FIRST-AND THIRD-PERSON APPROACHES : THE PROBLEM OF INTEGRATION

The author discusses the problem of integration of firstand third-person approaches in studying the human mind. She critically evaluates and compares various methodologies for studying and explaining conscious experience. Common strategies that apply reductive explanation seem to be unsatisfied for explaining experience and its subjective character. There were attempts to explain experience from the first-person point of view (introspectionism, philosophical phenomenology) but the results were not intersubjectively verifiable. Dennett proposed heterophenomenology as a scientifically viable alternative which supposed to bridge the gap between firstand third-person perspectives. The author critically evaluates his proposal and compares it to contemporary attempts to provide first-person methods.


INTRODUCTION
Cognitive science has a relatively short history, if compared to established scientific disciplines that are participating in this interdisciplinary science of the mind: neuroscience, psychology, computer science, linguistics, evolutionary biology, anthropology, educational sciences and philosophy.Josés Luiz Bermúdez stresses that "[T]he job of cognitive science is to provide a framework for bringing all these different perspectives together" [1; p.xviii].This is far from being an easy job, and cognitive scientists from different disciplinary backgrounds are dealing with concrete questions on how to combine different approaches and methodologies in their research.David Marr [2] famously proposed three levels of analysis (computational level, algorithmic/representational level and physical level) with the corresponding methods, and thus suggested the model of integration in classical cognitive science.Later approaches to cognitive modelling (connectionism, neural networks) and the much more important role of neuroscience reopens the problem of how to build coherent interdisciplinary theories.Valery Hardcastle in her book How to Build a Theory in Cognitive Science suggested that the answer to this conundrum revolves "around the common explanatory patterns one finds in cognitive science" and the shared interest in explaining how information is processed [3; p.9].She is aware that what counts as information and how different scientists explain it is not uniform across disciplines composing cognitive science.But it seems to me that if we stick with Bermúdez and his "short, but accurate, definition of cognitive science: Cognitive science is the science of the mind" [2; p.1], the problems of integration are much harder.Namely, many philosophers of mind and cognitive science (e.g.Nagel [4], Chalmers [5], Varela [6,7]) have pointed out that cognitive scientists, if they really want to explain the mental, they have to find the way how to deal with the experience and subjective aspects of consciousness.They emphasize that cognitive science has been neglecting the experiential.
Let me elaborate a bit on the philosophy of psychology from the historical and theoretical perspective.Elisabeth Valentine compares different frameworks in psychology with respect to three different aspects: the subject matter, the methods prescribed and the preferred theoretical analysis [8; p.128].She suggests that the subject matterconscious experience, behaviour and physiology -provide three different types of data and further determine the methods and theoretical analysis.She points out that the famous question "how to get inside the head" can be answered by three different methods roughly associated with three subject matters: asking (introspection) with conscious experience, guessing (inferences from behaviour) with behaviour and looking (neuroscience) with physiology [8; pp.128-129].Both behaviourism and cognitive science (classical and connectionist) are dealing with explaining behaviour, but they use different methods and theoretical analyses.Behaviourism is studying behaviour through objective observation and experiments using functional analysis in terms of stimulus-response relations.Classical and connectionist cognitive science study systems governing behaviour and use the computational model of mind, applying functional analysis and describing mental processes in terms of information processing [8; p.131].Valentine, with a very broad brush, suggests that "there are three main psychological approaches to explain mental phenomena: the experiential, the behavioural and the neurophysiological" [8; p.137].The experiential approach includes traditions of Verstehen, phenomenology, folk psychology and introspectionism.They tend to adopt a subjective and an individualistic approach with the focus on the subject's (first-person) perspective and explanation in terms of beliefs, desires, and reasons.The behavioural approach encompasses both behaviourism and cognitive psychology.It is based on the functionalist theory of mind and attempts to provide a causal analysis of the processes in the system.In contrast to the previous one, this approach is objective and general, as is the next, neurophysiological approach.At the level of neuroscience the task is to provide explanations in terms of physical embodiment.Daniel Dennett describes three different stances: intentional stance, design stance and physical stance [2], that roughly correspond to the division above.He argues that these stances help us explain and predict behaviour at different levels of abstraction, the physical stance at the most concrete level and intentional stance at the most abstract level, corresponding to folk psychology.He also suggests that with the method of heterophenomenology he is able to provide a scientifically respectable approach to conscious experience.Heterophenomenology has provoked a lot of discussion and disagreement, but before focusing on his method let us briefly look at Nagel's and Chalmers' understanding of conscious experience and the possible science of consciousness.

NAGEL AND CHALMERS ON CONSCIOUS EXPERIENCE
David Chalmers opens his paper "Facing Up the Problem of Consciousness" with the following observation: "Consciousness poses the most baffling problems in the science of the mind.There is nothing that we know more intimately than conscious experience, but there is nothing that is harder to explain.All sorts of mental phenomena have yielded to scientific investigation in recent years, but consciousness has stubbornly resisted" [9; p.200].What he sees as the most perplexing problemfamously hailed as "the hard problem of consciousness"is the problem of experience, the subjective aspect of mental processes.He argues that in contrast to the "easy problems of consciousness" (e.g. the reportability of mental states, the focus of attention, the deliberate control of behaviour etc.) which concern the explanation of cognitive abilities and functions, the hard problem is not a problem about the performance of functions.Easy problems may be solvable by the methods of cognitive science and neuroscience, but even if one manages to successfully explain the performance of cognitive or behavioural functions (e.g.verbal report, perceptual discrimination), Chalmers feels it will still be unclear as to why exactly the performance of these functions should be accompanied by experience [9].Or, in Thomas Nagel's words, "the fact that an organism has conscious experience at all means, basically, that there is something it is like to be that organism" [4; p.166].Nagel stresses the subjective character of conscious experience.Only creatures that undergo similar experiences can understand this "what is like to be" in an empathetic sense.So, famously, because of a differences in sensory apparatus and consequently in perception, there is no reason to believe that we can feel what is like to be a bat.Nagel argues that facts about consciousness can be only incompletely understood from an outside, third-person perspective.Knowledge gained from the external, objective, third-person perspective of the natural sciences or cognitive sciences would thus, according to Nagel, not suffice to understand what the bat can understand of its own experience from its internal firstperson subjective point of view.This epistemic form of subjectivity is associated with limits on the knowability and understandability about conscious experience [10].
We see that both authors emphasize the subjectivity of conscious experience and are concerned especially with the qualitative character of consciousness or qualia, sometimes also called phenomenal consciousness [11].According to Block, phenomenal consciousness properties include the experiential properties of sensations, feelings, perceptions, and also thoughts, wants and emotions.He maintains that these properties differ from any cognitive, intentional, or functional property belonging to what he calls access consciousness.Nagel, Chalmers and Block stress that feelings escape functional explanations of cognitive science.In the words of Joe Levine, there is the so-called "explanatory gap" [12] between causal explanation from the third-person perspective and the first-person experience of how it feels.
Nagel and Chalmers deny the possibility of a reductive explanation of conscious experience because reductive theories lack resources to give an answer to "what is like to be" [4] or the "hard problem" [9].But they both speculate about the possible solution.Chalmers argues that to account for conscious experience we would need "an extra ingredient" in the explanation.His suggestion is that we have to take experience as the fundamental feature of the world and construct a theory of experience with the aid of new psychophysical principles that will supplement physical theory.He further speculates about the double-aspect of information as the basic principle that might underlie and explain the emergence of experience from the physical.According to the double-aspect hypothesis, experience arises by virtue of its phenomenal aspect, while physical aspect is embodied in physical processing [9].Nagel ends his seminal paper "What Is It Like to be a Bat?" with the following speculative proposal: "It may be possible to approach the gap between subjective and objective from another direction.Setting aside temporarily the relation between the mind and the brain, we can pursue a more objective understanding of the mental in its own right.At present we are completely unequipped to think about the subjective character of experience without relying on the imagination -without taking up the point of view of the experiential subject.This should be regarded as a challenge to form new concepts and devise a new method -an objective phenomenology not dependent on empathy or imagination..... Apart from its own interest, a phenomenology that is in this sense objective may permit questions about the physical basis of experience to assume a more intelligible form.Aspects of subjective experience that admitted this kind of objective description might be better candidates for objective explanations of a more familiar sort" [4; pp.178-179].

METHODOLOGIES FOR STUDYING CONSCIOUS EXPERIENCE
So, are there any concrete proposals on how to scientifically study conscious experience and integrate the first-and third-person perspectives?Is it possible to improve the standard methods used in science or do we have to develop new ones?How are new attempts connected to the more traditional approaches to studying experience?Let me start with Dennett's illustration of two opposing teams tackling the problem of explaining experience.Dennett starts off with the two questions, posed by his colleague James Conant [13; p.1]: Descartes: How is it possible for me to tell whether a thought of mine is true or false, perception or dream?Kant: How is it possible for something even to be a thought (of mine)?What are the conditions for the possibility of experience (veridical or illusory) at all? Dennett adds a third version of the question: Turing: How could we make a robot that had thoughts, that learned from "experience" (interacting with the world) and used what it learned the way we can do?Dennett suggests there are two main reactions to Turing's proposal to trade in Kant's question for him: (A) Cool! Turing has found a way to actually answer Kant's question!(B) Aaaargh!Don't fall for it!You're leaving out . . .experience! Dennett declares himself as a captain of the A team (together with Quine, Rorty, Hofstadter, the Churchlands, Andy Clark and others), while Chalmers is a supposed captain of the B team.He thinks that his team will win, but admits that it will not be an easy task.It will take "a rather remarkable exercise of the imagination to see how it might even be possible", but following Turing's insight on how to recast Kant's question as an "engineering question" he thinks it will be possible to "trade in the first-person perspective of Descartes and Kant for the third-person perspective of the natural sciences and answer all the questions-without philosophically significant residue" [13; p.1].On the other side, Chalmers, together with Nagel, Searle, Levine and others, insists that he just knows that A team leaves out consciousness and does not address the hard problem.
So, let us look first at Dennett's heterophenomenology, and then turn to first person methods in the science of consciousness.

HETEROPHENOMENOLOGY
Dennett develops heterophenomenology as a method for studying consciousness and describes it in the 4 th chapter of Consciousness Explained as "the neutral path leading from objective physical science and its insistence on the third-person point of view, to a method of phenomenological description that can (in principle) do justice to the most private and ineffable subjective experiences, while never abandoning the methodological principles of science" [8; p.72].In his later paper, he characterized it as "a bridgethe bridgebetween the subjectivity of human consciousness and the natural sciences" [14; p.249].He thinks that heterophenomenology preserves the point of view of the subject (first-person perspective) and then conveys it to science (third-person perspective).He argues that it takes the subjects seriously, but without granting them infallibility, in contrast to the Cartesian tradition which he calls a "lone-wolf autophenomenology".The distinguishing character of the method is neutrality, a kind of agnosticism, "a deliberate bracketing of the issue of whether what they are saying is literally true, metaphorically true, true under-an-imposed-interpretation, or systematically-false-in-a-way-we-must-explain" [14; p.252] that is contrary to what we are used to in our everyday interpersonal communication.
The investigator starts the research by extracting verbal utterances that are transcribed and function as verbal reports.The method, however, is not limited solely to verbal reports, but may also include other types of data, such as behavioral reactions, visceral reactions, hormonal reactions, and other changes in physical states that are detectable by objective means.Dennett points out that the investigator has to be particularly cautious with verbal reports, since they require interpretation and assessment of speech acts as expressions of beliefs about a subject's subjective state.Verbal reports represent the most critical part and require the employment of the intentional stance as well as the move from raw data to interpreted data -subject's heterophenomenological word [13].Dennett stresses that his fictional world is populated with all the images, events, sounds, smells, hunches, presentiments, and feelings that the subject (apparently) sincerely believes to exist in his or her (or its) stream of consciousness.Maximally extended, it is a neutral portrayal of exactly what it is like to be that subject-in the subject's own terms, given the best interpretation we can muster" [2; p.98].
Dennett suggests that heterophenomenologists can, by carefully designing their investigations, bring data from the first-person point of view to objective science.He feels he has not discovered a new method, but has merely described and explained it.He considers it to be a "good old 3 rd -person scientific method applied to the particular phenomena of human (and animal) consciousness" [13].Scientists in various disciplines (e.g.psychophysicists, cognitive psychologists) that intend to study consciousness in a scientific way have already used it.Heterophenomenology also takes some features from philosophical phenomenological tradition (Brentano, Husserl) but in a naturalized variant.What is the exact relationship between Dennett's heterophenomenology and classical philosophical phenomenology is the topic of much heated discussion [15].But before that, let us give a word to the members of the team B who believe that heterophenomenologists leave out what they need to explainthe subjective experience.

FIRST PERSON METHODS
On the first sight Chalmers has radically different views about possible methods for the science of consciousness then Dennett.He argues that the job of a science of consciousness is to connect first-person data to third-person data [16].The latter are obtained by investigating processes like behaviour, brain processes and environmental interactions that are accessible by known scientific methods.First person data are about conscious experience an include those concerning visual experience (e.g. the experience of colour), other perceptual experiences (e.g.auditory, tactile), bodily experience (e.g.pain), mental imagery (e.g.recalled visual images), emotional experience (e.g.happiness, fear), and concurrent thought (e.g. the experience of deciding) [17].Chalmers takes for granted that there are first-person datait is a fact about our minds, and that "our direct knowledge of subjective experiences stems from our first-person access to them" [16].He has argued that reductive strategies to explain conscious experiences are doomed to fail [5,9].Even if scientist find out the complete functional explanation there will still remain the question why is this functioning associated with the particular subjective experience.First-person data are not data about objective functioning.Scientists are thus facing the problem to find good methodologies for collecting the data (both first-and third-person), express them in suitable language and find connecting principles.Chalmers' goal is to find "fundamental theory of consciousness", i.e. formulate simple and universal laws that underlie these connecting principles [16,17].In order to possibly achieve the goal scientists have to develop methods in both domains.
There was a fascinating development of methods in psychology and especially in neuroscience (e.g.brain imaging, single cell studies) in last few decades, as well as improvement in expressing data (e.g.computational models, statistics).It therefore seems that science about the third-person domain is well equipped.But what about the first-person domain?Do we have well developed methods for gathering first-person data?There were traditions in the experiential approach, especially introspectionism and philosophical phenomenology that nearly disappear from scientific investigation of the mind.But they are now coming back, transformed and accompanied with many new first-person approaches to the study of consciousness, including those, based on a Eastern meditative tradition [18].Nevertheless we have to take into account the well known obstacles: the lack of incorrigible access to experience, the changing of experience while we self-observe and the difficulties with expressing in language.
What I find perplexing is how two approaches, introspectionism and phenomenology are sometimes merged in new first person methods.I think the names phenomenal consciousness and phenomenal properties may provoke some misunderstanding by conflating qualitative and phenomenal structure [10].According to phenomenologists from Brentano on, the phenomenal structure of experience is much more then raw feels and covers the domain of the world as it appears to us.It involves not only sensory qualities but also intentionality.Introspection was introduced by Wilhelm Wundt as a psychological method where "one attends carefully to one's own sensation and reports them as objectively as possible" [19, p.103].This means that one describes the felt sensation and not the stimulus that provoke it.In a way, the introspectionists were studying the elements of sensation and looking for the basic constituents of mental states.On the other hand, for Brentano, the early phenomenologist, psychology starts with the mind-an active, creative entity which has intentions, for it implies and demands an object.The true subject matter of psychology is the mental act-such as judging, sensing, imagining, or hearing, each of which reflects a sense of direction and purpose.One cannot simply see; one must see something; and the act of seeing something is psychological or mental.Given this viewpoint, the task of empirical psychology is to study the mind of the agent at work, dealing with objects, purposes, and goals.... one cannot conceive of thoughts and judgement, let alone study them, except by taking into account one's inner phenomenal experience.And this can be accessed not by prompted introspection for one cannot observe at the same time that one experiences-but rather by simple phenomenal experience of one's inner mental life" [19; pp.101-102].
We can clearly see the difference between bottom-up approach of introspectionism and topdown phenomenological approach of Brentano.As Howard Gardner has pointed out, his topdown concerns are emerging in different forms, also in the view of the computer as an agent with plans, intentions, and goals [19; p.102].On the other side, we can understand why with the more accurate measurements of brain activity neuroscientists are interested in more finegrained descriptions of experience that can be linked to neural correlates.As Nahmias reflects: "Given the new tools we have to test and correlate the conscious experiences reported by subjects, should we shake off the shackles of behaviorism and reconsider some of the introspectionists' methods and goals?Specifically, might it be worthwhile (1) to try to train subjects to attend more closely to their experiences and describe them more fully and accurately; (2) to try to develop a more precise language with which subjects can report the contents of conscious experience; and ultimately, (3) to try to map out the internal structure of conscious experience to better understand its relations to neural processes" [20; p.12].
Let me illustrate the idea of correlating subjective experience and neural state with an experiment done by neuroscientist Antonio Damasio and his colleagues [21,22] .They set about to study the neurobiological substrates of feelings by connecting first-person experiences with third-person data obtained by modern imaging techniques.First, he and his team made sure that they were actually capable of triggering emotions in their test subjects.They asked them to recall a certain emotional situation in which they experienced one of the four feelings: joy, sorrow, fear or anger, and then to re-experience this emotion as vividly as possible.The experiment was then taken to the second room where test subjects were placed in a PET scanner.They were asked to imagine the past experience again and then signal with their hands when the required emotion was experienced.At that point the activity of their brains was recorded with the PET scanner, which eventually provided appropriate brain correlates for certain emotive states.The obtained results corroborated their hypothesis that in experiencing emotions there was high neural activity in the somatosensory parts of the brain (cingulate cortes, insula and SII, hypothalamus and several nuclei in the brain stem tegmentum), and at the same time activation/deactivation patterns for different feelings differed substantially.So, just as we can perceive that our bodies are in different states when we experience fear or joy, Damasio and his colleagues successfully showed that brain maps corresponding to these feelings also differ.
But is this approach really so different from Dennett's heterophenomenology?I think that Dennett would accept experiments like Damasio's without problems.The important difference between hetrophenomenology and first person methodologies described by Varela and Shear [18] is in the preparation of persons that are doing first person research.Persons doing introspection or phenomenological inquiry are trained to do this (see also Kordeš, this issue), while Dennett rely on ordinary people.He thus suggests that "if some of your conscious experiences occur unbeknownst to you (if they are experiences about which you have no beliefs, and hence can make no "verbal judgments"), then they are just as inaccessible to your first-person point of view as they are to heterophenomenology. Ex hypothesi, you don't even suspect you have themif you did, you could verbally express those suspicions.So heterophenomenology's list of primary data doesn't leave out any conscious experiences you know of, or even have any first-person inklings about" [13].On the other side at least some first person methods are concerned with pre-reflexive experience and see the role of introspection in explicating what was only implicitly present.Since this part of our knowledge is non-conscious and thus difficult to get at, Pierre Vermersch and Claire Petitmengin-Peugeot propose the explication session with an interviewer.Interviewer guides the subject through three stages: gathering descriptions of experiences, analysis and modelling of the descriptions and comparing the established models.The accessing the experience can be relieving a past experience or living the experience 'in the present [23, 24; p.46].In seems that guidance and help from well trained interviewersecond-person, is crucial for such an investigation and that "naive" introspection will not lead to interesting results.But who wants to stick to naïve introspection if there are better methods.

CONCLUSION
Common strategy in science is to apply reductive explanation but it is highly problematic if the same strategy can be used also in cognitive science, particularly in science of consciousness.Nagel and Chalmers have argued that because of the special qualitative character that accompany experience which is always subjective and from first-person perspective, strategies for explaining experiences must be different.We know from the beginning of psychology that there were attempts for a scientific approach to explain experience (introspectionism, philosophical phenomenology).Because the results were not intersubjectively verifiable scientist were highly sceptical about such methods.Dennett proposed heterophenomenology as a scientifically viable alternative, which supposed to be a bridge between first-and third-person perspectives.I think this method is quite similar to some new introspective methods although the ideology behind often points otherwise.I understand all of them more pragmatically as tools for obtaining more systematic and deeper knowledge of our mental life that can be possibly correlated with the functioning neural system but I doubt that any of the proposed methods can help Chalmers to solve the hard problem.