terug (Nederlands)

back (English)

Explaining consciousness and the duality of method


In consciousness studies, the first-person perspective, seen as a way to approach consciousness, is often seen as nothing but a variant of the third-person perspective. One of the most important advocates of this view is Dennett. However, as I show in critical interaction with Dennett’s view, the first-person perspective and the third-person perspective are different ways of asking questions about themes. What these questions are is determined by the purposes that we have when we ask them. Since our purposes are different according to the perspective we take, each perspective has a set of leading questions of its own. This makes that the first-person perspective is an approach of consciousness that is substantially different from the third-person perspective, and that one cannot be reduced to the other. These perspectives are independent, although complementary approaches of the mind.

Explaining Consciousness and the Duality of Method

Henk bij de Weg


A central issue in the philosophy of mind is the problem of the nature of the mind-body relation and how to explain this relation. Currently, one finds an almost unanimous rejection of Descartes’s substance dualism. The theories that have replaced it generally defend a kind of mind-body dependence insofar as they maintain that the purely mental does not exist but that mental properties are based on physical properties. The question how then the mental is dependent on the physical, has been answered in different ways in the course of time. As most important views I want to mention behaviourism (Carnap, Hempel, Skinner), identity theories, both type-type identity theories (Smart, Place) and token-token identity theories (Kim, Davidson), functionalism (Putnam, Fodor), and eliminative materialism, in which the mental is even completely explained away for the benefit of the physical (P.M. Churchland). Different as these answers are, they have one thing in common, anyhow: they approach their objects from the viewpoint of the investigator, the viewpoint of the third person. This is striking, for consciousness, the mental, has important first-person aspects, but these first-person aspects are as such not dealt with in theories of consciousness as just mentioned. Because of this, these theories do not agree with the ways the individuals that are the bearers of the phenomena described and explained experience them themselves. In this respect the theories are incomplete or at least one-sided.

It is no surprise that this shortcoming was criticized in one way or another, and that attention was called for the experiences of the first person. The most important of those criticisms are Nagel’s "What is it like to be" argument, Searle’s "Chinese room" argument, and Jackson’s Knowledge argument. Although so the existence of first-person phenomena was taken seriously in the philosophy of mind, the fact that they supposedly cannot be "caught" from an external point of view has not led to the development of a separate first-person perspective that is fundamentally different in character from the third-person perspective, although Nagel has gone rather far. It has been recognized that the first-person perspective has features of its own, indeed1, but the general trend has been to reduce the mental to the physical and to neglect the typical methodological features of the first-person perspective in favour of third-person features. The first-person perspective is made dependent on the third-person perspective.2 And even if the difference between the former and the latter is recognized, this does not imply that it is acknowledged that the first-person perspective has a logic of its own that is distinct from, but equivalent to, the logic of the third-person perspective.

In this paper, I cannot found these theses extensively. What I can do is show that the first-person perspective is a separate approach of consciousness that cannot be reduced to the third-person perspective. To this end, in section 3, I analyse the heterophenomenology that has been developed by Dennett as a method for integrating the first-person perspective into the third-person perspective. Then, in section 4, I develop some theoretical and methodological aspects of the relation between both perspectives that have only been touched in the preceding section. Section 5 gives my conclusion. I start, however, with a few terminological remarks in section 2.

2. Some terminological remarks

This paper is about the first- and third-person perspectives. As the terminology already suggests, it is about standpoints from which something can be viewed. If we talk about the first-person perspective, we mean that what the perspective is about is viewed from the standpoint of the bearer of the perspective. Or, as said so briefly and to the point by Velmans, the first-person perspective is about "how things appear from the subject’s point of view" (Velmans 1991, section 8). We can also say: The first-person perspective is the view from within. This indicates already the other way one can view what the first-person perspective is about, namely from the outside. Then one takes the standpoint of the third person, and then one gets a third-person perspective. Or as Velmans puts it: "[the] ‘third-person perspective’, refers to how events relating to the subject appear to an external observer (e.g. an experimenter)" (ibid.).

I just talked about "what the perspective is about". I want to call this the theme of the perspective. It has to be distinguished from its contents, which is the way the perspective has been filled in concretely. An example of a theme is "pain". From the first-person perspective it is "my pain"; the contents is formed by my pain experiences. From the third-person perspective the theme "pain" can be specified as "Henk’s pain" and the contents is then for instance Henk’s C-fibre firing. Strictly speaking, "my pain" is another theme than "Henk’s pain". However, I consider both themes as specifications of the general theme "pain" that are necessary to determine the perspective. As for themes for the first-person perspective, we may think of thoughts, emotions, and generally what are called "qualia", or also of the body as it is experienced. But first-person themes can also refer to the world outside the first person as it appears to and is experienced by the first person. Themes for the third-person perspective may refer to the way the brain and the body function physically, and to a person’s physical reaction to impressions of the external world. In conformity with the distinction between the first-person perspective and the third-person perspective, one usually talks of experiences and to experience in the former case, and of observations and to observe in the latter.

3. Perspective, meaning, and asking questions

In the preceding section, we have seen that the two perspectives can have the same general theme. However, the ways the theme is approached and filled in are different. These are determined by the relation of the person to the theme. Then, the question presents itself whether both perspectives on a theme are related to each other and whether one perspective can be reduced to the other. Dennett, one of the most important advocates of the superiority of the third-person perspective, answers these questions affirmatively. Or rather, an affirmative answer is implied, when he describes his method of heterophenomenology. Therefore, I want to examine his method here. In 3.1, I briefly describe the heterophenomenology. In 3.2 – in critical interaction with Dennett’s views – I introduce the distinction between two kinds of meaning and two ways of asking questions. In 3.3, I analyse the consequences of this distinction for the method of heterophenomenology.

3.1. Dennett describes the heterophenomenology extensively in his 1993a.3 It is, he says, "a neutral method for investigating and describing phenomenology" (1993a, 98), so a method for catching what I have called here the first-person perspective (cf. id., 70). That the method is neutral means that it is neutral "with regard to the debates about subjective versus objective approaches to phenomenology, and about the physical or nonphysical reality of phenomenological items" (id., 95). For studying his object, Dennett uses the verbal communication of people investigated in relevant experiments (id., 73-74).4

An analysis according to the method of heterophenomenology consists of a number of steps. The first step is the recording of verbal communication on videotape, sound tape, as electroencephalograph (sic: see note ), or no matter how. Dennett develops here the example of the recording of sounds of the communication between subjects and the experimenter during the experiment. Then, three trained stenographers prepare independently transcripts of these raw data (id., 74). By taking this step, "we move", according to Dennett, "from … the world of mere physical sounds … into … the world of words and meanings, syntax and semantics" (ibid.). However, a transcript itself is no longer given as data but it is an interpretation of these raw data (id., 75). This interpretation consists primarily in assumptions made by the stenographer about the language used in the verbal communication, correction of obvious mistakes, making sense of words not well understood on the basis of the speaker’s intention, etc. Or, as Dennett characterizes this process of interpretation: "We effortlessly … ‘make sense’ of the sound stream in the process of turning it into words" (id., 75; italics HbdW). In this way, the experimenter gets a text. In the next step the experimenter interprets the text as a record of speech acts. But this step requires that the experimenter adopts, what Dennett has called, the "intentional stance" with regard to the agent who has uttered the text:5 the agent is supposed to be a rational agent with beliefs, desires, and other mental states that acts intentionally (id., 76). In addition, Dennett mentions another problem, namely whether we need not presuppose the consciousness of the subject that has produced the verbal behaviour that made the experimenter to write down the text. It certainly might be possible that the subject is a zombie (or, for example, is simply confused). Dennett evades this problem by referring to the analogy of interpreting the text of a novel: just as a novel is a work of fiction but still can be read and interpreted as were it a real world, in the same way an experimenter can take the subject’s text as "that subject’s heterophenomenological world" that can be interpreted further (id., 81).

3.2. These are the main points of the method of heterophenomenology that are relevant to my discussion. According to Dennett the method is "maximally extended … a neutral portrayal of exactly what it is like to be that subject – in the subject’s own terms…" (id., 98; italics D.). But is it really neutral in its approach of the subject and does it really result in a portrayal of "what it is like to be" for that subject, its phenomenology, that is its first-person perspective? Dennett contradicts himself about the neutrality. For, although Dennett himself calls his method neutral in so far as it does not opt for a subjective or an objective approach, he also says that it starts out from the third-person perspective (id., 72, 96).6 And what else is a third-person perspective but an objective approach of the phenomena to be studied? (cf. Dennett himself id., 72). But this means that Dennett incorporates the first-person perspective into the third-person perspective. Obviously, Dennett does not see that catching the first-person perspective is something different than recording certain external manifestations of this perspective. However, one can know the first-person perspective only by finding out what is relevant from the first-person perspective itself, and not by finding out what seems to be relevant from the outside, from the third-person perspective. In order to make this clear, I shall examine what having a first-person perspective and having a third-person perspective involves. The first-person perspective is the "view from within"; the third-person perspective is the "view from the outside". What are we trying to say, when we say that a person has a certain view from within on the world in and around her and that this person experiences this world in a certain way? Then we want to say that this person (in this case the first person, P1) gives the things and events in or around herself a certain meaning. Now, this is also true for the third person (P3) that observes P1 and tries to find out what P1 experiences. P3, too, gives what she observes about P1 a certain meaning, founded on her view from the outside. But this does not imply that the meaning assigned by P3 agrees with the meaning assigned by P1. We can find this view also in Dennett, albeit implicitly, when he states: "You are not authoritative about what is happening in you, but only about what seems to be happening in you, and we are giving you total, dictatorial authority over the account of how it seems to you…" (id., 96; italics HbdW). "What is happening in you" is the field of the experimenter. It is her view on – and so her interpretation of – what is happening in you, and therefore it is not you but the experimenter that is authoritative about the contents of it. On the other hand, "what seems to be happening in you" is the field of the first person. It is her view – and so her interpretation –. Therefore, in this case, it is the first person that is fully authoritative about its contents.

In my 1996 and 2001 I have drawn a distinction between two kinds of meaning, which I have called meaning 0 and meaning 1. The former was "the meaning the people who make up social reality give to this social reality or to parts of it themselves; … it is their interpretation of their own lived reality". The latter I defined as "the meaning a scientist gives to an object, either physical or social in character; … it is the scientist’s theoretical interpretation of reality" (bij de Weg 2001, 19-20). In terms of the present study we can say it this way: Meaning 0 is the meaning given by a person herself to the phenomena that make up her own personal world as she experiences them or lives through them. It is her view from within, her interpretation from the first-person perspective. On the other hand, meaning 1 is the meaning given by a person to the phenomena that make up the world that she is confronted with as she observes them or studies them. It is her view from the outside, her interpretation from the third-person perspective. A consequence of this distinction between meaning 0 and meaning 1 is that an investigator that wants to study the first-person perspective has to deal with two kinds of interpretation, or, as Habermas says it, the investigator has to interpret on two levels: First, the investigator has to understand what the first person means when she interprets her phenomenal world; and second, the investigator has to interpret this interpretation within her own theoretical frame (Habermas 1982, 162-163). This double-interpretation problem is not seen by Dennett. However, in this paper I want to pass it over. Here I want to consider the logical consequences of the distinction between meaning 0 and meaning 1.

If we want to know something about a person, either how this person herself experiences or sees certain things from within (so from the first-person perspective), or as we ourselves observe from the outside what we as investigators consider to be relevant about this person (so from the third-person perspective), then we need to collect data (as Dennett does as well). In the first case, we want to find out the meaning 0 of a certain theme; in the second case, we search for meaning 1. However, we do not get relevant data automatically, but only by asking the right questions. What the right questions are is determined by the purpose of the questions in the first place, i.e. what kind of meaning we want to catch, and that depends on the theme specified. The questions that we have to ask in order to catch the meaning 0 – so the first-person perspective – are the same questions as the first person herself asks – implicitly or explicitly. These questions are different from those asked for getting meaning 1, if we as investigators ask questions from the third-person perspective. In the third-person perspective the object of the question is central: we want to know something about the object and what the object means for us as investigators. On the other hand, if we are interested in the first-person perspective, then we as investigators ask, as said, basically the same questions as the first person. But now not the object of the question is central to us but just its subject: we want to know something about the subject and what the object of the question means for the subject.

Typical questions from the third-person perspective are causal why-questions and the questions of definition that precede it. What is pain? What causes pain? Why has this patient pain? Questions like these lead to answers in terms of the structure of the brain and the nervous system, the irritation of nerves, and for the patient involved maybe to diagnosing an inflammation. However, from the first-person perspective questions like these are as such not interesting, and often they even cannot be answered. It is not so that the doctor asks the patient "How do you know that you have pain", and that then the patient answers by giving a definition or quoting a medical companion, adding "and that is what I feel". Instead, the doctor asks the patient what the symptoms are or what she feels, and then the patient describes her sensations: "I feel a sharp pain in my chest" or "I have a piercing pain in my foot". That the patient knows that it is a pain that she feels is taken for granted. If the patient knows that what she feels are the symptoms of a gastritis, it is usually not, because the pain that she feels feels like the pain of a gastritis, but for example because she has learned from books the place of the stomach in the body and what the symptoms of a gastritis are, and so has acquired third-person knowledge.7 For the first-person perspective, such questions about the agent’s sensations or, in general, about the agent’s experiences – and then I take sensations as a kind of experiences – are what questions of definition are for the third-person perspective. While, in the case of a definition, we ask "What is it?", in the sense of "What is it as such?" or "What is the essence of it?", in the case of a first-person question about experiences, we ask "What is it for me?" (i.e. "What is it for the subject of the question?"), or in Nagel’s terms "What is it like to be [a bat]?", "What is it like to have [pain]?" etc., i.e. for the bat or for the person having the pain (see Nagel, 2000).

For Dennett, the "what it is like to be"-question is characteristic for the first-person perspective. It is not the only question that can be asked from this perspective.8 Also the causal why-question has its counterpart in the first-person perspective, namely the reason-giving why-question. Here, too, the two types of questions differ in their relations to the asking subject. If one asks a causal why-question, one tries to get an answer that is objective in the sense that it is valid independent of the asking subject and that every other subject would give basically the same answer. On the other hand, the reason-giving why-question does not ask "why, whatever the case may be" but "why for me (as asking subject)", and so it does not ask what makes something happen but what the justification for it is. A reason can be a justification for the person herself (her reasons for choosing between several possible actions) as well as a reason for justifying to another person why she did what she did. Why did John rob the bank? A psychologist may say that his act was caused by the bad environment where John has grown up and the influence of his criminal friends, which made that a hold-up was considered as an acceptable means to get money. John himself justifies his act perhaps to the judge by saying that he immediately needed money, that he could not get a loan, and that he was in a desperate situation, because he was blackmailed.

3.3. I want to complete my criticism on the heterophenomenology. I started from the distinction between the first-person perspective and the third-person perspective. Then, I have shown that this distinction involves more than the casual viewpoint of the bearer of the perspective in relation to an object, but that it goes back to the nature of the meaning assigned. The first-person perspective is not only "from within" (in a sense, this is also the case for the third-person perspective, because it is also borne by a certain person9) but it is also subjective: it is the perspective as experienced by the bearer of the perspective. On the other hand, the third-person perspective is objective, not only because it is "from the outside", but especially in the sense that the object of the perspective, and not the bearer of it, is at the centre: basically, it is independent of the bearer of the perspective. Now it is so that if one wants to collect knowledge from and about a certain perspective, one needs to ask questions, and as we have seen, the different perspectives lead to different types of questions. It will be clear that this difference in types of questions has methodological consequences as well (see my 1996, where I discussed especially the logical and methodological foundations of the first-person perspective and its relation to the third-person perspective). If we want to know something about the first person as a first person, so if we want to try to catch her first-person perspective, then we talk of "understanding" (or trying to understand). If we want to approach the same person from the third-person perspective, then we talk of "explaining" (or trying to explain). We understand a person, if we answer a "what it is like"-question in terms of the experiences of and for this person, so as first-person experiences; or if we answer a why-question in terms of reasons for the agent. We explain a person, if we answer a "what it is like"-question in terms of, for example, brain structures or environmental circumstances, so in the objective sense, independent of the view of the person concerned; or if we answer a why-question in the objective sense, so in terms of causes.

It is a problem with Dennett’s heterophenomenology and his approach of the first-person "what it is like"-question that he places his method within the frame of the third-person approach – so within the frame of an objective approach –, and that in this way he tries to answer this question from the third-person perspective.10 For Dennett, the first-person "what it is like"-question – and consequently getting at meaning 0 – belongs in fact to the phase of data collection, which precedes theoretical explanation: "You [= the first person] get the last word. You get to edit, revise, and disavow ad lib, … so long as you avoid presumptuous theorizing about the causes or the metaphysical status of the items you report…" (Dennett 1993a, 96; italics HbdW). For theorizing is reserved to the investigator. That, according to Dennett, meaning 0 precedes meaning 1 is also the real sense of the quotation above, where Dennett states that only the first person is authoritative about what seems to be happening in her: the bearer of a first-person perspective provides the experiences that the investigator incorporates theoretically. Now it is not a problem of Dennett’s procedure that it is not acceptable as such, but that Dennett does not see that the first-person perspective is characterized by types of questions of its own that are distinct from those of the third-person perspective and that are independent of them. Collecting first-person data and having them confirmed by the persons involved that the description of these data as data11 is correct does not lead to catching the first-person perspective, if these data are used for answering questions that are asked from the third-person perspective.

As we have seen above, we can "understand" a person as a subject (and then we take the first-person perspective as point of departure), or we can "explain" that person as an object (and then we take the third-person perspective as our point of departure). If we explain, we as investigators give a meaning 1 to what we observe. For the correctness of these meaning 1 interpretations we are accountable to the community of investigators. This is also the case for meaning 0 interpretations, if we try to understand a person. But meaning 0 is the meaning as assigned by the person under investigation herself: the experiences investigated are the experiences as they are for the person experiencing them; the reasons ascribed to actions are the reasons as produced by the agent. We can, therefore, say that we understand a person only if this person confirms our interpretations. Thence it is so that Dennett’s "you get the last word" – i.e. that you, the first person being investigated, have "dictatorial authority over the account of how it seems to you, about what it is like to be you" (and I, HbdW, want to add "and about what your reasons are") (id. 96) – is not only relevant to the correct interpretation of the data as data. It applies also to the subsequent theoretical interpretation of these data by the investigator from the perspective of the first person. We can say that our interpretations concerning what something is like for the person involved or the reasons that an agent has are correct, only if this person or agent herself confirms them.

We can compare it with the case of studying a language: a linguist may give a complete description of the grammar of a language, but, in the end, the native speakers of this language are those who determine whether a sentence formulated by the linguist is grammatically correct (cf. bij de Weg 2001, 20-21). However, a meaning 0 interpretation made by an investigator on the basis of certain data is a theoretical interpretation. This is why it does not need to be so that the wording used by the investigator is equal to the wording that the person studied uses or would have used for the formulation of her own experiences herself. This is no problem as long as we manage to catch the meaning 0 with our interpretations. Therefore, Apel has put it this way that the subject must be able to confirm our meaning 0 interpretations in principle, e.g. after adequate training (Apel 1979a, 201-202). For, of course, it must not be so that generally giving a theoretical interpretation becomes impossible, because the subject involved does not know what giving a theoretical interpretation means. A consequence of this requirement of confirmation by the subject is, however, that the investigator has lost her privileged position towards the subject studied (Habermas 1982, 173). Unlike in case of an explanation from the third-person perspective (e.g. by means of the heterophenomenological approach), if an investigator wants to get valid meaning 0 interpretations in order to understand a person, she needs to communicate with this person not only in order to collect data but also in order to legitimize the theoretical interpretations.

4. The relation between the first- and third-person perspectives

Now I want to develop some important aspects of the relation between the first-person perspective and the third-person perspective that could only be touched in the preceding section. First I exemplify the distinction between the first-person perspective and the third-person perspective in order to state more precisely what having a certain perspective involves and what makes these perspectives different (4.1). Then I deal with the question whether it is possible to study the first-person perspective from the standpoint of the third-person perspective (4.2). Finally, I discuss the connection between both perspectives (4.3).

4.1. Let us look at the following figure:


This figure can be seen as a simple example of phenomena that usually are much more complex and that can be experienced or can be observed, e.g. music, a painting, a house or a scientific experiment, but also a religious perception, a world view, plans for the future, or an action, e.g. taking the train to Utrecht in order to visit the university library.

I look at the figure. Now we want to know how this looking takes place. So "looking" is the general theme of our investigation. We can examine this theme from both perspectives. In the case of the first-person perspective meaning 0 is involved; in the case of the third-person perspective meaning 1 is involved. The difference between both perspectives is expressed methodologically in a difference in types of questions: in questions asked from the first-person perspective the subject referred to in the theme is central; in questions asked from the third-person perspective the object of the theme is central (see 3.2). For our theme "looking", this means that we focus our attention on "my looking", if we take the first-person perspective as our point of departure – as opposed to, for example the investigator’s looking. If we take the third-person perspective as our point of departure, we focus our attention on looking at the hare/duck figure – as opposed to, for example, looking at a non-ambiguous figure (maybe we want to know whether different brain processes are involved in these cases). We can represent this difference in this way:

The left oval represents the first-person theme; the right one represents the third-person theme. That "the figure" is not enclosed by the left oval does not mean that the figure is not involved in the analysis from the first-person perspective. One cannot look just like that but one can look only at something. The theme for the first-person approach in this example is, nonetheless, my looking. Likewise, the "I" is not covered by the third-person theme, but being the one who looks, I am involved in the analysis from the third-person perspective.

In 3.2, I have indicated two leading questions for each perspective. For our example, we can specify and then answer them as follows:

for the first-person perspective:

Q: What is the figure like for me?

A: I see a schematic figure in which I recognize the outline of a hare.

Q: Why does the figure represent a hare?

A: According to me, the two protrusions at the left are two ears and the little dent at the right is a mouth.

for the third-person perspective:

Q: What kind of figure is that?

A: It is an ambiguous figure representing both a hare and a duck drawn for experimental purposes.12

Q: Why do some see the figure as a hare and others as a duck?

A: "The way an object is perceived is not determined solely by the external stimulus it presents to our senses. It is determined, at least sometimes and at least in part, by the antecedent cognitive state, educational background, or frame of mind of the perceiver" (Churchland 1995, 108). And Churchland continues answering the question with the help of the "recurrent network" that he has explained in the preceding pages of his book: "If the recurrent activity arriving at layer 2 [of the recurrent network] is such as to tilt its vectorial activity already in the general direction of a rabbit13 vector, then the input of [the] figure … will be much more likely to result in the activation of a rabbit vector over any other. Alternatively, if the recurrent activity tilts layer 2 toward a duck vector, then the input of [the] figure … will almost certainly result in that activation pattern instead. Here the cognitive activity at layer 2 is steered by information (or misinformation) from sources other than the sensory periphery" (id., 108-109; italics Churchland).

Here Churchland’s explanation as such is not under discussion. What is germane is that Churchland gives an explanation of my looking – or, as he says, perceiving; the difference is not important here –, and then my seeing and interpreting of the hare/duck figure in which I myself as subject am no longer present. He is interested only in how my looking and what I see and my interpretation can be grasped in terms of brain structures and processes and the input of these processes, e.g. my educational background, antecedental knowledge, etc. Just as for Dennett, my experiences (for example, that I interpret the figure as a rabbit – or a hare or a duck) are only input for the analysis, resulting in an explanation from the third-person perspective. However, from the first-person perspective my experiences as subject as such are central. What is relevant then is that I see the figure as a hare – or rabbit or duck – and that I can tell my reasons for it, even though I may be mistaken. Although these experiences are "reflected" in the structures and processes of my brain, they have a meaning for me only as experiences and not as brain structures and processes. From the first-person perspective, I normally have no knowledge of what is happening in my brain. It is these experiences that determine my thinking and acting as subject from the first-person perspective and that, for example, I communicate to others.

4.2. It is not so that it is not possible that I, the first person that I am, take a third-person perspective towards myself. Nor is it so that another, a third person, has no access to my first-person perspective, no matter how. For taking a perspective is approaching a general theme in a certain manner either by highlighting the subject or the object, guided by the choice for questions of a certain type.14 And there is no reason why I (the first person) cannot ask the same questions about myself that another (a third person) asks about me or that I ask about another from the third-person perspective. But on the other hand, there is also no reason why another cannot ask the same questions about me that I ask about myself. However, unlike me, the other has no direct access to the answers on these questions (and in this sense Nagel is right). Nevertheless, in order to answer the questions from the first-person perspective and so to get an indirect access to my experiences it is not sufficient that the third person fits the answers on questions from the first-person perspective in her own theoretical frame of interpretation, for then, as in Dennett’s heterophenomenology, the first-person perspective functions only as input for a third-person interpretation, and it is meaning 1 that is assigned. In order to catch my first-person perspective as such, and with this my meaning 0, the third person will have to take towards me a, what Habermas called, "performative attitude" (Habermas 1982, 164), and the third person will have to, as Apel called it, "verständigen" herself with me (Apel 1979b, 7). That is, she must try to understand the meaning that I give to what I see, hear, feel etc., and make sure that this is what I mean by coming to an agreement with me that this interpretation is right. This can be done only by communicating with me. So, while taking a third-person perspective is basically a one-sided process, because the third person is opposite to the first person, trying to catch a first-person perspective requires a process of mediation between the third person and the first person about the sense and meaning of the first-person’s conceptions, interpretations and experiences. In this sense, catching the first-person perspective from the outside is a confrontation of two frames of meaning (see bij de Weg 2001, 24-25; Giddens 1986, 144-148).

One can object that the other cannot know my first-person perspective, because the third person has only indirect access to the first-person’s experiences: the third person sees me from the outside, anyhow; I see myself from within. I do not want to minimize this problem. On the other hand, in daily life it is quite normal that we try to understand another person and that we comprehend what another person feels and that we sympathize with her. This is why I think that the fact that there is a problem of indirect access of the first-person perspective cannot involve that it is a priori impossible for a third person to know the first-person perspective of another person. For why would this be impossible for us as philosophers or investigators while we can do it in daily life: taking the perspective of the other? I do not want to exclude beforehand that certain inner phenomena can be investigated. If there exist problems for the third person to investigate the first-person perspective from an outside position, I think that these problems are not fundamental but practical, and those are there to be solved like any scientific problem. This does not imply, incidentally, that there might no "immunity to error through misidentification relative to the first-person pronoun" (Shoemaker 1984), or that the first person has no authoritative knowledge of her own experiences. These issues concern the verification of knowledge statements from the first-person perspective, not the question of access to the first-person perspective.

4.3. Experiencing is linked to the first-person perspective, and observing to the third-person perspective. This does not mean that her own experiences are not relevant to the third person in that capacity. As we have seen, the basis of the distinction between experiencing and observing and between the first- and third-person perspectives is the distinction between meaning 1 and meaning 0. To put it in a different way than above, we can say that meaning 0 is the meaning given in daily life. We may call it "commonsense", if we do not forget that meaning 0 can have a high degree of sophistication. People can do very complicated things in the practice of their professional lives, for instance as an artisan, and people can maintain very complicated social relations, without being able to express exactly what they do, although they do it skilfully. Moreover, we can describe meaning 1 also as the meaning given to things set apart from the routine of daily life. It is abstraction and conscious reflection. Just this abstraction and reflection makes that meaning 1 has to be translated into practice before it can be applied. In science abstraction and reflection take the form of theoretical interpretation.

However, as I have argued in my 199615, for an observer or a theoretical interpreter assigning meaning 1 is never detached from assigning meaning 0. For a practical or theoretical investigator abstraction and reflection, i.e. theoretical interpretation, on the base of observation can refer only to a part of her world or life. They cannot encompass her whole world or life, even not if the investigator studies the parts that make up her world or life one by one. Large areas of what makes up the life and world of the investigator and that have developed from the early childhood as parts of her life and world, or that have been added later, belong to her commonsense background and they form the frame for her theoretical interpretation, i.e. her assigning of meaning 1. How much developed and elaborated the observations and theoretical interpretations of an investigator may become, there will always be a limit where theoretical interpretation ends and where only commonsense meaning, i.e. meaning 0, exists and where the investigator must say: "that is the way things are and that is how I feel it". In this sense, experience is basic to all investigation, both to the theoretical interpretative part and to the practical observational part. Or what is the same thing, in this sense the third-person perspective is based on the first-person perspective.

There is also a more limited sense that observation is founded on experience, as Velmans has made clear (2000, 175). But in the light of my present analysis I want to present the argument in a different way. Both subjects and investigators ask questions – explicitly or implicitly –, if they want to get at the meaning of something, albeit questions of a different type. Returning to our example of the hare/duck figure, there the answer on the third-person what-is-question has been that it represents an ambiguous figure drawn for experimental purposes. The causal why-question can be answered again by taking Churchland’s quotation. Now it is true that a theoretical interpretation as exemplified by these questions and answers has an intersubjective meaning in the sense that the interpretation can be shared by other investigators and that there can be agreement among investigators about the meaning of the interpretation, even though not all investigators may endorse it. On the other hand, it is also true that a theoretical interpretation is not unrelated to what it means for a particular investigator, since the investigator is a subject for herself. This is why the hare/duck figure is not only a theoretically grounded figure with a meaning 1 shared by the community of investigators and set apart from the ordinary course of life. As such it has at the same time a meaning for the private investigator who knows what it means being an investigator and giving a theoretical interpretation. And in this regard, there is no difference between the way that the subject examined by the investigator interprets the hare/duck figure and the way that this investigator interprets it: both interpretations are interpretations in the light of their own lived realities and in this way both interpretations are meaning 0 interpretations. Likewise, Churchland’s causal explanation of the interpretation of the hare/duck figure in the brain is not an intersubjective causal explanation with a meaning 1 just like that. This explanation is meaningful for the individual investigator who endorses it or at least understands it, and for the investigator it derives its subjective meaning from the scientific and learned background frame that says what counts as a valid causal explanation and what the theoretical terms employed mean. So, also causal explanation is a kind of assignment of meaning 0 in the light of the reality of the investigator, a reality formed in this case by the research community. Since it is just so that observation is based on theoretical interpretation (Popper 1959), here again we see that observation is ultimately based on assignment of meaning 0. Seen in this way, and in this respect, we can therefore endorse Velmans’s thesis that "in terms of phenomenology there is no difference between ‘observed phenomena’ and ‘experiences’ " (Velmans 2000, 175; cf. Bermúdez 2000, 145). Or as we can say it, too: the third-person perspective is the first-person perspective of the investigator and her observations are her experiences from the first-person perspective. However, there are two important differences between the first-person perspective of the subject that is examined and the scientific and learned first-person perspective of the investigator. The former is a commonsense perspective formed in the practice of daily life, or at least it is for a large part. Usually it is implicit. The latter, on the other hand, is essentially well-thought-out and constructed, and it can easily be made explicit (and that is in fact the reason why we call experiences by an investigator "observations"). The other difference is a consequence of the first one: being a commonsense perspective, the first-person perspective of the subject is basic. But the first-person perspective of the investigator that provides the frame for theoretical interpretation is a limited perspective and, as we have seen above, it cannot exist without a founding wider commonsense perspective: science is a limited, albeit an important, niche in the practice of life.16

That the third-person perspective is founded on the commonsense perspective that the first-person perspective is in the practice of daily life does not mean that the third-person perspective can be reduced to the first-person perspective as a methodical approach. Methodically, both perspectives are independent approaches of a general theme with a set of questions of their own, with one perspective focusing on the objective aspects of the theme and the other one on the subjective aspects. Moreover, each perspective represents a particular, what Habermas and Apel called, "knowledge-interest". For the third-person perspective, it is "the interest in controlling an objectified environmental world". However, as Apel adds, this interest "may … be sublimated, and as such it goes far beyond its manipulative paradigm, compromising e.g. all theoretical objectifying of world-pictures and even being presupposed in our present attempt at laying out leading knowledge-interests philosophically, in a very abstract model of their structure" (Apel 1979b, 6-7). So the essence of this interest is not so much that the knowledge actually is used for manipulation and that the theoretical insights have been formulated with a technological purpose, but that the knowledge can be used for manipulation and that the theoretical insights can be transformed into technological rules (cf. ibid.). In psychological terms, we may say that the possibility of control is a disposition of this kind of knowledge.

As for the knowledge-interest of the first-person perspective, it is what Apel describes as "the interest in communicative understanding" or, with a German word, the interest of "Verständigung" (id., 6-8). "Verständigung" is the substantive belonging to the verb "verständigen", which we met already on p. *. There it was a means to find out the first-person perspective of a subject; here it is the interest behind the first-person perspective as a methodical approach. Briefly, we can say that methodically the first-person perspective is to understand in order to understand better.

I do not want to discuss here the complex relation between understanding and Verständigung. I have done that elsewhere (1996, 2001). Nor do I want to go into Apel’s and Habermas’s theory of knowledge-interests. But that the first-person perspective and the third-person perspective are linked to different knowledge-interests makes clear that they are independent views on reality. These perspectives are complementary to each other in the same way as described by Apel for the methods of understanding and explanation (Apel 1979b, 12-13; remember that above on p. * I linked the first-person perspective and the third-person perspective to these methods):

  1. The perspectives supplement each other: what is known, experienced or observed from one perspective cannot be known, experienced and observed from the other perspective.
  2. The perspectives exclude each other: they have different intentions, expressed in different types of questions. The first-person perspective is subject directed; the third-person perspective is object directed.
  3. For both reasons, the perspectives cannot be reduced to each other.

In fact, this reformulation of Apel’s complementarity thesis for the methods of knowledge into the idea of complementarity of the first-person and third-person perspectives is a summary of my preceding analysis and my criticism on Dennett.17

5. Conclusion

Consciousness has subjective and objective aspects. I think that hardly any philosopher can be found who does not agree. But within the philosophy of mind, the mainstream is formed by views that reduce the subjective to the objective in one way or another. Now I do not want to deny that it is possible to explain the subjective aspects of consciousness from the objective aspects. By denying that, I should ignore many important results in the field of consciousness research. But reducing the subjective to the objective, when one approaches the mind, passes over what the subjective makes that it is subjective: the typical perspective that is called the "first-person perspective", and that is characterized by its own point of view and a set of leading questions of its own. As a consequence, there are two different methods for approaching consciousness: the first-person approach and the third-person approach. In order to show this, a big part of this paper has been dedicated to the analysis of Dennett’s heterophenomenology. For actually, reducing the subjective to the objective is also what Dennett does, when he explains this method. What Dennett does not see is that the first-person perspective and the third-person perspective have different purposes, and so are led by different questions and interests. It is just this that makes that the former is substantially different from the latter, and that a first-person method is not a third-person method in disguise but that first- and third-person methods are independent though complementary methods for approaching the mind.


1Cf. Shoemaker 1996, Wright et al. 1998

2Cf. Stubenberg 1998, 39

3See also Dennett 1982 en 2001. His 2001 is a draft, but it is entirely in agreement with Dennett 1993a. Therefore, I’ll use it to a limited extent.

4Later, in his 2001, Dennett says that he unintentionally made a restriction here, and that "to these verbal reports most be added all the other manifestations of belief, conviction, expectation, fear, loathing, disgust, etc., including any and all internal conditions (e.g. brain activities, hormonal diffusion, heart rate changes, etc.) detectable by objective means". Now I think, just like Dennett, that "the other manifestations of belief, conviction, expectation, fear, loathing, disgust, etc." can be analysed equally well from the first-person perspective as the verbal reports. However, the "internal conditions" mentioned by Dennett are irrelevant to the first-person perspective, that is to phenomenology. From within, either they cannot be experienced at all or they cannot be experienced as such, but as certain feelings at most. A person can acquire knowledge about her internal conditions as such only by taking a third-person perspective towards herself. This applies also for the heartbeat that one feels. That it is the heart that makes this movement is no phenomenologically experienced knowledge, but it is something one has learned, or that one has inferred from observations of other people. Apart from that, my criticism will also hold, if Dennett would drop this extension of the scope of heterophenomenology. Therefore, I concentrate my analysis on verbal communication, as described by him.

5See for the "intentional stance" Dennett 1993b.

6In his 2001 Dennett even says that "heterophenomenology is nothing but good old 3rd-person scientific method".

7Cf. my remark about the heartbeat in note 4.

8Incidentally, I do not want to list all possible questions that can be asked from both perspectives. I want to show only that questions asked from these perspectives differ in their relations to the asking subject.

9See section 4.3.

10In his 2001 this is even clearer than in his 1993a.

11So not as a theoretical interpretation of first-person experiences.

12In this case the what-is-question is trivial. Of course, that need not be so, if the investigator is confronted with a new or not immediately recognized phenomenon, for example in astronomy. A what-is-question can also be an end in itself. Cf. Stegmüller 1983, 114.

13It is striking that for Wittgenstein the figure is a hare or a duck, while for Churchland it is a rabbit or a duck.

14Such a choice may be, and usually will be, implicit, if my own experiences are involved.

15Cf. also Apel 1979a

16I have developed this point in my 1996, chapter V, and my 2001.

17The thesis of complementarity of the first- and third-person perspectives is defended also by Velmans, although Velmans did not draw his inspiration from Habermas and Apel but from Russell. See Velmans 2000 and the references to his own articles there.


Apel, K-O 1979a, Die Erklären:Verstehen-Kontroverse in transzendentalpragmatischer Sicht, Frankfurt am Main: Suhrkamp Verlag

Apel, K-O 1979b, "Types of Social Science in the Light of Human Cognitive Interests", in: S.C. Brown ed., Philosophical Disputes in the Social Sciences, Brighton, Sussex etc.: Harvester Press etc., pp. 3-50

Bermúdez, José Luis 2000, The Paradox of Self-Consciousness, Cambridge, Mass. etc.: The MIT Press

Churchland, Paul 1995, The Engine of Reason, the Seat of the Soul, Cambridge, Mass. etc.: The MIT Press

Dennett, Daniel C. 1982, "How to study consciousness empirically or nothing comes to mind", Synthese 53, pp. 159-180

Dennett, Daniel C. 1993a [1991], Consciousness Explained, London etc.: Penguin Books

Dennett, Daniel C. 1993b [1987], The Intentional Stance, Cambridge, Mass. etc.: The MIT Press

Dennett, Daniel C. 2001, "The Fantasy of First-Person Science", on website: http://ase.tufts.edu/cogstud/papers/chalmersdeb3dft.htm

Giddens, Anthony 1986, New Rules of Sociological Method, London etc.: Hutchinson

Habermas, Jürgen 1982, Theorie des kommunikativen Handelns. Band I. Handlungsrationalität und gesellschaftliche Rationalisierung, Frankfurt am Main: Suhrkamp Verlag

Nagel, Thomas 2000 [1979], "What is it like to be a bat?", in: Mortal Questions, Cambridge, UK etc.: Cambridge University Press, pp. 165-180

Popper, Karl R. 1959, The Logic of Scientific Discovery, London etc.: Hutchinson

Shoemaker, Sydney 1984, "Self-reference and self-awareness", in: Identity, Cause, and Mind, Cambridge, UK etc.: Cambridge University Press, pp. 6-18

Shoemaker, Sydney 1996, The first-person perspective and other essays, Cambridge, UK etc.: Cambridge University Press

Stegmüller, Wolfgang 1983, Probleme und Resultate der Wissenschaftstheorie und Analytischen Philosophie, Band I. Erklärung – Begründung – Kausalität. Studienausgabe, Berlin etc.: Springer-Verlag

Stubenberg, Leopold 1998, Consciousness and qualia, Amsterdam etc.: John Benjamins Publishing Company

Velmans, Max 1991, "Consciousness From a First-Person Perspective", Behavioral and Brain Sciences 14 (4), pp. 702-719

Velmans, Max 2000, Understanding Consciousness, London etc.: Routledge.

Weg, Henk bij de 1996, De betekenis van zin voor het begrijpen van handelingen, Kampen: Kok Agora

Weg, Henk bij de 2001, "The Commonsense Conception and its Relation to Scientific Theory", Philosophical Explorations 4, pp. 17-30

Wright, Crispin, Barry C. Smith, Cynthia Macdonald (eds.) 1998, Knowing Our Own Minds, Oxford etc.: Clarendon Press


terug (Nederlands)

back (English)