped within this perspective is what they consider the constituents that structure the mind. This has led other authors to refer to this model as the constitution hypothesis (Shapiro, 2011) or extended mind (Rowlands, 2010). Given that my analysis is directed at testing the status of the concept of corporeality within the different variants of embodiment, I have maintained my classificatory proposal, which situates the body at the centre of the scene. In what follows, I will summarise examples of contributions that make up this hypothesis.
In a classic article, Clark and Chalmers (1998) propose that the brain does not carry out its functions isolated from the body and the external world, but instead, in an integrated way. They advocate, for example, that an agenda can be a constitutive element of the human cognitive system. Their argument relies on a functional theory of the mind within which mental states are functional states that can be carried out by different material systems (Held, Knauff, & Vosgerau, 2006). So, to remember a way to get to a location using merely memory or notes taken before in an agenda are functionally equivalent procedures that provide the same information. Clark and Chalmers’ conjecture, based on the multiple realizability causal model (Looren de Jong, 2003), is interesting in that, although it retains some foundations of the C-R model (debate that is outside this paper’s reach), it proposes giving our bodies and a group of objects the status of cognitive tools: The body is the primary tool enabling the intelligent use of environmental structure. It acts as the mobile bridge that allows us to exploit the external world in ways that simplify and transform internal problem solving. … The body is thus the go-between that links these two different (internal and external) sets of key information-processing resources. (Clark, 2008, p. 208)
Similarly, Robert Wilson (2004) proposes that using pencil and paper can collaborate to resolve mathematical problems by extending our cognitive system. When we resolve an equation on paper we can annotate in the margins, cross out wrong steps to eliminate them from the final resolution, highlight a result to lend it relevance, amongst other possibilities. All of these strategies are not only helpful towards reducing cognitive demand but also make up a part of the mind.
One problem deriving from this perspective is how to decide the limits between the components that make up the aforementioned continuum. For example, it becomes unclear where the body ends and the environment begins: is the skin the limit (Maturana, 2006)? Or is it the phenomenological space where our body image ends (Gallagher, 2000)?
A related problem lies in determining which elements of the continuum are structural parts of cognitive processes themselves. Normative tools (i.e., rules) have been proposed to circumscribe this problem. In this way, Shapiro (2011) proposes that if C is a constitutive part of an event or process P, then C must exist within the same spatial–temporal coordinates as P. So, for a process P, if C takes place before P or if C takes place in spatial coordinates different from those of P, then C is not constitutive of P. However, I believe this proposal cannot resolve the problem. Let me analyse the case of visual perception where a subject is wearing sunglasses. How should these be considered? As a constitutive part of the cognitive-perceptual process or as an external causal influence? What role does the sun itself play? Would not the density of the air particles in the environment be a constitutive part of this process? The risk lies in asserting that “everything is cognition,” affirmation that, as an explanation, has no value.
Ecological body hypothesis
Within the cognitive sciences some have argued that the hypotheses aforementioned make up a group of weak and unproductive embodied models because they maintain elements of the classic C-R models and are not innovative (Kiverstein, 2012). Nonetheless, these same theorists believe it is possible to propose a serious embodied model, committed to a new status of the body: the replacement hypothesis (Glenberg, Witt, & Metcalfe, 2013; A. D. Wilson & Golonka, 2013) or, as I call it, the ecological body hypothesis.
This hypothesis argues against the use of abstract constructions such as representations and concepts to think about cognition. Instead, the theorists propose that cognition should be thought of as the result of the relationship between perception–action bodily couplings that integrate nonlinear dynamic systems (L. M. Ward, 2002). Here, the body is neither a collaborator in constructing complex processes nor is it a mediator of interactions with the environment. From this point of view, physical interactions are, in and of themselves, cognition, and it is unnecessary to invoke other elements pertaining to any levels other than the bodily one—understanding the body as a material, organic entity—to arrive at a successful explanation of cognitive processes.
This conception of the body is congruent to developments within movement studies that propose that any gesture can fulfil not only a metonymical function—as a surface realisation standing in the place of, for example, a competence—but also an epistemic function—it can be a unit in the production and structuring of knowledge (Hostetter & Alibali, 2010; Roth & Lawless, 2002). Likewise, it has been sustained that during verbal communication, gestures made by participants drastically raise the level of understanding of the semantic content being communicated (Lozano & Tversky, 2006). Similarly, during a sight-reading musical task, the use of gestures tends to be congruent with temporal and morphological aspects of the musical discourse at play, which reduces the cognitive weight that the task entails (Pereira Ghiena, 2014).
Experimental studies also showed that blocking the movements of the corrugator supercilii muscle with cosmetic Botox injections diminishes the ability to process phrases describing states of anger or sadness, but not those expressing happiness (Havas, Glenberg, Gutowski, Lucarelli, & Davidson, 2010). These studies suggest that physical movements, including structures such as muscles, are in and of themselves cognitive agents and that their participation in cognitive processes does not necessitate conjuring up an abstract, representational level. This affirmation will be developed with greater detail in the following example (for an extended discussion of the example, see Clark, 1997; A. D. Wilson & Golonka, 2013).
Let us consider the case of a baseball player (A), who must catch a ball thrown by another player (B). The ball is far from A and looks smaller due to its being at a great height. To carry out a successful movement, A must get closer to the place where the ball will fall and intercept it. A cognitive explanation of this using C-R terminology would propose that the player uses information provided by the environment (such as initial direction, speed, and angle of the ball) to generate an internal representation and make predictions regarding where the ball will fall, a strategy that entails a high degree of cognitive demand. However, and given that estimations of trajectories based on human visual perception have a large margin of error, A probably will not be able to catch the ball.
A dynamic explanation based on the ecological body hypothesis would propose, instead, that from the moment the ball is in motion it makes up, along with A, a pair of terms linked by a relationship of prospective control that is not representational, but instead grounded in perception-action bodily couplings. This means that both the player and the ball enter a dynamic field of continual regulation that leads A to adopt one of these possible solutions: (a) align himself with the trajectory of the ball and run so the ball appears to move with a constant speed or (b) run laterally so it seems to move in a straight line. In both strategies, the ball’s potential property of being caught—affordance (Gibson, 1977)—is specified in a direct and continuous way by visual information, without internal simulation or computational predictions.
Dynamic Systems Theory (DST) and Radical Embodied Cognitive Science (RECS): The body as a physical datum
I have decided to approach in a separate section those contributions linked to the body-as-a-physical-datum hypothesis, a perspective that is a subgroup of the contributions pertaining to the ecological body hypothesis, adopted by DST (Thelen, 2000a; Thelen, Kelso, & Fogel, 1987; Thelen & Smith, 1994) and RECS (Chemero, 2000, 2009).
Dynamic Systems Theory (DST)
Theoretical framework
DST is a theory based on a group of mathematical instruments that allow for the formalisation of a dynamic system, that is, a collection of elements that undergo certain changes over time, by which it is possible to describe the system’s behaviour. It is a proposal that emerged from the field of mathematics, directly linked to chaos theory, vector field theory, and complex systems theory (Guckenheimer & Holmes, 1986/2002). All dynamic studies of the behaviour of a system entail the process of identifying the parts susceptible to variation and characterising all possible trajectories towards change. Both operations enable the formalisation with mathematical rules, which allows for the description of how a system evolves between two moments t0 and t1. Although this focus was at first used to describe physical systems, it has been applied to research on human cognition for several years.
DST and development
Such is the case of Esther Thelen who, in line with the generic foundations of embodied cognition, argued that the way to overcome the fallacies of the C-R model when attempting to understand developmental processes was to adopt a dynamic focus of micro-level behaviour (i.e., mechanical physical interactions; Thelen et al., 1987). Thelen’s main concern was to find a way to mathematically describe patterns of human development—considered as an open system—and how they evolve. This is a crucial point in her theory given that only open systems can experience transformational change in their organisation (such as those that define human cognitive development, here deemed phase shifts) because they maintain an equilibrium different from that of thermodynamic systems, usually defined as the state in which all the environmental conditions that affect a system compensate each other so the system is unable to experience any state of change. Every system is defined in terms of one or more types of behaviours it prefers over others, usually called attractors. For Thelen, then, to speak of behaviour is equal to speaking of attractors of varying stabilities, and to study development is analogous to mathematically describing the attractors’ states that generate the continuous stabilisation and destabilisation cycles of a system.
Previous studies have argued that theories such as this can prove useful when studying basic phenomena, but are not so when attempting to explain how these phenomena escalate to allow for the emergence of the capabilities that make human beings distinct from other species (like rational thinking or creative imagination; Kiverstein, 2012). For Thelen, instead, DST is suitable for the investigation of human cognition because the principles that guide the organisation of complex dynamic systems are independent from the material substrates where they are applied. Thus, the dynamic paradigm can be equally employed to study any psychological phenomena, inorganic chemical reactions, and galaxy formation (Thelen & Smith, 1994). Complex cognitive processes are no different from the less abstract mental operations upon which the former are ontologically based in that they too are rooted in real and concrete processes. They both occur in a specific physical space and time: “What, where, are these explicitly symbolic representations if they are not also distributed patterns of activity in time?” (Thelen & Smith, 1994, p. 335).
One phenomenon that Thelen could reinterpret with DST is the coordination of the movements involved in a baby’s kicks and standing up. According to her, the leg muscles act like springs functioning under the control of gravity (Thelen, 2000b). The movements involved in the acts of standing up and kicking are equivalent to the oscillatory movements of a spring and are not determined by any mentally encoded cognitive plan. Motor development is, simply, a product of natural coordination or the self-organisation of certain physical forces around one or several attractors in our infant bodies. The unit of analysis of dynamic studies includes, thus, the individual and the changes in the behaviour she exhibits over time. The appearance of new developments (be they motor, cognitive, or social) is explained through the relative dynamic status of a complex dynamic system and by the substitution of one dynamic topography with another (Thelen & Bates, 2003).
Thelen’s position is embodied given that the dynamic principles that apply to processes and events on any level of complexity are equivalent to those that describe the behaviour of complex cognitive processes, and that it is possible to study micro-level behavioural development using a DST perspective. Thus, it becomes plausible to sustain that human cognition is a complex system’s activity that can be described as a function of our actions in the world, where actions are understood as non-propositional, mechanical, emergent, and dynamic forms of corporeal behaviour. For Thelen, cognition is (in its entirety) embodied because it can be explained by analysing distributed bodily activity considered as a dynamic system that produces concrete physical changes in the world.
Radical Embodied Cognitive Science (RECS)
RECS’s project
Radical Embodied Cognitive Science’s project is based on the assumption that cognitive phenomena can be explained without using representations. Instead, this projects’ proposal is that a valid and profitable way to explain cognition is to replace representational theories with dynamic theories that include coupled oscillators in their explanations (Chemero, 2000).
Oscillator theories appear within the field of physics as an attempt to create a model for the emergence of order from chaos, using certain logical forms of self-organisation (autopoiesis; Strogatz, 2003). Experimental studies showed that, for example, after a concert, the way a crowd’s applause is synchronised and desynchronised follows a mathematical model defined by a series of states and phase shifts (Néda, Ravasz, Brechet, Vicsek, & Barabási, 2000). A round of applause goes through specific dynamic states (such as cacophony, synchronisation, and the decline of synchronisation) that are equivalent to the states by which other systems, such as cardiac muscle cells and fireflies, organise spontaneously. These states result from automatic autopoiesis processes that have no relation to computational processes or conscious decisions, but instead to the mechanical interactions between our bodies and the world.
The term oscillator refers to any circuit that, while receiving a steady source of power, can produce an alternating current signal regardless of its specific waveform. Thus, it is plausible to say that spectators act as coupled oscillators and that their type of applause can be described in terms of dynamic behavioural archetypes.
There are three types of oscillators: (a) relaxation oscillators, (b) spring-mass oscillators, and (c) hybrid oscillators. Relaxation oscillators accumulate voltage slowly, but are quick to liberate it. These oscillators synchronise effectively with themselves and with a rhythmic input, but cannot do without the input. This type of oscillator’s biological archetypes are neural systems that owe their activity to the difference between potentials that, once reached, favour the propagation of electrochemical impulses. Spring-mass oscillators cannot synchronise with rhythmic signals as successfully as relaxation oscillators because they have a determined mass that lends them a momentum. Paradigmatic examples of this type of oscillator are physical systems, such as springs. Last, hybrid models of oscillators can synchronise quickly with input signals but depend partially on inertia, a characteristic that allows them to “follow a rhythm” even in the absence of a signal.
For Chemero, to take embodiment seriously is equivalent to adopting a radical, anti-representational position based on the dynamic foundations of the oscillators described above (Chemero, 2009). For the author, to state this does not imply making a metaphysical affirmation such as, “cognitive systems are ontologically non-representational”; instead, it is an epistemological affirmation that states it is unnecessary to make use of representations to come up with the best explanation of the cognitive system.
The Watt governor
Chemero, like Thelen, states that human cognition can be understood in mathematical terms. To defend this position, the author makes use of the Watt governor as an example (see van Gelder, 1995). A centrifugal (or Watt) governor is the component of a machine by which the speed of the machine is kept nearly uniform. It is part of a mechanical feedback chain that provides a parameter that is a function of angular velocity. This parameter can take the form of a movement that acts upon a negative feedback control valve that feeds the motor to maintain a uniform speed. It is composed of two or more masses that rotate around a central spindle. Because of the centrifugal force, the masses distance themselves from the central spindle and, in doing so, create an opposing force to a spring system causing a lever system to transform the radial movement of the masses into an axial sliding along the collar (Maxwell, 1867).
For Chemero, the best explanation of how these type of systems work is not the computational-representational explanation, but the self-regulatory explanation provided by DST. Thus, the machine’s speed changes are a function of the speed valve’s value, which depends, in turn, on the angle of the centrifugal arm. These two dynamic systems are linked: a change in the angle of the centrifugal arm fluctuates the total dynamic of the system due to a modification in the machine’s speed, and any change in the machine’s speed variates the overall functioning of the system because it alters the centrifugal arm’s angle.
By discovering the equations that regulate the system, Chemero argues, we can claim to have described the regulator’s behaviour. And if we agree with that (particularly in what relates to the definition of “behaviour”), we must also accept that finding the equations that regulate human mechanical action is equivalent to explaining human cognition (Chemero, 2009).
The philosophical foundations of embodiment
In this section I present some of the philosophical compromises involved in the metaproblem of the body (Zaner, 1971) to clarify some epistemological issues regarding embodiment. These underlie the embodied cognitive and psychological programme. For that, I briefly explore the contributions of two authors (Gabriel Marcel and Maurice Merleau-Ponty) who propose to approach the phenomenological study of the body as lived experience.
Against what is embodiment really rebelling?
Beyond being an alternative to the C-R model of human cognition, the proposal that the body is a fundamental part of cognitive processes is, on a philosophical level, a response to objectivist and subjectivist ontological and epistemological strains and, more generally, to all forms of realism. For the latter, cognition is a psychological function that allows subjects to incorporate into their repertoire of mental content the effective organisation of objects of the environment in an automatic and unconscious way. Thus, concepts with which we operate cognitively are objective, natural, and have an exact correspondence to ontological structures in the world. This belief is based on principles such as:
The world comprises a discrete totality of objects (independent of the cognitive subject) which exhibit rational and objective logical relationships between them.
These relationships are self-evident, univocal, and permanent for human understanding, and are acquired in a direct way.
There exists a true and complete description of the being, the essence, of the world. Attributes connect concepts to the real world, that is, link the meaning (intension) of a concept with the objects (extension) that follow its definition.
Truth always involves some correspondence.
Given that concepts reflect reality by means of names that refer to objects in the world and predicates that denote properties of them, according to this perspective there exists only one valid and universal conceptual system, and it is semantically measurable (Muñoz Gutiérrez, 2006). For embodiment, reason is not universal (it is not a part of the universe’s structure), but is instead a property that emerges from the nature of our bodies in the world. This means that our cognitive structure is based on (or is equivalent to) bodily interaction with the environment. In this sense, embodiment has taken on an interactionist-realist perspective (Lakoff & Johnson, 1999) or an experientialist one that states that truth is relative to our conceptual system that, in turn, is anchored in our lived embodied experience (Casasanto, 2013; Varela et al., 1991/1993). Although this concept does not belong in any fundamental way to any of the discussed psychological hypotheses of embodiment, it is fundamental within the field of philosophy. From an experientialist perspective the body should not be considered a biological entity but a phenomenological one, as a body lived through our experiences.
Gabriel Marcel
Gabriel Marcel’s philosophical project is directed at unravelling the metaphysical conditions of existence. To that end, he analyses the fundamental experiences inscribed in our condition of human Beings. This line of thinking takes on the form of a philosophical anthropology that aims to analyse the lived experiences of subjects (Marcel, 1950/1964). For Marcel, modernity—at the hand of realists—led to a crisis of the human subject, who was displaced either by abstract reason or by ineffable idealism. This modern loss resulted in the urgent ontological need for fundamental knowledge of oneself as a subject (Marcel, 1927/1952). However, in this quest for consciousness, the Being who inquires “who am I?” is also the Being who is expected to respond. This query of man by man cannot be equivalent to other inquiries into objects of knowledge. The Being cannot exclude itself from this inquiry, because the act of exclusion, in and of itself, distorts the problem. Regarding one’s own Being, it becomes impossible to create a relationship of n’importe qui similar to those we assume with objects, understood as those entities that are indifferent to us. To be an object and to be a subject are mutually exclusive alternatives (Marcel, 1949b) that entail different types of reflection. Marcel distinguishes between a primary reflection (pensée pensée) that allows for the apprehension of entities as objects, and a secondary reflection (pensée pensante) that allows subjects to apprehend their condition of Beings without, in the process, dissolving the unity of lived experience, which proceeds by unifying rather than by dividing.
All Being is an embodied being [être-incarnée] that exists as an embodied consciousness, and secondary reflection is always embodied thought, meaning it is recognised in the body by means of the body, that is, in the Being itself (Marcel, 1935/1949a). The body is something I have (in an objective sense) and, as such, is “the datum common to my consciousness and to other possible consciousnesses” (Marcel, 1927/1952, p. 18), but it is also something I am in the phenomenological sense of the word. From this perspective, all inquiry that takes the body as its object will be incomplete, because it considers the unit of analysis to be a body and never my body.
Maurice Merleau-Ponty
For Merleau-Ponty, the problem of the body is part of a greater issue tied to the criticism he makes of classical theories of sensorial perception. These theories mistakenly assumed the existence of an inert and lazy subject, limited to the role of a passive receptor of external information. Merleau-Ponty proposes a new theory of the body to make possible the emergence of new perception theories that consider perceptive experience as a mutually penetrating, complementary, and overlapping relationship between the body and the object of perception. Perception is more than just a gnoseological fact; it is the Being’s private experience of opening up to the world in its own original ontological modality.
For Merleau-Ponty, the body is a type of knowledge and an expressive unit that can only be known through lived experience, that is, through embodied performances (Merleau-Ponty, 1945/2005). Things and the world, in general, are cognisable only through the constructive activity of the body. However, this does not imply that the body acts as a bridge or a mediator. In fact, quite the opposite: the body is a decisive moment in the constitution of the objective world by embodied action, it is a stage director of reality (Merleau-Ponty, 1964/2010, p. 21). The body and the world are two elements chiasmatically linked, intertwined.
This reflection on duality leads inevitably to the act of considering said duality not as a function of its components, or in terms of a transcendental unit before unfolding, but instead as an incomplete body–world dialectical interaction (without the third dialectical term, synthesis). As the author states, it is necessary “to begin there: There is no identity, nor is there no identity or no coincidence, there is inside and outside in as much as one rotates around the other” (Merleau-Ponty, 1964/2010, p. 317).
In this sense, corporeality cannot be considered an object or a consciousness. Instead, it is a third form of reality characterised by a chiasmatic immanence, that is, Merleau-Ponty considers it a linking unit, a dual unit, or a unitary duality. The body cannot be reduced to an objective Being-in-itself or a subjective Being-for-itself. It is always a lived body, and consciousness is always embodied consciousness.
Varieties of embodiment: A proposal and one criticism
In the preceding sections, I have put forth various hypotheses that have guided research within embodied cognitive sciences. What I have attempted to highlight is that, within a paradigm that appears to be based on congruent research perspectives and rooted in the same philosophical and epistemological principles, there coexists a wide range of dissimilar and even opposing versions of the “body,” if this were to be considered the unit of analysis common to the perspective.
It should be noted that, until now, I have said nothing about other embodied theories, i.e., not necessarily cognitive. To offer a comprehensive review of the issue, I discuss, in the following paragraphs, some of the key contributions of social embodied theories. After that, I present a possible hierarchical organisation of the cognitive and philosophical perspectives reviewed in the earlier sections of this article and a prospective epistemological criticism.
More varieties: Social embodied theories
Over the last 50 years the body has not only been a central topic within cognitive studies. Other perspectives and disciplines have addressed the problem of the body, proposing points of view not necessarily congruent with the hypotheses delimited before. Because the main aim of this article is to contribute to the epistemological analysis of the embodied paradigm within cognitive sciences I will not dwell at length on this point. However, I will provide a brief characterisation of different theoretical strands that have dealt with embodiment from a social point of view.
The body as a field of socio/normative inscriptions
It was Foucault (1976/1980) who, in his writings about the history of sexuality, proposed for the first time that the body is a construction dependent on history and a field on which different regimes of bio-power act in a capillary fashion. In Foucault’s anti-essentialist philosophy, the body is a social body and embodied behaviour is historical. The traditional distinction between the natural body and culture is abandoned. The body is inseparable from the ideational and cultural meanings that constitute it. Judith Butler, while criticising Foucault’s notion of genealogy, agrees with him in that the body is a site where regimes of discourse and power inscribe themselves (Butler, 1989, p. 601). In line with Foucault, and from sociology, Bourdieu proposed that bodies are subject to certain habitus. A habitus is “a system of lasting, transposable dispositions which, integrating past experiences, functions at every moment as a matrix of perceptions, appreciations, and actions” (Bourdieu, 1977, pp. 82–83). In this fashion, thought, perception, and embodied action schemes are revealed on the basis of a social genesis. This puts in crisis those theories that conceive social practice as a mechanic reaction determined by a set of circumstances, and blocks simplified notions that understand “context” as “cause.” Different authors have explored the richness of the concept of habitus to explain embodied social forms of doing and feeling, and embodied normative practices (see, e.g., Joranger, 2016; Pizanias, 1996; Voestermans & Verheggen, 2013).
The body as gendered body
The feminist and critical epistemological stances have interrogated the existing relations between body, sex, gender, and cultural practices (see, e.g., Frerks, Ypeij, & Sotiria König, 2014; Grosz, 1987, 1994). For Grosz, for instance, one of the main problems for feminist epistemology is related to the ways in which the body is produced, and the ways in which bodies reinscribe themselves onto their sociocultural environment (Grosz, 1994).
Given that the body is the primary object of social production and inscription it is considered a key concept to understand the social constitution of sexual identities and other phenomena such as agency, domination, violence, and class differences (Morgan, 2005; Tucker, 2010). Bodies are always, in the feminist epistemology, gendered bodies.
The body as a dialectic term of the cultural self
From the areas of sociology of the body and anthropology of the body, certain authors have noted that the denial of the Cartesian substance dualism proposed by some cognitive approaches of embodiment does not solve the phenomenological problem of the self. In other words, even if we accept that the self is inseparable from the body we still need to explain what relations exist between these two terms. Body and self (organism and lived body) do not coincide because, as Gadow explains, “for consciousness, the body is still the other” (1980/2004, p. 70). This unsolvable difference enables relational levels between the body and the self. For this author, for example, there exist four relational levels that are qualitatively different: (a) lived body, (b) object body, (c) harmony of (a) and (b), and (d) subject body. The importance of the lived body becomes quite clear when considering, for instance, ritual performativity, where the body becomes an instrument of meaning (Bell, 1997). Because of this, some state that embodiment, understood like so, can be considered as a paradigm for anthropological research (Csordas, 1990, 2002).
The institutionalised body
Some authors belonging to institutional psychology developed, in sympathy with Foucault’s and Grosz’s approaches, the idea of the body as an institutionalised form of subjectivity. This interpretation usually has two readings. On one hand, it is possible to think of bodies as legal entities that are conformed and act according to the institutional patterns of laws and rules, by their transgression, or by their ignorance (Folkmarson Käll, 2016). Institutions pierce bodies determining them (Costa Paz, 2014), and they, because of their involvement, re-signify the former (Gärtner & Ortmann, 2016, p. 94).
On the other hand, other authors, adopting a broader definition of “institution,” think of the body as an institution (Baremblitt, 1992/2005; Schvarstein, 1992). For instance, for Baremblitt, institutions are decision trees, a series of prescriptions, prohibitions, and different options that guide social life. They are, at the same time, logical, dispositional, ethical, and aesthetical, and they translate into a spectrum of cultural instantiations, such as laws, principles, maxims, codes, rules, guidelines, customs, or habits (Baremblitt, 1992/2005). In this way, bodily articulations (i.e., corporeal tone, movements, gestures, and actions) can be seen as an expression of the culturally determined institution of the body.
The semiotic body
Within developmental psychology and psychoanalysis much attention has been paid to certain embodied meaning-making phenomena that take place in early childhood due to the interactions between baby and parents (see, e.g., Bruner, 1975; Lacan, 1966/2009; Leuzinger-Bohleber, Emde, & Pfeifer, 2013; Mahler, 1975/1977; Spitz, 1965/1996; Thibault, 2004). In these studies, it is assumed that, from birth, there coexist two modes of being: one material (biological) and one semiotic (product of the meaning inscriptions that others impose on the baby). The body is, from the beginning, a socially semiotised body and not only a means to interact with the physical environment. As Booth states, the human body is social because “it is born and develops by being embedded in an emotional network of social attention and communication. Human communication is an intercorporeal affair, a scene of bodies in relationships of movement and anticipation” (2016, pp. 6–7).
Thanks to these communication networks, social experiences involve, de facto, the creation of semiotic bonds that demand a labour of interpretation or the establishment of an intersubjective common reference. After the Peircean semiotic theory, authors such as Thibault (2004) have proposed that the body is the first locus where meaning is generated as a result of complex semiotic trajectories that link the dynamic of the individual with that of its context.
A hierarchical organisation
In what remains of this section I would like to propose that the different cognitive hypotheses (including the philosophical ones) I analysed can be hierarchically organised to reach a better understanding of the structure of the embodied programme of cognition. It is manifest to me that phenomenology, as explained by Gabriel Marcel and Maurice Merleau-Ponty, emerges as the most comprehensive theory. It is the only one that appears to highlight five aspects of embodied cognition which have recently given rise to the 4E+A acronym: embedded, embodied, enacted, extended, and affective (D. Ward & Stapleton, 2012). Within the reviewed philosophical contributions, the world is conceptualised as carnal spatiality, the body as an expressive totality, and the processes of knowing as embodied and phenomenological states of possession of the world. Furthermore, subjective aspects of lived experience are considered fundamental to a complete understanding of what it means to speak of corporeality in an experiential sense (to be a body) and not merely in a mechanical and possessive sense (to have a body).
Other hypotheses presented in this paper consider the body in terms of only a few of the 4E+A cognition elements. The body-as-a-physical-datum hypothesis is particularly interesting in its consideration of the body as a mechanical entity susceptible to being divided into its parts and formalised according to mathematical models. It would appear to place objective, physical, and geometric space as prior to the experience of the corporeal dimension. But the latter, in turn, is not truly at the base of our experience, but instead in the teleology of our cognoscent conscience (Ramírez, 2013). That is, it is a construction and not an a priori fact. On the other hand, those contributions congruent with DST omit phenomenological aspects such as consciousness of intentionality or lived experience from the concept of corporeality. At the same time, they eliminate factors that other hypotheses include, such as environment (as understood by extended body hypothesis) and intentionality (as conceptualised by body-in-action hypothesis).
DST appears to be based on a double reductionism of cognitive experience to a group of corporeal mechanisms. This reductionism can also be found in some cognitive development theories (Castorina, 2002) and neurobiological theories of cognition (Garza & Smith, 2009). This dissociative epistemological movement is characteristic of functional explanations that disassemble complex systems to analyse them (de Jong, 2003). Here, it has the objective of reducing human cognition to a set of nomological and mechanical explanations with the hope—and no guarantee—that the change in level of explanation is equivalent to a better understanding of cognitive phenomena than the one proposed by representational theories. I say “double reductionism” because DST operates, on the one hand, through an ontological reductionism that transforms Merleau-Ponty’s chiasmatic immanence into a set of anatomical structures and biological functions. But, on the other hand, it also resorts to an explicative reductionism in pretending to explain all cognition with a mathematical apparatus. According to my understanding, this results in a diminished and very limited version of human cognitive processes.
I believe, like Overton (2008; Overton & Palermo, 1994), that embodiment owes to itself a revisiting of the philosophical contributions that paved the way for the emergence of the embodied perspectives within the study of cognition. This would encourage epistemological discussion and the consolidation of a relational perspective that can analyse the multilevel relationships that exist between the body as a lived experience and the body as a physical object, two concepts already highlighted by Gabriel Marcel in his works. It is necessary to consider the bodily dimension within the bio/socio/cultural framework that characterises it, testing the types of relationships between conceptual pairs elsewhere considered as polar opposites, such as experience–action or individual–society. To that end, I also believe that social embodied theories have much to contribute because they integrate key aspects of humaneness like development, normative structures, historical determinants, social representations, semiotic operations, phenomenological experiences, cultural practices, and symbolic meaning attributions.
Consequently, I would like to propose that only a non-excisionist and dialectical theory that integrates diverse elements into relationships that are not necessarily causal or linear, but instead dialectical, could prove helpful in explaining human cognition. A theory like that could account for how our body, as a unit of meaning, participates in the co-construction of cognitive processes in specific material and cultural contexts, while still maintaining the psychological subject as central. The task that remains is that of designing a new epistemological landscape that allows for the emergence of a novel programme bearing these characteristics that, surely, is soon to come.
Footnotes
Acknowledgements
I wish to thank Dra. Diana I. Pérez (UBA) for her encouragement, and accurate suggestions that improved the quality of this paper. Thanks also go to the whole team of IIF-SADAF and LEEM for the enriching discussions held during 2015–2016. I am also grateful to Dra. Ana M. Talak (UNLP) and Dr. José Antonio Castorina (FLACSO) for their useful comments and corrections on a previous draft. Comments made by anonymous reviewers were very useful.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
