The WE-turn of Action: Principles

1.1 What is the WE-turn?

This work sketches a philosophical system called the “WE-turn,” which, at its core, advances the thesis that the proper subject or unit of somatic action, self, goodness, responsibility, rights, well-being (or well-going), and freedom should shift from the individual “I” to the multi-agent system “WE” that necessarily includes the “I.”

1.2 From Action Theory to Axiology

The WE-turn has an argumentative structure that derives many WE-turns one after another in a domino-style, through various inferences from numerous premises. The first WE-turn is a thesis of philosophical action theory, from which various WE-turns on values are derived. In this sense, the WE-turn directly links the theory of action and the theory of value or axiology in a way that the latter is based on the former.

1.3 Toward an Alternative Model for Human–AI Relationship

Members of the “WE” here include AI and other artifacts, non-human animals, and the natural environment. By including these various agents other than the “I” and other humans into the “WE,” the WE-turn also proposes an alternative model for the ideal relationship between humans and AI to the conventional one.

2.1 Two Incapability Theses

Among the assumptions of the WE-turn are the two fundamental incapability theses of a human individual.

2.1.1. The first incapability thesis of single action: No individual can perform ones’ somatic action alone.

2.1.2. The second incapability thesis of full control of necessary others: No individual can fully control any other agents whose affordances are necessary for the performance of one’s somatic action.

In this work, only the first thesis plays a role. But the second becomes essential when we talk about the internalization of Self-as-We, or first-person awareness of the WE-turned self by “I,” that will not be discussed here.

2.2 From Capability to Incapability

Roughly and generally speaking, Western philosophy, especially its modern and contemporary versions, has credited the human individual subject, and her alone, with various agentic capabilities, such as intellectual capability, communicative capability, and capability to control nature and technology. The most important capabilities are the capability of a single action and especially that of single-self decision-making. Many modern and contemporary philosophies claim that a person ought to be capable of independent action, and therefore constitute a self-sufficient, self-standing, and autonomous individual or the “naked I.”^[1] Human dignity, irreplaceability, or freedom, they argue, consists in the self-sufficiency, self-standingness, and autonomy of the individual.

In contrast, the WE-turn focuses not on capability but on incapability of individuals, and argues that it is precisely because they are incapable that they are irreplaceable to “WE.” In particular, the “fundamental incapability” mentioned above implies an objection to the very idea of the human being as a self-standing individual that has been accepted as the norm by modern philosophy. The foregrounding of these “incapacities” dismisses the “naked I” as an illusion that can perform actions without the help of others or “WE,” and replaces it with a being who, as we will see later, is always already a member of “WE.” The “WE” is not a removable costume that reveals a “naked I” when we take it all off.

3.1 What is the WE-turn of Doer?

The first WE-turn is the WE-turn of action or doer, that is, the claim that the subject or doer of any human somatic action is not an individual “I” but a multi-agent system as the “WE” that includes the “I.” It is the turn of the subject or unit of somatic action from “I” to “We,” from “I do X” to “We do X.”

3.2 Derivation of the WE-turn of Doer

This shift is derived from the following three assumptions: (1) the first incapability thesis, (2) the thesis of doer as necessary and sufficient agent, i.e., the claim that the doer or subject of a somatic action is nothing but an agent that is necessary and sufficient for its performance, and (3) doer-externalism: the doer or subject of a somatic action need not be always aware or conscious of one’s doing the action.

3.3 Fundamental Incapability

While humans are capable of doing many things, they also have various incapabilities. Among those, there are universal and inescapable incapabilities that all humans have without exception and that we cannot overcome no matter how hard we try. Let us call them “fundamental incapabilities.” One of the most serious ones, for example, is the incapability to live forever.

Another serious one is the “incapability of single action,” mentioned above. Although it is widely held that a single person should be able to perform so-called “solo actions” such as jogging or singing solo. However, the first incapability thesis professes that this is not the case.

Let us exemplify this by referring to the somatic action of “riding a bicycle,” which at first glance seems to be a typical example of a single action. In order to ride a bicycle safely and successfully to its destination, “I,” the cyclist, must ride the bicycle properly. But that is not enough. The bicycle must work well without a tire getting punctured or losing its chain along the way. Also, without proper transportation infrastructure, such as roads and signal systems, the bicycle cannot safely reach its destination. Such transportation infrastructure is supported by various people. Moreover, without the historical background of the bicycle, which was invented by someone in the past and spread widely through a process of various improvements, and without the industrial system in place today in which bicycles are systematically produced and routinely sold at bicycle shops on street corners through complex distribution channels, “I” would not be able to ride a bicycle. A wide variety of people are integral to this historical process and industrial system.

Riding a bicycle would also not be possible without adequate atmospheric oxygen levels, atmospheric pressure, and gravity. In this way, the somatic action of “riding a bicycle,” which at first glance seems to be performed by “I” alone, requires not only “I” but also various people other than “I” – including those who are alive today and those who have already passed away – to carry out the action. It needs the inputs of those who intentionally support bicycling here, those who have no such intention, and in some cases even of those who want to hinder it. Furthermore, it needs a wide variety of non-human creatures, inanimate objects, the natural environment, artifacts, and social institutions.

The somatic action of “riding a bicycle” is therefore only possible when all of these agents, whether intentional or merely causal, exert a certain agency. In other words, if any one of these agents was missing, the act of bicycle riding would not be successfully performed.

3.4 Multi-Agent System

In addition, bicycling is not possible if each of these agents simply exert their agency on their own. They need to be synchronized in such a way that makes bicycling possible. In other words, these various agents must constitute what we might call a “multi-agent system.” It is this multi-agent system as a whole that supports the somatic action of bicycle-riding in an integral way.

It is simple to determine what is, and what should not be considered an element of such a multi-agent system. If an agent is a necessary agent for bicycling, it should be included in the multi-agent system. Conversely, if an agent is not a necessary agent, it should not be included in it. Simply put, a multi-agent system is a system that consists of all necessary agents for a certain somatic action.

3.5 Multi-Agent System as the Necessary and Sufficient Agent

As a result, any somatic action can be performed without problems as long as a multi-agent system with all necessary agents is in place, while a multi-agent system must be in place for the action to be performed. The formation of a multi-agent system is the necessary and sufficient condition for a somatic action to take place – it is the “necessary and sufficient agent” for that action.

On the other hand, the individual “I” is a necessary agent for a somatic action, but not a sufficient one. In order for the action to take place, the agency of the “I” alone is not sufficient, and the help, support, or affordance of other agents is also necessary. In the end, the necessary and sufficient agent of the somatic action is not the “I,” but a multi-agent system that includes the “I.” This is a significant corollary derived from the first incapability thesis.

3.6 Fundamental Equality among Members of a Multi-Agent System

The role that individual agents in a multi-agent system play for the performance of somatic action and the mode and degree of their agencies are different for each agent. In the case of bicycling, the most important and central role must be played by the “I” on the bicycle. However, all agents are on an equal basis in the sense that they are necessary agents indispensable for its performance.

The most appropriate personal pronoun for such a multi-agent system is the plural first person nominative, i.e., “WE.” On the other hand, there are also multi-agent systems that do not include any “I,” which should be referred to by the impersonal (rather than the third personal) “THEY” – not “WE.”

3.7 Metaphysical Minimalism

There are multiple metaphysical options open for the ontological status of a multi-agent system as a “WE.” For example, it can be considered an abstract entity, or a mereological sum with each agent as a part. Some might admit its existence, while others do not. On the whole, the WE-turn attempts to propose the idea of a multi-agent system as a “guideline” for people to live their lives. Therefore, the WE-turn should preferably be compatible with any metaphysical standpoint that admits its function. Thus, it takes the position of “metaphysical minimalism,” an argumentative strategy that minimizes commitment to any particular metaphysical position.

The WE turn’s metaphysical minimalism is also demonstrated with respect to the concept of “agent” introduced earlier. An Aristotelian, for example, is committed to its reality, while a Humean is not. The WE-turn, however, is indifferent to the realism vs anti-realism debates concerning the agent at all. Even if it does not exist and therefore has no “ontological significance,” it is enough for the WE-turn if it has “explanatory power” to be useful for the discussion of somatic actions.

3.8 Subject of Somatic Action as a Necessary and Sufficient Agent

Next let us look at the second assumption, namely, the assertion that the subject of a somatic action must be a necessary and sufficient agent of that action, and not merely a necessary agent or a sufficient agent. This seems to be fairly obvious, so, as with the first assumption, let me demonstrate it with reference to an example, Japanese sumo wrestling.

Who is it that is sumo wrestling? The “subject” of sumo, which is a two-person combat sport, is, needless to say, a duo made up of two sumo wrestlers. It is true that each wrestler, a member of this duo, is a necessary agent who is indispensable for the performance of sumo. However, a wrestler cannot perform sumo by himself. What he can do by himself is called “one person sumo,” which is a comical performance that is different from the original martial art. A single wrestler is not a sufficient agent of sumo, but merely a necessary agent. And a mere necessary agent is not enough to be the subject of sumo.

Next let us consider a trio consisting of two sumo wrestlers and one spectator. If this trio exists, then there is no doubt that a sumo bout will be held. The trio constitutes a sufficient situation for the performance of sumo. On the other hand, it is not necessary for three people to be present for a sumo bout to take place. A duo consisting of two sumo wrestlers constitutes both the necessary and the sufficient conditions for sumo to be performed. This duo of necessary and sufficient agents, and only this duo, can be called the multi-agent system of the somatic action of sumo.

3.9 Non-Conscious Doer?

However, there is still one problem that needs to be cleared before one can really carry out the WE-turn of doer. The archetypal example of a doer is a human being. Humans are beings with consciousness and a mind (whether or not they are identified with the brain or other bodily mechanisms or functions). As such conscious beings, when performing a somatic action, humans often consciously have an intention for that action, and while performing the action, they have a sense of agency or a sense of being the doer in real time. Even if they do not have a clear intention or a sense of agency, they can always make their intention or sense of agency explicit if they want to.

On the other hand, “WE” as a multi-agent system, whether it is an “abstract entity,” a “sum,” a real object, or just a simple explanatory device, is an entity that does not and cannot have a mind or consciousness in principle. Can one think of such a non-conscious entity as the “doer,” instead of conscious entities as the model case?

3.10 Doer-Externalism

As mentioned above, the third assumption for the WE-turn of doer is that the doer or subject of a somatic action need not be always aware or conscious of the ones doing the action. This assertion can also be paraphrased as “the doer may exist ‘outside’ the realm of consciousness.” I will call this view “doer-externalism.”

The opposite of this doer-externalism is doer-internalism, that is, the claim that the doer must always be able to be conscious or aware that he or she is performing the somatic action, i.e., the doer must always be “within” the realm of consciousness.

3.11 How to Remedy “Externalism Allergy”

It is probably impossible to convince argument against those who axiomatically accept doer-internalism. We can, however, impose certain limitations on the doer-externalism in order to alleviate – if not completely eliminate – some of the objections that internalist may have against it.

Behind the internalist refusal of externalism, which extends the concept of doer to include non-conscious entities, there seems to be a legitimate fear that such an externalist extension would result in an excessive and unjustified conceptual expansion of both “consciousness” and “doer.” In the following, we will attempt to alleviate the internalist’s “externalist allergy” by placing a double constraint on externalism to prevent these unwarranted conceptual extensions.

3.12 Concerns Against Animism

Let us begin by looking at the restrictions that prevent undue extension of the “consciousness” concept. The internalist concern here can be paraphrased as the suspicion that extending the concept of the “doer”–whose paradigmatic example is conscious beings–to nonconscious entities may, in turn, lead to the implicit attribution of germinal forms of consciousness or mindedness to those nonconscious entities. Such an expansion of “consciousness” is similar to animism, which holds that even stones and old tools contain “spirits,” and would be seen by some internalists as an easy and unjustified “padding” or “cheapening” of the concept of “consciousness.”

3.13 Licklider’s Non-Animistic Externalism

In the face of these internalist concerns, we would like to first show that externalism without the smuggling of consciousness into non-conscious existence – let us call it “non-animistic externalism” – is possible, or at least has been proposed previously.

An example of non-animistic externalism can be found in the concept of “Man-Computer Symbiosis” formulated by J.C.R. Licklider (1915–90), also known as the “Father of the Internet,” in 1960.^[2] Like Norbert Wiener’s “cybernetics,” Licklider’s idea of a “symbiotic system” does not consider machines and their users, humans, separately, but posits a system composed of both of them, which itself is considered as an agent, and then tries to optimize its performance. In his view, “[t]he resulting partnership [between human brains and computers] will think as no human brain has ever thought and process data in a way not approached by the information-handling machines we know today.”^[3]

By “partnership” he means a “system” – in his words, a “symbiotic system” – consisting of humans and computers. In his vision of a future in which a “system” composed of humans and computers will outperform its constituent elements, both humans and computers, he sees the “system” itself (not humans or computers) as the subject of actions such as “to think” and “to process data.”

It is important to note that even though Licklider treated the “man-computer symbiosis” as an “actor,” he did not believe that a prototype of consciousness or mind resided in the “system” of the symbiosis. In Licklider’s idea, we can clearly read a “non-animistic externalism” that even a “system” without any form of consciousness can be a doer.

3.14 Techno-Externalism

Of course, the fact that Licklider adopted the idea of “systems as doers” without making the system conscious does not mean that his “non-animistic externalism” has been proven the only “correct” view of human–computer interaction, nor do we have to agree with his ideas just because his externalism is embedded in the design philosophy of computers widely used today. But Licklider’s case shows that externalism can also be recast as “techno-externalism,” so to speak, in the context of computer technology architecture, which does not involve the smuggling of “consciousness” into non-conscious entities, and which is far removed from animistic ideological genealogies.

3.15 Non-Conscious and Mindless Multi-Agent System

In light of this situation, we will add the following restriction to doer-externalism: “consciousness” must not be attributed to the non-conscious entity that is considered to be the doer. Under these restrictions, there is no attribution of consciousness or mind – prototypical or otherwise – to the multi-agent system as “WE.” The internalist need not fear the “unjustified expansion of consciousness” with respect to the WE-turn.

3.16 Overextension of Doer

Let us now address the second fear that internalists seem to have about externalism, namely, their wariness of the overextension of the concept of “doer.” If we adopt externalism, which unconditionally allows all non-conscious entities to qualify as “doers,” then any system that causes any consequential activity would be considered a “doer.” A weather system such as a typhoon would become a “doer,” and a storm caused by a typhoon would become a kind of “somatic action.” As a result, “somatic action” becomes a concept almost synonymous with “causal activity in general.” This is clearly an unjustified “padding” of the concept of “action.” The internalists are right to be concerned.

3.17 Necessary Inclusion of Humans

To dispel this second internalist fear, let us place a further restriction on externalism. This is, the restriction that the non-conscious entities that can be considered “doers” are limited to those systems that necessarily include conscious entities and humans in particular. This restriction means, for example, that a “stone” that is not such a system cannot be called a doer. Also, a typhoon can no longer be considered a doer, since it is not a system that necessarily includes a conscious being. On the other hand, Licklider’s “man-computer symbiosis,” as its name suggests, is a system that necessarily includes a conscious being, a human being, by definition, and thus continues to qualify as a doer under this limitation without any problem. The same can be said of a multi-agent system as a “WE.” A multi-agent system, which by definition always has a conscious being, an “I” as a member, also qualifies as a doer under this restriction. This restriction would at least somewhat alleviate the criticism that the concept of “doer” is inappropriately inflated in relation to externalism.

Consciousness-entity can include artificial imitations of life forms. Therefore, the possibility that a system that includes artificial entities such as robots and AI can be a doer is not eliminated under this restriction.

3.18 The WE-turn of Doer Derived

Now, we are finally ready to introduce the WE-turn of doer. First, as a corollary to the incapability of single action, we have obtained the assertion that only a “WE,” as a multi-agent system that includes an “I,” is the necessary and sufficient agent of somatic action. By adding the second assumption that only agents that are both necessary and sufficient are the subjects of physical action, or doers, we can conclude that “WE,” and not “I,” is the doer of somatic action. Furthermore, by adding a third assumption of “restricted doer-externalism,” which states that we can recognize non-conscious agents as doers, but only in systems that necessarily include conscious agents, we have been able to secure that “WE” as a multi-agent system is the doer, in a way that suppresses the objections of internalist to some extent.

This constitutes the shift from the “I” of the doer of a somatic action to a multi-agent system or “WE” that includes “I” – in other words, the WE-turn of doer.

3.19 Solitary Actor as a Self-Contradiction

From the perspective of the WE-turn of doer, all somatic actions are seen as collective actions performed by a “WE” of some kind. As an indispensable member of “WE,” “I” is always participating in the collective actions of “WE.” “I” is always acting as a member of the “WE,” supported, helped, and sustained by the other members of “WE” as well as the “WE” itself.

Conversely, considering that you are acting and living on your own is nothing more than a form of myopic hubris. Every time you act, an invisible “WE” is actually standing by you and supporting you. You just cannot see it because you think you are acting and living on your own.

This means that you or I are never truly alone, or that we cannot be alone in the first place. Even if someone feels that they are all alone in the world with no friends or family, as long as that person is alive and performing somatic actions, there are a number of necessary agents supporting those actions, and a “WE” made up of all those agents is standing around that person. The “solitary actor” is a contradiction in terms.

3.20 Naked I as an Illusion

As long as “I” continues to live and act, “I” will always be with many instances of “WE” of different kinds, which are constantly appearing and disappearing one after another. A “WE” has been with the “I” from the beginning, and the “I” cannot escape from it. For the “I,” a “WE” is not a removable costume. Conversely, there is no such thing as a “naked I” that appears after having taken off all the “WE” costumes. After having stripped off all the “WE” there is nothing left.

3.21 Mutual Indispensability between “I” and “WE”

In this sense, “WE” is indispensable for the “I” in that the “I” cannot exist without a “WE.” On the other hand, the “I” is also indispensable for “WE.” As mentioned earlier, if the “I” were not there, the “WE” would cease to be “WE.” A multi-agent system would not be “WE,” but “THEY.” In the end, “I” and “WE” are both indispensable for each other. There is a mutually-dependent inescapable relationship between the two.

3.22 “I” is Indispensable Because “I” is Incapable

“WE,” for which the “I” is indispensable, comes into existence because the “I” is characterized by the fundamental incapability of solitary action. It is precisely because the “I” is a “incapable being” rather than a “capable being” that “I” has an indispensability for “WE” that will never waver. In the WE-turn, the indispensability of the “I” is based on “I’s” incapability rather than capability. “I” is an indispensable being precisely because “I” is incapable.

3.23 Incapability that is Open to “WE”

The word “incapability” has a negative meaning, referring to the lack of capability. However, the fundamental incapability of solitary action tells us that any “I” is always a member of a “WE” and supported by its other members, and that “I” is indispensable for the “WE.” It is the incapability that reveals us that “I” is open to “WE.” Shortly it is the incapability that is open to “WE.” In this sense, the fundamental incapability has a positive significance.

4.1 Mental Activity as a Somatic Action

With additional assumptions, the WE-turn of doer triggers further WE-turns. Let us now address the WE-turn of mental activities such as thinking and decision making.

This WE-turn takes mental activities such as thinking, intending, and decision-making as kinds of somatic actions for the reason that for all of them, systemic somatic activities or conditions are indispensable, and therefore, the implications of the WE-turn of the doer also apply for these kinds of actions, e.g., thinking. The upshot is the WE-turn of thinking, or the shift from “I think (cogito)” to “We think (cogitamus).”

4.2 Not Mere Influencers but Equal Members of a Thinker as a “WE”

It is widely acknowledged that thinking is not immune from the influence of many factors, but rather always affected by those factors such as the opinions of others, information from mass or social media, social and cultural values, the language in terms of which it is framed, and so on. There is no thinking that occurs in a vacuum.

But many have been taking only “I” as the thinker while dismissing other factors as mere influencers to this unique thinker. The WE-turn challenges such a standard idea, claiming that “I” and other factors are equal thinkers, or more precisely, equal members of the thinker that is a multi-agent system or “WE.”

4.3 “I” as the Only Necessary Internal Agent for Thinking

However, the “I” is unique among all members of the thinking “WE.” Whenever “I” pertains to “WE’s” thinking, it always and invariably knows, in real-time, that it is doing so. Let us call the agent that can be aware of its participating in a somatic action the “internal agent.” The “I” is an indispensable internal agent for thinking because it always and necessarily remains internal. In contrast, no other member of the thinking “WE” has such a necessary or inevitable internal agency. The “I” thus has a monopoly on the status of the indispensable and necessary internal agent.

4.4 Confusion of the Unique Necessary Internal Agent with the Thinker

Though the “I” is the only necessary internal agent of thinking, on the present view the “I” cannot be its unique subject. There is a difference between being the only necessary internal agent of thinking and being the subject of thinking. (Here the subject means a special type of agent, i.e., the necessary and sufficient agent.) The view that takes “I” as the subject of thinking or the unique thinker may result from the confusion of the only necessary internal agent of thinking with the subject of thinking.

4.5 The WE-turn of Decision Making

The same can be said of decision-making. Just like the case of thinking, the WE-turn of decision-making from “I decide” to “We decide” can be seen as a special version of the WE-turn of doer. All decisions are taken to be collective decisions made by “WE.” Single decision-making by the “I” is impossible on this view, and the “I” no longer has the monopoly for the capacity to decide.

4.6 “I” as the Only Necessary Internal Agent for Decision Making

However, the “I” is still in a unique position in decision-making in comparison to other members of “WE”, for two reasons. First, as in the case of thinking, the “I” is the only necessary internal agent. The “I,” and only the “I,” is always and necessarily aware, in real-time, that it is participating in the decision-making process. Furthermore, in decision-making, the “I” has another unique status of being the only permanent decision-maker whose presence is required throughout the process. Even though all decisions are made collectively by “WE,” not all members of “WE” are present at the final stage of the decision. The “I,” on the other hand, is always a final decision-maker. However, the “I” is not always the only final decision maker. There can be cases where both an “I” and a “you” are finalizing the decision, for instance. However, the “I” is always one of the final decision-makers, while the “you” and any other agents need not be present at this stage. Only the “I” always has the final say.

However, the “I” does not have a monopoly on decision-making authority. It is nonetheless usual to confuse this necessary presence at the stage of the final decision with a monopoly of decision-making authority as such.

6.1 Generalized First Incapability Thesis

The “I” as a member of the multi-agent system is responsible for a certain part of any act of the “WE,” that of the bicycle rider, for instance. Can the “I” perform its part of this action alone? The WE-turn’s answer is again “no.” The “I,” just as any other agent in such a multi-agent system, cannot perform their partial action alone, just by themselves. This follows from the first incapability thesis that is extended to cover any somatic actions, partial or not.

6.2 First Layer of “I”

Many agents should participate even in the “I’s” partial action. Some of them are within the “I’s” body: the intelligence, sensory perceptions, the central nervous system, sensory organs, muscles, bones, and so on. The others are outside the body: the “I’s” biological ancestors, oxygen in the air, and so on. Those inside and outside agents constitute a smaller multi-agent system for the partial action that the “I” is responsible for.

This smaller multi-agent system cannot be called a “WE” because it does not contain a separate “I” any longer. Its members such as the intelligence and the muscles are not the “I,” but mere parts of the “I.” All members of this smaller system are also members of the larger multi-agent system for bicycling. So, the former is a proper subset or a subsystem of the latter.

The same is the case for even a smaller agent, say the muscles. As a result, a nested structure of a multi-agent system and its subsystems takes shape.

The multi-agent subsystem for the rider’s partial action is the first and the most fundamental layer of the “I” or I₁, upon which other layers are embedded. This first layer has a distinct property of we-ness. This we-ness has two components. The first is plurality: it always and necessarily has multiple members both inside and outside the body. The second is the vagueness of its boundaries. Any multi-agent system or its subsystem should have a boundary. They do not extend indefinitely. But no one can clearly determine where these boundaries are. Any attempt to forcibly define a boundary can always be undermined by counterexamples just outside the drawn line. The WE’s boundaries are thus ontologically determined but epistemologically vague.

6.3 Consciousness as a Bundle of Two Vectors

The second layer of “I,” or “I₂,” is constituted by the “vectors of consciousness.” In the WE-turn, consciousness is seen as a bundle of two vectors, or two arrows, called the “active vector” and the “passive vector.” Each vector has a certain direction and is a two-dimensional structure with two poles, the starting point and the terminal point. (The starting point and the terminal point here correspond to, roughly, noesis and noema in Husserl’s phenomenology, respectively.) While “I₂” is the starting point from which the “active vector” heads toward the object, it is also the terminal point towards which the object heads in the “passive vector.”

6.4 Aching Me

The prototype, archetype or paradigm of “consciousness” with “I₂” with its poles is somatic pain. The “I₂” is both “the endpoint towards which pain is directed” and “the starting point from which it is directed towards the object via pain.” The former can be called the “patient of pain,” and the latter the “agent of pain.” The archetypal metaphor for “I₂” is “aching me” as an “agent/patient of pain.”

6.5 Evolutionary Acquirement of Pain

Somatic pain can be thought of as a bodily self-monitor with an alarm function that triggers actions to quickly and effectively escape from external environments that are harmful to the body (such as predators). “Aching me” and, by extension, “I₂” in general are “self-monitoring representations of the body in consciousness.”

This kind of “body awareness” represented by the monitor function of “pain” and the accompanying “aching me” have been effective for the preservation of individuals and species, and so it can be said that it is a bodily mechanism that certain living things – including cephalopods such as squid and octopus, and vertebrates such as humans – have acquired over the course of evolution. In other words, it can be said that not only humans, but also squids and octopuses, among others, have “I₂” – or at least a prototype of it.

6.6 Singularity and Distinctness

The physical pain sensation, which has the function of inducing a quick escape from a harmful external environment, represents the “body” as a singular entity, clearly and sharply differentiated from that environment. The “aching me” thus presumes a representation of the body based on the singularity of the subject that is clearly distinct from the external environment. We can assume, on the one hand, that this singularity is a characteristic of internal representation that contributes to controlling the body as a whole and enabling it to act quickly. On the other hand, the clear distinction between the body and the environment is a characteristic of internal representation that helps to identify harmful elements in that environment and to quickly react to them.

6.7 Extended Body/Mind vs Multi-Agent System

There are various ways to theorize about the range of the body of which the “aching me” is the internal representation. It has also been suggested that this body can be seen as extending beyond the biological body into the part of the environment that is connected to the body (the extended mind theory). However, this extended “I₂” or “extended body” mostly involves the vicinity of the physical body.

On the other hand, the multi-agent system of the WE-turn extends far beyond the immediate neighborhood of the body in space and time. More importantly, while multi-agent systems definitely do have a certain boundary, we cannot clearly draw that boundary. As argued above, the boundary of a multi-agent system is determined ontologically, but indeterminate epistemologically. Multi-agent systems have indeterminate boundaries, in other words, the “WE” is simultaneously vague and distinct.

6.8 “I-ness” as a Natural Illusion

On the other hand, the singularity and clear distinctiveness of the “aching me” and “I₂” in general – let us call these “I-ness” – are in sharp contrast with the “we-ness” of “I₁” – that is, its “plural nature” and vague distinctiveness. As a result, the “I-ness” conceals the “we-ness.” (Of course, “aching me” and “I₂” in general can have vague boundaries as well, but these are still far clearer than the vague boundary of the “WE.”)

In the sense that it conceals this fundamental fact of “we-ness,” the “I-ness” of “aching me” and “I₂” can be said to be a kind of “illusion.” However, this “illusion” is an unavoidable one that has been acquired evolutionarily, because it is effective for the self-preservation of both single individuals and the whole species. In that sense, it can be called a “natural illusion” in analogy with the Humean term “natural belief.”

6.9 Third Layer of “I”

Like “I₂,” “I₃” is also an “intra-conscious entity.” However, while “I₂” is essentially limited to a single consciousness – for example, a single sense of pain – “I₃” is not restricted to a single consciousness, but takes itself as extending over time beyond a single consciousness and across multiple consciousnesses. It is the conscious representation of the body as something that straddles a single consciousness.

6.10 Depersonalization

The reason for setting “I₃” as the being that transcends a single consciousness and distinguishing it from “I₂” is that the state of having “I₂” but lacking “I₃” can be observed in certain types of psychopathological symptoms, which are called the “depersonalization-derealization disorder.” Patients in such a state can feel their “I” and “the world,” but, in this condition, the “I” and “the world” are perceived and experienced as separate entities that have lost their coherence and meaning. In terms of the WE-turn, such symptoms represent a situation in which the “I₂” is functioning but the “I₃” has become dysfunctional for some reason.

6.11 Fourth Layer of “I”

While “I₂” and “I₃” are both layers of the internal consciousness, the “I₄” is not an element of consciousness. It is the being that is “constructed” and “fictionalized” as the “leading character of the I-narrative” in the interaction of “I₁” and “I₃.” The “I-narrative” here typically refers to a “story” in which “I” is both the protagonist and the narrator. In this story, the “I” is depicted as facing numerous difficulties and overcoming them to continue moving toward certain goals. The “I₄” is a being that has been constructed or fictionalized as the leading character of such an “I-narrative” or “I-story.”

It is true that this fictional entity is not always consciously perceived, but it is functional as an entity that continues to exist even if it is not consciously perceived. In this sense, “I₄” is an extra-conscious entity. Typically, “I-stories” like this are spun linguistically, but it is also possible to imagine non-linguistic, or proto-linguistic narratives that feature a form of such an I-fiction.

6.12 Evolutionary Acquisition of “I-narrative”

Such “I-narratives” are currently only possessed by a limited number of living things, who are able to feel physical pain in the sense discussed above. They are “natural” narratives that have been acquired evolutionarily because of their efficiency in preserving individuals and species. However, the WE-turn does not exclude the possibility that artifacts will also be able to develop such “I-narratives” in future.

6.13 Different “I-narratives”

It is also possible for different patterns of “I-narratives” to be constructed and shared in various societies. For example, it is possible to experience “death and rebirth” before and after the initiation rites for joining an adult group, and to live a different life before and after that. In that sense, the “I-narrative” here and “I₄” as the protagonist of that narrative can also be social constructs, at least in part. Such narratives are thus both natural and social at the same time.

6.14 “I” as a Four-Layered Structure

Finally, let us summarize the relationships between these four layers. The foundation of the “four-layered self” is a subsystem of a multi-agent system as its first layer with “we-ness” in the form of an intrinsic multiplicity with vague boundaries. The “second layer” is a self-monitoring conscious representation with “I-ness” (which sharply opposes “we-ness”) in the form of singular individuality with clear boundaries. The “third layer” is constituted by the awareness that this “second layer” extends beyond a single consciousness. And the “fourth layer,” which is the protagonist of the I-narrative, is constructed in the interaction of these first three layers.

The idea of the “I” as the four-layered structure described above fits into the WE-turn framework in all respects. However, this is not the only view of “I” that is compatible with the WE-turn – the relations of the “I” and the “WE” can also be modeled in a variety of different ways.