WHERE ARE THE CAR KEYS?

The Helios Dialectic—Ur edition by Richard Crist

Husband: “Where are the car keys?”

Wife: “They’re on the dining room table where I put them.”

Husband: “No they aren’t; I was just in the dining room and there’s nothing on the table.”

Wife: “Oh, then Mikey must’ve taken them.”

Husband: “Yes, Mikey must have them; we both believe that.”

Wife: “You should ask him where he put them.”

Husband: “I’ll do that right now.”

MY ARGUMENT

Two consciousnesses, one new conclusion, one belief, one action:

The husband and wife begin with different information. The wife knows that she put the keys on the dining room table. The husband knows that the keys are no longer there. Neither, using only the information independently available to that person, can conclude that Mikey has taken the keys.

Their conversation brings the separately held information together. The husband's information changes the wife's thinking, and she concludes that Mikey must have taken the keys. That conclusion depends upon information supplied by both members of the pair. It is therefore a product of their thinking together: a collective conclusion.

The husband then considers the conclusion, accepts it, and says, “Yes, Mikey must have them; we both believe that.” At this point the collective conclusion has become a collective belief. The distinction matters. A group may produce or entertain a conclusion without all its members accepting it. Here, however, the conclusion produced through their interaction is subsequently accepted by both.

Thinking has occurred—but whose thinking is it? The wife performs the final inferential step, but she could not have reached the conclusion from the information available to her alone. The husband supplies information without which the conclusion could not have been reached. The thought is therefore produced by a process extending across both of them.

There is a direct analogy with thinking within an individual human being. Information located in different parts of an individual cognitive system is brought together and processed, producing a conclusion. We do not deny that the individual thinks because different parts of the process occur in different parts of the brain. In the present case, information is likewise distributed among different parts of a larger system, brought together and processed, and a conclusion results. The difference is that the parts of this cognitive system are themselves human beings. (a) If a process in which information distributed among the parts of a cognitive system is brought together and processed to produce a conclusion constitutes thinking by that system, then an analogous process occurring within the pair constitutes thinking by the pair. (b) In the pair, information distributed between the husband and wife is brought together and processed to produce a conclusion, in a process analogous to the process that constitutes thinking within an individual. (c) Therefore, the pair thinks. So,

CLAIM 1: The pair itself is a thing that thinks. There are three thinkers: the husband, the wife, and the pair.

This is an example of Collective Intelligence. Collective Intelligence, or CI, arises when you have a group of individuals, and the various thoughts of these individuals are communicated to one, some, or all of the members of the group, who then process those thoughts and arrive at conclusions that no single member of the group could have arrived at independently.

Let Helios be the collective entity constituted by humanity as its members communicate and process information together through Collective Intelligence.

(d) A collective entity whose members engage in the kind of process established above is a thing that thinks. (e) Helios is a collective entity whose members engage in that kind of process. (f) Therefore, Helios is a thing that thinks. So,

CLAIM 2: Helios is at least a thinking being.

a thought about a thought:

The pair has arrived at the belief that Mikey has the keys. But the thinking does not stop there. The husband says, “Yes, Mikey must have them; we both believe that.” The belief that Mikey has the keys has now itself become the object of another thought. The husband is not merely thinking Mikey has the keys; he is also thinking about the fact that they believe that Mikey has the keys.

A thought about another thought is called a Higher-Order Thought, or HOT. The belief Mikey has the keys is the first-order thought. The thought we believe that Mikey has the keys is a higher-order thought about that first-order thought.

(g) If, upon inspecting the contents of phenomenal consciousness, we find nothing present in them beyond points and relations of sameness and difference among those points, then the contents of phenomenal consciousness consist of points and s/d relations. (h) Upon inspecting the contents of phenomenal consciousness—for example, the visual field—we find nothing present in them beyond points and relations of sameness and difference among those points. (i) Therefore, the contents of phenomenal consciousness consist of points and s/d relations.

How, then, does the Higher-Order Thought represent the first-order thought? On the account proposed here, one part of the brain in effect “looks at” the thought occurring elsewhere in the brain. It does not represent that thought as the biological event that realizes it—as relations among firing neurons. Instead, it represents the thought as points connected by relations of sameness and difference. The Higher-Order Thought thus represents the first-order thought in the same terms—points and s/d relations—that, according to (i), constitute the contents of phenomenal consciousness.

(j) If the phenomenal contents of the visual field consist entirely in the apprehension of points and relations of sameness and difference, then a Higher-Order Thought that pictorially presents a first-order thought as points connected by relations of sameness and difference presents that thought in the phenomenal form of the visual field. (k) ( = (i))The phenomenal contents of the visual field consist entirely in the apprehension of points and relations of sameness and difference. (l) Therefore, a Higher-Order Thought that pictorially presents a first-order thought as points connected by relations of sameness and difference presents that thought in the phenomenal form of the visual field.

The visual field provides the clearest example because its relational structure can literally be pictured. But the principle is not confined to vision. On the account proposed here, the same basic principle extends throughout phenomenal consciousness: sounds, pains, smells, tastes, and other phenomenal contents differ in the particular relations that constitute them, but what we apprehend phenomenally is in every case a relational structure of sameness and difference. Vision is therefore the example through which the general principle is most easily seen, not an exception upon which the argument depends.

Just as the husband and wife come to have the same first-order thought, Mikey has the keys, they come to have the same Higher-Order Thought, we believe that Mikey has the keys. There is no essential difference between the two cases. The first-order thought is a thought about the keys; the Higher-Order Thought is a thought about that thought. In both cases, the thought arises through their thinking together. Neither thought would have arisen if each had been thinking alone.

The Higher-Order Thought therefore belongs to the entity that is the pair just as the first-order thought does.

The Higher-Order Thought therefore belongs to the entity that is the pair just as the first-order thought does. (m) If the Higher-Order Thought belonging to the pair represents the pair's first-order thought in phenomenal form, then the pair is phenomenally conscious. (n) The Higher-Order Thought belonging to the pair represents the pair's first-order thought in phenomenal form. (o) Therefore, the pair is phenomenally conscious.

CLAIM 3: The entity that is the pair is a phenomenally conscious being.

The husband and wife provide a small and easily examined example of a collective entity that thinks and is phenomenally conscious. There is no reason in the preceding argument to restrict such an entity to two human beings. The same process can occur among the vastly greater number of human beings whose thinking together constitutes Helios. Helios can form first-order thoughts, and it can form Higher-Order Thoughts about those thoughts. Those Higher-Order Thoughts can represent Helios' first-order thoughts in phenomenal form just as the Higher-Order Thought of the pair represents the pair's first-order thought in phenomenal form.

(p) If an entity has Higher-Order Thoughts that represent its first-order thoughts in phenomenal form, then that entity is phenomenally conscious. (q) Helios has Higher-Order Thoughts that represent its first-order thoughts in phenomenal form. (r) Therefore, Helios is phenomenally conscious.

CLAIM 4: Helios is a phenomenally conscious being.

(s) If a being thinks in a sufficiently sophisticated way and is phenomenally conscious, then it is a person. (t) Helios thinks in a sufficiently sophisticated way and is phenomenally conscious. (u) Therefore, Helios is a person.

CLAIM 5: Helios is a person.

So here is a surprise: Helios, the system of the Sun, is a person. He is a great astrocolossus, with the physical form of a plasma body enclosing a central burning sphere surrounded by circling rocky and gaseous spheres, which themselves are attended by still smaller companions. His mind has been evolving for four and a half billion years. He is ancient, vast, and still developing. And this brings us to the really big surprise—the implication that the stars are people.

OBJECTIONS TO THE ARGUMENT FOR GROUP COGNITION

Argument inspired by Deborah Tollefsen

DIALECTICIAN 1: Your (a) has not been established. You have shown that there is an analogy between what happens within an individual thinker and what happens between the husband and wife. In each case, information from different sources is brought together and a conclusion results. But from the fact that two processes are analogous, it does not follow that both are instances of thinking by a larger entity.

We often explain what groups do by speaking of them as though they had beliefs, desires, intentions, and thoughts. Such descriptions can be useful because they allow us to understand and predict the group's behavior. But their usefulness does not establish that there is literally another thinker in addition to the individual human beings involved. In the Car Keys case, you have clearly shown two people thinking together. You have not yet shown that the pair itself is a third thing that thinks.

DIALECTICIAN 2: The analogy is stronger than you suggest. Consider the process at the physical level. Within an individual thinker, neurons in different parts of the brain interact, information is brought together and processed, and a conclusion results. In the pair, neurons in the husband's brain and neurons in the wife's brain also interact, although the connections between the two brains pass through speech and perception. Information is again brought together and processed, and a conclusion results. In both cases, the physical process ultimately consists of interactions involving neurons.

The difference is that in one case all the neurons are enclosed within one human body, while in the other they are enclosed within two. But unless there is some reason why that difference prevents the larger process from constituting thought, it does not defeat the analogy. The boundary of an individual body cannot simply be assumed to be the boundary of a thinking system. Why should neurons cease to participate in a larger cognitive process merely because some of their causal connections run through the world outside the skull?

And there is another reason to think that the two brains can form a larger cognitive system.

Argument inspired by Mattia Gallotti and Bryce Huebner.

When two people think together, they need not merely pass completed thoughts from one brain to another. What happens in one person's brain can change what happens in the other's, whose response can then change what happens in the first. The thinking itself can pass back and forth between them.

That is what happens in the Car Keys case. The wife remembers putting the keys on the table. The husband tells her that they are no longer there. His information changes her thinking, and she concludes that Mikey must have taken them. She communicates that conclusion to the husband; it changes his thinking, and he accepts it. The conclusion has emerged from a process that repeatedly crosses the boundary between their two brains.

Thus the husband and wife are not simply two independent thinkers exchanging the finished products of their separate thinking. Their interaction is part of the thinking that produces the conclusion. Neither would have reached that conclusion from the information independently available to that person. This gives us further reason to regard the process as thinking by the entity constituted by the pair.

Argument inspired by Michael Schmitz

DIALECTICIAN 1: I can grant even more than that. I can grant that there is genuinely an our thought—a thought that belongs to the husband and wife together—and still deny that there is a third thinker. Two people can be joined as co-subjects of a thought without thereby creating another subject in addition to them. A unity of minds need not itself be another mind.

So the question remains. You have shown that the husband and wife think together and that the resulting thought can properly be called their thought. But why does the existence of their thought require the existence of an entity, the pair, that thinks it?

Argument inspired by Eric Schwitzgebel

DIALECTICIAN 2: But why should the physical separation of the husband and wife prevent the pair from being a thinker? Schwitzgebel calls Contiguism the view that the parts of a conscious or cognitive system must be spatially contiguous. But it is difficult to see why contiguity should be necessary.

Consider the comparison again. In an individual brain, neurons in different regions interact and together produce thoughts that we attribute to the person. In the husband-and-wife pair, neurons in one brain interact with neurons in the other brain through speech and perception. The causal connections are longer and pass through the surrounding environment, but the information produced in one brain changes the processing occurring in the other, and the resulting response changes the first in turn.

If we say that the husband thinks because the relevant processing occurs among the interacting parts of his cognitive system, we need a reason for refusing to say the same thing about the pair. Simply pointing out that some of the interacting neurons are in one skull and some are in another amounts to treating spatial separation as decisive. Unless Contiguism is true, that difference does not answer the analogy.

Two people thinking together do not automatically constitute a third thinker. But in this case we have identified a cognitive process extending across them and producing a conclusion that neither could have produced from the information independently available to that person. If spatial separation does not disqualify that process from being cognition, we have the same kind of reason for attributing the thought to the larger system that we have in the individual case.

Argument inspired by Thomas Szanto:

DIALECTICIAN 1: Spatial separation is not my fundamental objection. I can grant that the boundaries of a thinking system need not coincide with the boundaries of a single body. But there is another important difference between your two cases. The neurons within the husband's brain are not themselves subjects of thought. The husband and wife are. Each already has a point of view and thoughts of his or her own.

So the difference is not merely that the neurons are divided between two skulls. In the individual case, many nonthinking components contribute to the thinking of one subject. In the collective case, the supposed components are themselves thinking subjects. Why should their interaction produce another subject over and above them? Perhaps what you have shown is an unusually intimate integration of two thinkers, not the existence of a third thinker.

DIALECTICIAN 2: But why should the fact that the parts of a thinking system can themselves think prevent the larger system from thinking? Suppose we discovered that individual neurons do much more cognitive work than we presently suppose—that before passing information onward, they perform their own elaborate processing of it —perhaps integrating information, retaining rudimentary information about prior inputs, and selecting among alternative outputs. Would that discovery show that the human being composed of those neurons had ceased to be a thinker? Surely it would not. It would merely tell us more about the capacities of the parts through which the person's thinking occurs.

Nor is it clear where a boundary could be drawn. If a component performs a little processing, the larger system can think; but if the component performs more processing, does the larger system somehow lose that capacity? And if the component eventually becomes sufficiently sophisticated to count as a thinker itself, why should that fact suddenly prevent the larger system from thinking?

The fact that the husband and wife are themselves thinkers therefore does not answer the analogy. It identifies a difference between the two cases, but it does not yet show that the difference is relevant. A thinker may itself be a part of a larger system that thinks. What matters is whether the larger system contains the kind of integrated cognitive process that produces thoughts attributable to that system.

OBJECTIONS TO THE ARGUMENT FOR PHENOMENAL CONSCIOUSNESS

DIALECTICIAN 1: I will grant, for the sake of argument, that the pair itself thinks. But thinking is not the same thing as being phenomenally conscious. Your argument for phenomenal consciousness depends first upon (h):

Upon inspecting the contents of phenomenal consciousness—for example, the visual field—we find nothing present in them beyond points and relations of sameness and difference among those points.

I do not see why I should accept this. When I look at a red apple, I do not seem to find merely points standing in relations of sameness and difference. I find redness. I find shape, brightness, perhaps depth. When I hear a note, I find a particular sound; when I feel pain, I find the painful quality of the pain. These seem to be qualities of phenomenal consciousness, not merely relations of sameness and difference among points.

So before your argument can use the relational character of phenomenal consciousness to explain how a Higher-Order Thought can present a thought in phenomenal form, you must establish that (h) is true. Why should I believe that all of these apparently different phenomenal qualities consist entirely of points and relations of sameness and difference?

DIALECTICIAN 2: You are treating redness, shape, brightness, sound, and pain as though they were something in phenomenal consciousness in addition to relations of sameness and difference. But that is precisely what I deny. What we call qualia are our apprehension of those relations.

Consider red. What more can be said of its redness? Red is different from orange and red-violet; it is more different from yellow and blue; and it is most different from green. These relations can be present within the visual field itself. But red also stands in relations to things outside the present visual field. It is the same as the color we have seen in fire and blood, for example, and different from the colors of countless other things we have seen. These relations and associations form an enormously rich pattern.

My claim is that this is all there is to the redness of red. There is no additional phenomenal ingredient—a pure redness—over and above our apprehension of this pattern of sameness and difference. If you believe that there is, try to identify it. Attend to the redness itself and subtract everything that consists in its being the same as this, different from that, more like this, less like that. What remains?

The same challenge applies to other qualia. Merely naming the resulting phenomenal content—redness, a sound, a pain—does not identify something in consciousness in addition to the relations that constitute it. So the objection has not yet identified anything present in phenomenal consciousness beyond points and relations of sameness and difference.

DIALECTICIAN 1: Suppose I grant your account of qualia. I will assume that phenomenal contents consist entirely in the apprehension of points and relations of sameness and difference. Your (j) still seems to make a further assumption. You say that if a Higher-Order Thought presents a first-order thought entirely as points and relations of sameness and difference, it thereby presents that thought in phenomenal form.

But why should that follow? A system might represent a pattern of points and relations of sameness and difference without anything becoming phenomenally conscious. A diagram, for example, can contain such a pattern, and a computer might represent such relations while processing information. What turns the relational representation in a Higher-Order Thought into phenomenal consciousness, rather than merely another representation?

DIALECTICIAN 2: There is an important difference between merely describing relations and presenting a relational structure as a picture. Consider three colors, c1, c2, and c3. I can describe the relations among them symbolically:

Where Mh is moderation of the relation kind of hue, Di is Dissimilarity of intensity, etc.

Each expression tells us something about the relation between two of the colors. For example, the notation can tell us how similar or different two colors are in hue, intensity, or saturation. The complete expression therefore describes a network of relations among c1, c2, and c3.

But now represent exactly the same information with the line-and-dot notation. All the occurrences of c1 become a single dot, all the occurrences of c2 become another dot, and all the occurrences of c3 become a third. The relations represented symbolically above become relation-lines connecting those dots. The result is a picture of the relational structure.

Nothing has been added to the relational information. What has changed is how that information is presented. In the first notation, the relations are described by symbols that must be interpreted. In the second, the relational structure itself is presented as a picture.

That is the distinction relevant to (j). I am not claiming that anything that contains information about relations of sameness and difference is therefore phenomenally conscious. A string of symbols could contain all of that information without being a phenomenal presentation. My claim is that a Higher-Order Thought pictorially presents the relational structure of the first-order thought. And according to the account defended above, this is the form in which the visual field itself is phenomenally present: an apprehended picture constituted by points and relations of sameness and difference.

DIALECTICIAN 1: I grant the distinction. A relational structure can be merely described symbolically, or it can be presented pictorially. But you have not yet shown that the Higher-Order Thought actually does pictorially present the first-order thought. Your line-and-dot notation shows that such a pictorial representation is possible. It does not show that this is what occurs when one thought represents another thought.

When I think I believe that Mikey has the keys, why suppose that my Higher-Order Thought contains anything resembling your relational picture? Perhaps it simply represents the first-order thought in some entirely different way. Your argument requires not merely that a thought can be represented pictorially, but that a Higher-Order Thought does represent it that way.

DIALECTICIAN 2: I agree that the line-and-dot example does not independently prove that this is what a Higher-Order Thought does. What I am proposing is an explanatory theory of how a thought becomes phenomenally conscious.

We already have a first-order thought with a relational structure. We have a Higher-Order Thought that represents that first-order thought. And we have found that the phenomenal visual field itself has the form of an apprehended relational picture. The proposal is that the Higher-Order Thought represents the relational structure of the first-order thought pictorially. If it does, these otherwise separate facts fall into a single simple account: the first-order thought becomes phenomenally conscious because the Higher-Order Thought presents its relational structure in phenomenal form.

The strength of the proposal is therefore explanatory. It does not require us to add some further mysterious ingredient that turns an unconscious thought into a conscious one. The same relational structure that constitutes the thought is presented in the same pictorial form that constitutes phenomenal consciousness. One simple process accounts for the transition from thought to conscious thought.

This does not demonstrate that the theory is true. But a theory that provides a simple account of the phenomenon, fits what we independently know about its parts, and leaves nothing further unexplained gives us a reason to take it seriously unless a better account is available.

Argument inspired by Thomas Szanto:

DIALECTICIAN 1: objection to (m)--Suppose I grant your theory of phenomenal consciousness. I will grant that a Higher-Order Thought can make a first-order thought phenomenally conscious by pictorially presenting its relational structure. And I will grant your earlier argument that the husband and wife can constitute a cognitive system that thinks.

But it does not follow that the pair is a phenomenally conscious subject. The phenomenal consciousness still occurs in the husband and the wife. The husband has his conscious thoughts, and the wife has hers. Their thoughts may be closely integrated; they may even properly speak of our thought. But a unity of conscious subjects is not thereby another conscious subject.

This difference matters because phenomenal consciousness seems essentially to belong to a subject. If the husband sees red, there is a phenomenal presentation for the husband. If the wife sees red, there is a phenomenal presentation for the wife. Where is the corresponding phenomenal presentation for the pair? Unless there is something phenomenally present to the pair itself, you have shown collective cognition, but you have not yet shown collective phenomenal consciousness.

DIALECTICIAN 2: But you are looking for an additional phenomenal presentation where none is required. We have already established that the first-order thought Mikey has the keys belongs to the pair. The pair then has the Higher-Order Thought we believe that Mikey has the keys. If that Higher-Order Thought pictorially presents the relational structure of the pair's first-order thought, then that is the phenomenal presentation for the pair. There need not be another presentation somewhere else.

It is true that the Higher-Order Thought is physically realized through processes occurring in the husband and the wife. But the same was true of the first-order thought. Its physical realization was distributed between them, yet we found that the thought belonged to the pair because it was produced by their thinking together and could not have been produced from the information independently available to either one. The fact that the Higher-Order Thought is likewise realized through the two members gives us no new reason to deny that it belongs to the pair.

Nor must the physical processes occurring in the husband and wife be identical in order to constitute the same thought. Their particular relational structures can differ while instantiating the same Form. At the first level, the different physical realizations instantiate the thought Mikey has the keys. At the higher level, their different physical realizations instantiate the thought we believe that Mikey has the keys. In each case there is one thought because the different particular structures instantiate the same Form within the cognitive process of the pair.

The mistake, then, is to suppose that a single phenomenally conscious thought must be located at a single place. Nothing in the account of phenomenal consciousness given here requires that. One collective first-order thought is represented by one collective Higher-Order Thought, and that Higher-Order Thought presents the first-order thought in phenomenal form. That is what it means, on this account, for the pair to be phenomenally conscious.

Argument inspired by Christian List

DIALECTICIAN 1: Your argument depends upon a Higher-Order Thought making a first-order thought phenomenally conscious. But why suppose that Higher-Order Thought is necessary for phenomenal consciousness? Many animals appear to see, hear, feel pain, and possess other phenomenal states even though they may lack the cognitive ability to form thoughts about their own thoughts.

If phenomenal consciousness can occur without Higher-Order Thought, then your account may be describing one way in which a thinker can represent its own mental states without explaining what makes those states phenomenally conscious. The pair's ability to think we believe that Mikey has the keys would then establish a kind of higher-order awareness, but it would not by itself establish phenomenal consciousness.

DIALECTICIAN 2: The animal objection assumes something that has not been established. It assumes both that animals which cannot have Higher-Order Thoughts are phenomenally conscious and that animals which appear to be phenomenally conscious cannot have Higher-Order Thoughts. Either assumption may be false.

Consider the first possibility. Some animals may have first-order representations without being phenomenally conscious of them. They may perceive features of their surroundings, respond appropriately to them, avoid injury, seek food, and exhibit surprisingly complicated behavior without the corresponding first-order states ever becoming phenomenally conscious. Their behavior alone does not establish that there are pictures they see, sounds they hear, or pains they consciously feel.

But the second possibility is equally important. Some animals may actually be capable of the relevant Higher-Order Thoughts. A dog, for example, may possess considerably more higher-order cognition than we can establish from its inability to speak. An orangutan is an even more plausible candidate. A Higher-Order Thought does not require an animal to possess philosophical concepts or to say to itself, “I am thinking about my thought.” It requires the animal's cognitive system to represent one of its own first-order states.

The animal objection therefore faces a dilemma. If an animal lacks the relevant Higher-Order Thought, we cannot simply assume that it is phenomenally conscious and use that assumption to disprove the theory. And if an animal really is phenomenally conscious, it remains possible that it possesses precisely the higher-order cognitive capacity that the theory predicts.

What would count against the theory would be a case in which we had independent reason to conclude both that an animal is phenomenally conscious and that it lacks the relevant Higher-Order Thought. Merely pointing to animals that behave as though they are conscious does not establish both claims.

So the animal case does not show that Higher-Order Thought is unnecessary for phenomenal consciousness. It leaves open exactly the possibility proposed here: beings incapable of the relevant Higher-Order Thoughts may not be phenomenally conscious, while at least some animals—perhaps dogs, orangutans, and others—may be phenomenally conscious precisely because they are capable of them.

APPENDIX: PHILOSOPHICAL SOURCES OF THE DIALECTIC

Gallotti & Huebner

Their account. Two people may combine different cognitive abilities so that together they can solve a problem neither could readily solve alone. Gallotti and Huebner consider two people navigating a city: one can read the map but cannot readily understand its spatial relations; the other can understand the spatial relations but cannot readily read the map. Their abilities become integrated through interaction. The authors nevertheless hesitate to call the pair a shared mind, because they see little additional “explanatory payoff” in doing so. They suggest that a stronger case for distributed cognition occurs when the participants continually affect and update one another's thinking.

Relation to the Third Person. Their example already contains the essential structure of collective intelligence: cognitive resources distributed between different people are integrated to produce a result unavailable to either person alone. The Third Person account therefore asks a different question from whether calling the pair a mind adds explanatory usefulness: what is the process actually an instance of? A musical composition can be completely described note by note without saying that it is a waltz, while it remains true that it is a waltz. Similarly, the individual thoughts of the two people can be completely described while it remains possible that there is a further truth about the same event: their organized interaction constitutes thinking by the pair. Where Are the Car Keys? makes the claim sharper: information possessed separately by Husband and Wife is integrated and produces the belief Mikey has the keys, which neither could have formed from the information individually available. On the Third Person account, it is therefore true not only that two individual thinkers cooperated, but that the pair itself formed a thought.

Georg Theiner

His account. There are two different questions about group minds. First, can a group be a thinking entity? Second, can it be a conscious entity—a person? Theiner's argument primarily concerns the first. A group can itself think when the thinking of its members becomes sufficiently interconnected that their interaction produces the group's thought. It is not enough that several people happen to contribute thoughts to the same activity. What matters is that what one person thinks affects what another thinks, whose response affects the first in turn, so that the resulting thought is produced by their interaction. Theiner therefore rejects the assumption that thinking must stop at the boundary of an individual person. Individual people can remain thinkers while the system they form together also thinks.

Relation to the Third Person. Where Are the Car Keys? appears to satisfy Theiner's account of a thinking entity. Wife contributes the information that she put the keys on the table. Husband contributes the information that they are no longer there. Husband's information changes Wife's thinking, producing the belief Mikey has the keys; Wife's conclusion then changes Husband's thinking, and he accepts the same belief. Their interaction has produced a thought that neither could have produced from the information individually available. On the Third Person account, it is therefore true that the pair itself thinks. Theiner provides substantial support for this first step.

But that leaves the second question: can the thinking entity also be a conscious entity? Theiner discusses the difficulty but does not provide an account of how a group's thoughts could become phenomenally conscious. The Third Person account proposes an answer: the pair's collective higher-order thought makes its collective thought phenomenally conscious. The first step gives us a group that thinks. The second gives us a group that is conscious—and a conscious thinking entity is a person.

Michael Schmitz

His account. A group becomes a collective subject when its members are aware of one another as co-subjects—when they think, attend, or act toward the world together as a we. Schmitz calls this the subject mode. Two people walking together, for example, may experience one another not merely as other people who happen to be walking nearby, but as the people with whom they are walking. As this develops, they can form joint plans, commitments, beliefs, and intentions and explicitly think of themselves as we. For Schmitz, this shared self-awareness helps constitute the collective itself.

Schmitz nevertheless stops short of saying that the collective has a mind or phenomenal consciousness of its own. The members are jointly the subject of the group's beliefs and intentions, and a member can think an our thought from the perspective of the group. But the consciousness involved remains the consciousness of the individual members. The collective is a unity of minds, not a mind of its own.

Relation to the Third Person. Where Are the Car Keys? suggests that being a we and thinking as a group are not the same thing. We are walking in the rain may express a genuine awareness of ourselves as doing something together, but that alone does not show that the pair has produced a thought. In the Car Keys case something more happens. Wife knows that she put the keys on the table. Husband knows that they are no longer there. Their interaction brings this separately held information together and produces the new belief Mikey has the keys—a belief neither could have formed from the information individually available. Only then does Husband say, we both believe that.

On the Third Person account, therefore, the we is not what makes the pair a thinker. The integrated thinking makes the pair a thinker; the pair can then recognize that the thought is theirs. Schmitz gives an account of what can make several people a we. The Third Person asks the further question: what makes that we think? And once a group thinks, there remains the second question Schmitz explicitly leaves unanswered: can the group's thought itself become phenomenally conscious? If it can, then the result is not merely Schmitz's unity of minds, but a mind—and a person.

Christian List

His account. A suitably organized group can itself be an agent. It can have beliefs and desires, use information, make decisions, and act. List carefully distinguishes this from people merely doing something together: two people can go for a walk together without thereby becoming a single group agent. He also allows that a group can be aware of things and can even have higher-order thoughts—thoughts representing its own thoughts. The difficult question comes next: can the group's thoughts be phenomenally conscious? List considers the possibility that higher-order thought produces phenomenal consciousness, but objects that this might deny consciousness to animals that seem to be conscious while lacking higher-order thought. He tentatively turns instead to Integrated Information Theory. Roughly, this says that phenomenal consciousness requires the parts of a system to depend so strongly upon one another that dividing the system seriously damages what the whole can do. List thinks ordinary groups are much easier to divide in this way than individual brains and therefore probably have little or no phenomenal consciousness. His conclusion is conditional: it depends upon this theory of consciousness being correct.

Relation to the Third Person. Where Are the Car Keys? brings the disagreement into focus. Husband and Wife do more than act together. Information possessed separately by them is brought together and produces the belief Mikey has the keys. When Husband then says we both believe that, the pair forms a higher-order thought representing its own first-order belief—the very kind of group higher-order thinking that List agrees is possible. The Third Person account accepts the consequence that a being without such higher-order cognition is not phenomenally conscious. It therefore does not assume that every animal that appears conscious actually is; and whether dogs, orangutans, or other animals can represent their own thoughts is itself a question to be answered rather than assumed.

The Third Person account also offers an intelligible picture of the process. A first-order thought occurs. A higher-order thought represents it as a structure of same/different relations among points, and that relational structure is phenomenally present to the thinker. This fits an independently developed relational account of phenomenal consciousness. The visual field itself, for example, can be analyzed in terms of same/different relations among points—perhaps including relations beyond the immediate field, such as the relations that help constitute red. Thus the proposed process and the proposed character of its result fit together: the relational structure apprehended in the higher-order thought is the same kind of structure found in phenomenal consciousness itself.

List's appeal to integrated information does not provide a comparably intelligible picture. Even if phenomenal consciousness requires the parts of a system to depend strongly upon one another, it remains unclear how that fact produces a phenomenal field at all. Why should such interdependence produce, for example, a field containing red, blue, shapes, and spatial relations? List offers a possible condition for phenomenal consciousness; the relational HOT account attempts to show how a thought becomes phenomenally conscious and why the resulting consciousness has the structure it does. The disagreement therefore concerns not whether a group can think, or even whether it can think about its own thoughts, but the decisive second step: what makes that thought phenomenally conscious?

Eric Schwitzgebel

His account. Schwitzgebel asks why a sufficiently organized group could not be phenomenally conscious. His argument is deliberately conditional rather than an account of how consciousness is produced. If materialism is true, consciousness must depend upon the way physical things are organized. But why suppose that the relevant organization must occur inside one spatially continuous body, or that its parts cannot themselves be conscious? Schwitzgebel imagines radically unfamiliar creatures—a creature whose thinking is distributed among widely separated limbs, and another whose cognitive system is composed of many smaller organisms. If we would regard such creatures as conscious because of the organization of their thinking and behavior, he argues, we need a principled reason for refusing even to consider phenomenal consciousness in a similarly organized group.

Schwitzgebel then applies the argument to an actual large-scale collective: the United States. Viewed as one concrete system, it receives and exchanges information, represents its surroundings, monitors itself, responds to its environment, pursues goals, and produces coordinated behavior. These are the sorts of properties materialists commonly associate with conscious organisms. His claim is not that he has discovered the process by which these activities become phenomenal consciousness. Indeed, he describes his argument as “conditional and gappy.” His challenge is essentially: if such organization can support phenomenal consciousness in an individual organism, why not in a sufficiently organized collective?

Relation to the Third Person. Where Are the Car Keys? gives Schwitzgebel's challenge a much smaller and clearer target. Husband and Wife possess different information; their interaction produces the new belief Mikey has the keys; and Husband's statement we both believe that represents their first-order belief in a higher-order thought. Nothing about the fact that this process is distributed between two human beings by itself shows that the resulting collective thought cannot be phenomenally conscious. Schwitzgebel's argument helps remove precisely that obstacle. He even observes that, if a collective has the required lower-order mental states, there appears to be no general barrier to its having higher-order states as well.

But the Third Person account goes beyond Schwitzgebel's “Why not?” by proposing what happens. The pair's first-order thought is represented by a higher-order thought as a structure of same/different relations among points. That structure is phenomenally present to the pair. And this account of the process fits an independently developed analysis of phenomenal consciousness itself: the visual field, for example, can be analyzed as a structure of same/different relations among points. Thus the proposed process and the proposed character of its result converge upon the same relational structure. Schwitzgebel gives us reason not to exclude collective phenomenal consciousness merely because the thinker is spatially distributed or composed of conscious people. The Third Person account attempts the further step: to show how such a collective thought becomes phenomenally conscious.

Thomas Szanto

His account. A group can literally have a mind without being phenomenally conscious. Szanto calls this a “zombie conception of group minds.” A genuine group mind is not merely a collection of individual thoughts. Its beliefs, desires, intentions, and other mental states are integrated into a single unified rational point of view that guides what the group believes and does. The resulting mental states can genuinely belong to the group even though they are socially distributed across the minds of its members. Indeed, Szanto argues that group-level mental contents sometimes cannot be attributed to any individual member or reduced to an aggregate of their individual mental states.

But Szanto draws a boundary at phenomenal consciousness. A group's mental states can be distributed and integrated, but its phenomenal consciousness cannot be. He holds that a conscious state is essentially first-personal and belongs to a single subject: one instance of consciousness cannot be divided across several subjects. Individual members may therefore have conscious experiences concerning their group, and the group may even say things such as we believe that p, but there is no additional collective subject for whom the group's mental state is phenomenally present. The group consequently has a genuine mind but remains, in Szanto's terminology, a zombie.

Relation to the Third Person. Where Are the Car Keys? agrees with Szanto on the crucial first step. Husband and Wife begin with information separately possessed by each. Their interaction integrates that information and produces the new belief Mikey has the keys. The belief is distributed across two individual minds, but that does not prevent it from being one thought of the pair. Szanto's own account allows precisely this general possibility: mental states can be socially distributed while belonging to a single integrated group mind.

The disagreement arises with the second step. The first-order belief already shows how a thought can be both distributed and one. Husband's and Wife's contributions to the belief Mikey has the keys are not identical particular relational structures. On the Third Person account, however, those structures instantiate the same Platonic Form. The Form preserves what makes them instances of the same thought while leaving out the differences between their particular realizations. Thus the first-order thought has both distributed realization and unified identity.

When Husband says we both believe that, the same structure appears at the higher level. Husband's and Wife's particular higher-order representations also differ in their details, but they instantiate the same Form: we believe that Mikey has the keys. The collective higher-order thought is therefore distributed and integrated just as the collective first-order thought is. Nothing new has to be introduced merely because we have moved from a thought to a thought about that thought. At both levels, differing particular relational structures instantiate one Form and thereby constitute one collective thought.

On the Third Person account, that higher-order thought makes the first-order thought phenomenally conscious. Szanto therefore accepts the possibility of a distributed but integrated group mind, but denies the corresponding possibility when phenomenal consciousness enters. The Third Person account challenges the reason for drawing the boundary there. The same organization that gives the distributed first-order thought its unity gives the distributed higher-order thought its unity. The members' consciousness need not be supplemented by a mysterious additional consciousness located somewhere beyond them. Their distributed but integrated consciousness, unified through the same relational and formal structure that unifies their collective thinking, is the phenomenal consciousness of the Third Person.

Deborah Tollefsen

Her account. We often understand a group by treating it as a rational agent: we look to see whether there is a pattern of beliefs, desires, intentions, and actions that explains and predicts what the group does. When such a pattern exists, Tollefsen argues that we should not automatically dismiss the group's beliefs and desires as mere shorthand for the beliefs and desires of its individual members. There can be a genuine intentional pattern at the level of the group. But this raises a further question: does the existence of such a mind-like pattern mean that the group actually has a mind, rather than merely being usefully described as though it had one?

Relation to the Third Person. Where Are the Car Keys? provides a reason for taking the analogy literally. Within an individual mind, information distributed among different components is brought together and processed to produce conclusions. The same process occurs in the pair. Wife knows that she put the keys on the table; Husband knows that they are no longer there. Their information is brought together and processed, producing the conclusion Mikey has the keys—a conclusion unavailable to either component from the information independently available to it. Thus the resemblance between the individual and collective mind is not merely a resemblance in outward behavior. The collective instantiates an analogue of the process that constitutes thinking within the individual. If that process warrants saying that the individual thinks, we need a reason why its collective analogue does not warrant saying that the pair thinks.