I. Problems and motivations introduced

The distinction between first-person and third-person perspectives has been crucial for philosophers of Mind for several reasons. First, consciousness is one of the central problems in philosophy of mind, and it seems that for a subject S, what we mean by ‘conscious states of S’ cannot be revealed; or an adequate analysis of S’s conscious states cannot be given without also referring to a special subjectivity derived from the first-person perspective in which S is involved. Second, what philosophers mean when they use a variety of terms such as ‘subjective feel’, ‘qualia’, ‘phenomenal character’ and ‘what-it-is-likeness’ of a certain group of mental states seems to be intimately related to what they mean by ‘first-person perspective’. Third, the distinction between first-person perspective (hereafter FPP) and third-person perspective (hereafter TPP) seems to be crucially important for conceivability arguments as well. Some philosophers such as Eric Marcus (2004) base their argument against the conceivability of zombies on the distinction in question. We could also observe that the friends of zombie, too, use the distinction in the process of conceiving.

If so, undermining the legitimacy of this distinction would lead us to some important consequences with respect to above-mentioned issues. Here, we are aware that philosophers are usually not concerned with an investigation of the development of FPP and TPP, since, we think, they take these two notions in a manner that they do not need to be investigated developmentally. For the purpose of this paper, we will not treat these two notions in the same way; instead, our main goal will be to question the ontological status of these two notions and the distinction between them by tracing back the root of the distinction to early childhood with the help of developmental psychology.

But before explaining the motivations behind this inquiry, we need to attract attention to what we understand from these different perspectives in the first place. Surely it is not easy to give explicit definitions of FPP and TPP, but we should at least have a rough idea concerning what a perspective is and what the distinction between them amounts to. One can think of a perspective as a “way of thinking about things.” When construed this way, FPP and TPP for an agent are just different ways of thinking about things. This understanding of FPP and TPP will of course include concept application, which we believe should be avoided when we investigate whether an infant has these distinct perspectives in the sense that adults have. So, in our context, perspectives will roughly be like standpoints through which one observes or experiences things or events.

Now, let us first express what encourages us to query the developmental root of these two notions and the distinction between them. We have two intuitive problems in mind that motivate us to question the legitimacy of FPP-TPP distinction.

First, we observe that we experience our own activity from a first-person perspective, and the comparable activity of others from a third-person perspective; and that these two perspectives have little in common and have fundamental differences. Yet we, as adult human agents, are able to recognize the intended actions of others, the object-directed emotions and motivations of others, and the object-directed epistemic states of others in pretty much the same terms, as we understand these intentional activities when we ourselves are engaged in them. How is this possible?

Second, it seems clear to us that given a certain event, an average adult human being distinguishes his own first-person perspective from any other perspectives. On the other hand, current developmental psychology shows that infants cannot make such a distinction at least in the same sense adult human beings make. There are several reasons for this:

a) As cognitive actions, newborn babies present only some newborn reflexes. Infants, under the age of two, do not have a properly working cognitive system yet. Most of cognitive capacities such as the ability to explore the properties of objects, the ability to form mental representations via images (mental pictures of objects) and concepts (or categories in which similar objects or events are grouped together) etc. only develop by the end of the second year (Berk, 2003:143-52).

b) Infants also lack certain types of consciousness which an average adult human being has. Self-consciousness in the strong senseFootnote 1 and a kind of second-order (perhaps phenomenal) consciousness can be given as examples here. The first one is a type of creature consciousness, and the second is a type of state consciousness. Roughly speaking, self-consciousness in the strong sense involves higher-order awareness of oneself as a self, as being with mental states and a subjective inner life. Correspondingly, phenomenal conscious states have distinctive subjective feels or qualia that enable us to say that there is something distinctive which it is like to undergo the relevant experience.Footnote 2 Here, for our purpose, it does not matter that these rough definitions are not free from controversy. What we need in order to present the main point of this paper is that infants (especially newborn babies) do not know that they are doing what they are doing when they are doing something. In other words, they are not aware of their actions. What they do at this sensorimotor stage are just sensorimotor reflexes, which do not require any ability to distinguish themselves from other people or objects.

c) Infants under the age of two years do not have a fully developed “psychological self” either. Here, again, we do not need to know what the self is in a philosophical sense. What is important for us is to know that in order to talk about FPP and TPP, we need to have a kind of psychological entity or a philosophical (phenomenal) agent to whom we can ascribe FPP. And let us assume that what psychologists mean by “self” is the most appropriate entity to take the position of that psychological entity or philosophical agent. Footnote 3

II. Infants’ inability to have distinct fpp and tpp

Now, as we expressed earlier, because of the reasons such as the above three, it seems plausible to think that infants (especially newborn babies) do not make a distinction between FPP and TPP. Indeed, they do not seem to have a first-person perspective. But what does this mean? Does it mean that they do not feel pain? The answer is a little bit complicated since the same physical phenomenon and undesired behavioral reaction occur when we physically stimulate them, say when we prick them with a needle, as in the case of adult human beings. And assertions such as “That baby is in pain” are true. In this sense, they surely feel pain. On the other hand, since they do not have a fully developed psychological self, yet, due to the reasons mentioned above, there seems to be no phenomenal agent who can be the feeler of that pain and whom we can comfortably ascribe the first-person perspective.

Here, rejecting the above reasons, one might claim that we cannot be sure that newborn babies do not have a psychological self, perhaps they are born with an innate self. This point is not important, because we can carry the story to an earlier stage (stages in the womb for example) until we find a stage at which the organism does not bear a psychological self indisputably.

One might also have such a worry that if we make a distinction between “being a phenomenal agent who is the feeler of the pain” and “having a concept of himself as a phenomenal agent who is the feeler of that pain,” it is true that lacking the latter doesn’t cancel out the former. So, infants might lack such a concept of themselves as a phenomenal agent, but they might still be a phenomenal agent. But considering the three reasons above, it is more plausible to think that infants not only lack a concept of themselves as phenomenal agents, but also lack the requisite individuality which is necessary to be a phenomenal agent, or at least necessary to talk about the existence of a phenomenal agent in the pure sense.

Now, we can illustrate the picture we had in mind based on what we have said so far as follows:

figure 1

This picture implies that in between these two stages somewhere a fully developed psychological “self” (the phenomenal agent) must appear. And in the light of developmental psychology this happens gradually at around the end of the second year. We shall return to this stage later.

III. Approaches evaluated

Let us go back to the first problem, and link both problems with the help of a possible solution of the first, so that we can think more deeply about the emergence of the self and the distinct FPP and TPP. There is a traditional answer to the first problem in the literature, the problem of how we understand the relations between other people and their environment: We model their objective activities, which we can only know directly as behavior, by imagining from a first-person perspective, what we would experience when doing what they are doing. However, this knowledge of other minds by “analogy” or by what some others call “simulation,” cannot be the way that we initially learned to understand others. For this simulation view does not take into account the problem of how we come to know our own minds or selves, which does not seem to be immediately transparent in first-person conscious activity. How do we, then, come to recognize that the behavior of others is the same as our own comparable behavior, and come to realize that they have first-person conscious (intentional) states that go along with their behavior, but which are not directly perceivable in the behavior? The problem seems to be that even in our own case, we have no a priori basis for distinguishing between our subjective first-person (intentional) states and their objective properties involving our bodies. So how do we acquire this knowledge that is needed to perform the simulation?

There is a thesis asserted by John Barresi and Chris Moore, which aims to solve the above problem, and which we should pay attention to for the purpose of this paper. They claim that in our infancy, we develop our own self and/or ego by means of shared activities that have taken place in several situations. In those situations, we are engaged in imitation involving the same or similar objects or contagious emotional sharing that sometimes involves some particular object or person. This could also be joint attention to some object or event, or some other form of joint activity involving objects. Their story goes as follows.

In the above situations, through what they call the intentional schema, they hypothesize that humans, and, possibly, some other organisms, are able to link together into a single concept or schema. This schema involves both the third-person properties of the relations between us and the environment, as we perceive it in the activity of the other, and the first-person properties of the same relations as we experience in ourselves when also engaged in this same activity. According to them, this allows us to understand the relations between us and the environment both from FPP and TPP. Later, through the use of this schema in memory, we can come to recognize the relations of the same sorts previously jointly experienced with others, in situations not involving shared or matched activities. Having integrated the first-person and third-person properties of these relations in these previous joint activities, we can now recognize them from their first-person properties taken alone in our own case, or from their third-person properties taken alone in the case of others (Barresi and Moore 1996: 107-21).

Let us restate the original problem here. How do we know that other people, too, have a qualitatively similar FPP as we have? The analogy method does not work here, since it does not explain how we know that we have a distinct FPP in the first place. What we infer from what Barresi and Moore offer is this: Infants are not aware that they have distinct FPP either. They first get an abstract schema from joined activities, a schema in which both FPP and TPP are integrated into one single representation. This schema contains both first-person and third-person properties of the world, since for the infants there is only one integrated perspective. But later on, through the development of their imagination (as a distinct faculty from the perception), they begin to imagine both their own and other people’s situations in shared activities, where the distinction between FPP and TPP starts to appear. We think this account is quite plausible.

What we have in mind is something like what the following diagram implies:

figure 2

IV. Intersubjective perspective

Now, In the light of Barresi and Moore’s account with respect to the first problem, we can now have a better idea of the emergence of the self and thereby the first-person perspective with respect to the second problem. In order to lay out the number of perspectives in a given case, let us imagine two babies under the age of, say, fifteen months, and two physical events that happen to each of them separately, which cause a certain neuronal activity in their brains:

Event A

  • Causes c-fiber firing in baby 1’s brain.

  • Baby 1 is crying.

Event B

  • Causes c-fiber firing in baby 2’s brain.

  • Baby 2 is crying.

If these two babies had a fully developed psychological self, we would expect there being eight perspectives in the above picture: Take the event A. From the standpoint of baby 1, he would have a perceptual FPP and an imagined TPP (the perspective of how he looks like from the standpoint of baby 2.) From the standpoint of baby 2, there would also be two perspectives, asymmetrically a perceptual TPP and an imagined FPP (the perspective of what it is like to be baby 1) Four other perspectives would occur if we took the event B.

However, based on the reasons (a, b, c) we have given in mentioning the second problem, and in the light of what Barresi and Moore hypothesize about the second problem, we believe that instead of eight distinct perspectives, there is only one perspective in the above picture.

As shown in the diagram, before the baby have a fully developed self, she must have a primitive inter-subjective self. Why? First, the most general reason is that it is highly improbable that a human infant who does not have any psychological self can suddenly have a distinct (personal) psychological self. Instead, as research in developmental psychology indicates, the human infant must gradually acquire the distinctive psychological self in question, in which case the development must include some intermediate stages. Second, it is more probable that the infant first acquire a general form of self, and afterwards he must gain a fully distinguished form of self. Third, experiments in developmental psychology imply that babies under certain age cannot fully distinguish their own (phenomenal) mental state from those of others. For example, in the case of baby 1 and baby 2 above, when the event A occurs, and respectively baby 1 cries, baby 2 begins to cry too, since baby 2 cannot yet distinguish her own mental states (being in no pain) from the baby 1’s mental state (being in pain). But after certain time, gradually, baby 2 comes to realize that it is the baby 1 who is originally in pain at which time her own crying is understood as originating from baby 1’s internal state. It is at this stage that a distinct psychological self begins to appear.

Hence, once we are convinced of the existence of this primitive inter-subjective self, we think that there would be some important consequences. If there is such a stage at which infants have an inter-subjective self, then we should somehow apply a perspective to this inter-subjective self. This perspective would be an integrated one, i.e., potentially include both FPP and TPP in itself, both of which will appear separately at later stages. Let us think a little bit about this intermediate stage. If infants such us baby 1 and baby 2 at this stage could speak, what kind of descriptions about their internal life would they give? First, those descriptions that contain information about their mental states would not have distinct personal pronouns, since for them there are no distinct persons or phenomenal agents; instead, there is only one psychological self. Second, those descriptions, which contain information about mental states, would be from an integrated perspective that is neutral to FPP and TPP, which adult human beings have.

Let us illustrate this using the same picture. Take the event A once more: From outside it appears that something physical happens to baby 1, and causes c-fiber firing in her brain. She starts to cry. Baby 2 starts to cry too. Since hypothetically they can speak, they utter the same sentences from one single perspective instead of different ones from distinct FPP and TPP. Respectively they utter:

Baby 1:

“Inter-subjective self is in pain.”

(Instead of “I am in pain”)

Baby 2:

“Inter-subjective self is in pain.”

(Instead of “Baby 1 is in pain”)

A better way of understanding what would happen in the above case is to imagine “x is in pain” being used in a way similar to the impersonal “it is raining” or “it is sunny.” So, instead of the above sentences, they might utter:

Baby 1:

“It is paining”

(Instead of “I am in pain”)

Baby 2:

“It is paining”

(Instead of “Baby 1 is in pain”)

V. Conclusion

What can be derived from this diagnosis? If we have described the picture correctly, serious consequences would occur. Let us attempt to express them with no hesitation. First, since FPP and TPP, which most adult humans possess, are developmentally dependent on each other, their ontological status would also be inherently dependent on each other. Second, the epistemological status of the descriptions that are gained through FPP and TPP would also be inherently dependent on each other. Third, without the presence of the other, they would not appear as distinct in the first place. This means that it is highly probable that without interacting with another human being, infants could not develop an FPP, which would affect the philosophical point of view in dealing with some serious problems of philosophy of mind.

Recall the three motivations behind why we attempted to analyze the ontological statuses of FPP and TPP, and let us see how those concerns we had are affected by the above implications.

First, the distinction between FPP and TPP is crucial because an adequate analysis of what we mean by ‘a subject’s conscious state’ can only be given by means of an adequate understanding of FPP and its differentiation from TPP. But if, developmentally, FPP is rooted in an integrated perspective and desperately need TPP to emerge as a distinct perspective in the first place, then a subject’s conscious states will most likely be rooted in a kind of shared consciousness as well, and will ontologically need other subjects’ conscious states to emerge as distinct conscious states. Second, we are concerned with terms such us ‘subjective feel,’ ‘qualia,’ ‘phenomenal character’ and ‘what-it-is-likeness’ that philosophers of mind often use as properties of a certain group of mental states because these terms are closely connected with what those philosophers of mind mean by ‘first-person perspective.” If the above implications are correct, however, it seems that the properties expressed by these special terms can only be perspective-dependent properties--a perspective that cannot come into being without the existence of at least one another perspective. Third, the distinction between FPP and TPP is also crucial for conceivability arguments against physicalism because the arguments in question employ a fictional philosophical being called ‘zombie’ whose only difference from a normal human being is that it merely lacks a first-person perspective. But if FPP is not an inborn possession that normal human beings have innately, and rather it is gained afterwards, then it is clear that we were all zombies once in our past and achieved to be a normal human being by gaining a first-person perspective later on. Other than these, further conclusions can surely be derived from the above implications.

To sum up, we attempted to examine the distinction between the first-person and the third-person perspectives from the standpoint of developmental psychology. Our aim was to show both that we, as adult human beings, do not innately possess a first-person perspective and a grasp of the distinction between the two perspectives, and that a deep consideration of the emergence of these two perspectives as distinct ones would lead us to some significant philosophical consequences that cannot be ignored for certain issues in philosophy of mind, cognitive sciences and other related areas.