Arbeitseinheit 14 - Physiologische Psychologie
 
M. Piefke
Approaches to the Self: Views from Philosophy, Social Psychology,
and the Neurosciences
In my talk I will focus on different approaches to the
self, a notion which is subject to ongoing interdisciplinary debates, particularly
in philosophy, social psychology, and the neurosciences.
I
will first turn to philosophical concepts of the minimal or momentary
self as supposed by William James and Galen Strawson, which are very similar
to the concept of self which Jun Tani referred to in the morning and which
he employs in his work with robotics. Specifically, I will depict the
so called immunity principle which characterizes a basic implication of
the use of the first-person pronoun, and I will illustrate the distinction
between the sense of self-agency and the sense of self-ownership. Subsequently,
I will attempt to apply the model of the distinction between self-agency
and self-ownership to the disruption of self-referential information processing
observed in schizophrenic patients. Secondly, I will focus on similarities
and differences between concepts of the extended or narrative self and
their theoretical implications, in particular with regard to the debate
on fictional and non-fictional aspects of the extended self. The interrelations
between the narrative self and episodic autobiographical memory will be
illustrated. Assumptions on the neural underpinnings of episodic autobiographical
memory and taking a self-perspective will be introduced on the basis of
recent functional neuroimaging data. Eventually, interrelations between
the narrative self and the minimal self will be discussed with respect
to future perspectives of interdisciplinary research on the characteristics
of the self.
The Minimal or Momentary Self
The
minimal self is unextended in time, that is, it is a momentary self which
is limited to what is accessible to immediate self-consciousness. The
minimal self must not be aware of having self-experiences, however, it
involves the sense of self-ownership and the sense of self-agency in the
context of both motor action and cognition. I will describe this distinction
in detail, but prior to this, I want to outline a further element which
is essential to the definition of the minimal self: it concerns an implication
of the use of the first-person pronoun in self-reference. The use of the
first-person pronoun in self-reference is immune to error through misidentification.
For example, when I say "I think it is raining outside" I certainly can
be wrong about the rain but not about the I. That is, I cannot misidentify
myself when I state that it is me who is thinking. This is referred to
as the immunity principle. Thus, access to myself in first-person
experience is immediate and non-observational. It is the immediate self
that is referred to here, and this immediate self is the pre-reflective
point for origin of action, experience, and thought. Now, there are two
closely related aspects of minimal self-awareness: these are the sense
of self-agency and the sense of self-ownership. According to
Gallagher (2000), the sense of self-agency is characterized by the sense
that I am generating or causing an action, that I am the initiator or
the source of the action. For example, the sense that I am the one who
is causing something to move, or the sense that I am the one who is generating
a certain thought in my stream of consciousness represent the sense of
self-agency. In contrast, the sense of self-ownership is characterized
by the sense that I am the one who is undergoing an experience. This would
be the case when I sense that my body is moving regardless of whether
the movement is voluntary or involuntary. In a review of philosophical
conceptions of the self, Shaun Gallagher (2000) proposed a neurocognitive
model of immediate self-awareness based on the distinction between the
senses of self-agency and self-ownership. Referring to this model, he
attempts to comprehend disruptions of self-referential information processing
observed in schizophrenia. In patients with schizophrenia, symptoms like
thought insertion, delusions of control, and auditory hallucinations are
assumed to be based on a dysfunction of self-monitoring processes (Feinberg,
1978; Frith, 1992). Gallagher (2000) supposes the senses of self-ownership
and self-agency to coincide and to be indistinguishable in normal voluntary
or willed action and thus specifies the dysfunction of self-monitoring
processes seen in schizophrenia as a disruption of the sense of self-agency
with normal functioning of the sense of self-ownership. In cases of delusions
of control or thought insertion, a schizophrenic patient would have the
sense that it is his body that is moving or that a certain thought occurs
in his stream of consciousness, that is, he obviously has a sense of self-ownership.
However, the movement or the thought would not seem to him to be self-generated,
but rather would appear to be alien and out of the subject's control.
This might be referred to as a disruption of the sense of self-agency
which is assumed to be based on a dysfunction of neuronal processes in
several areas of the prefrontal cortex (Daprati et al., 1997; Frith, 1992).
The Extended or Narrative Self
I
will now turn to the extended or narrative self. The extended or narrative
self is extended in time, that is, it includes memories of the past and
intentions towards the future. It is supposed to be a more or less coherent
self or self-image which is constituted through narration, that is, through
the various stories we tell about ourselves. Importantly, it is strongly
associated with episodic autobiographical memory. The narrative self establishes
continuity between our past and our future. It is a self with individual
identity which encompasses continuous experience. On the background of
the aforementioned model of the minimal self this conception of the narrative
self raises questions concerning the relation between the minimal and
the narrative self: Do we have to comprehend the extended self a succession
of momentary minimal selves? Or can we assume minimal selves as being
abstractions from a more substantial continuity that is the more genuine
self? One early and famous answer to these questions was given by David
Hume (1739) in his "Treatise on Human Nature". According to Hume, the
extended self is a bundle of impressions strung together by the imagination.
Thus, the extended self is a fiction, albeit a useful one since it lends
a practical continuity to life. The narrative theory of the self is a
contemporary reading of this view. One version of the narrative theory
of the extended self was supposed by Dennett (1991). Referring to the
finding that neuronal processing is distributed across a variety of brain
regions he claims that we cannot suppose a real "neuronal center of experience"
Thus, he concludes that there can be no real simplicity of experience
at one time and no real identity across time that might be labeled the
self. According to Dennett, we could at best refer to a minimal biological
self as something real, but this is nothing more than a principle of organization
which involves the distinction between self and non-self and is found
throughout living nature. Obviously, this is not sufficient for the generation
of coherence and continuity across time observed at the level of human
experience. It is language which enables us to render our experiences
coherent and continuous across extended time periods. By means of telling
stories we create what we call our selves and extend our biological boundaries
to encompass a life of meaningful experience. Dennett defines the self
as an "abstract center of narrative gravity". Thus, according to this
definition, an individual self consists of the abstract and moveable point
where the various stories that the individual tells about himself, or
that others tell about him, meet up. This view at least has two basic
implications: Firstly, we cannot prevent ourselves from "inventing" our
selves, and secondly, the self is nothing substantially real, but rather
an empty abstraction. This figure illustrates Dennett's view of the self
as the center of narrative gravity built up at the point were the various
stories referencing an individual meet up. This example which is reduced
of the high complexity typically characterizing self-referential narratives,
depicts the self as being composed of the narrative of being a spouse,
the stories I tell about myself, the narrative of being a parent, the
narrative of being a sports fanatic, the stories that others tell about
me, and the narrative of being a teacher. A related view of the self was
suggested in the context of neuroscience. However, this view from neuroscience
has lead to conclusions which are different from Dennett's conclusions
in at least one important aspect. It was Michael Gazzaniga (1995, 1998)
who introduced the model of the "left hemisphere interpreter" in the context
of ongoing debates on the phenomenon of consciousness and interpretative
capacities of the human brain. Gazzaniga's basic assumption suggests that
one function of the brain's left hemisphere is to generate narratives
by means of an "interpreter". This "interpreter" is supposed to be some
kind of capacity, or a set of mechanisms, residing in the left hemisphere
of the human brain. It is assumed to weave together autobiographical facts
and inventive fiction in order to produce a personal narrative that enables
the sense of a continuous self. Gazzaniga concludes from his model of
the "left hemisphere interpreter" that the self is not a fiction since
the normal functioning of the "brain's interpreter" tries to make sense
of what actually happens to the individual. Nonetheless, he has to concede
that the narrative self in the non-pathological case might be a "bit fictional"
(Gazzaniga & Gallagher, 1998, p. 713) since we cannot prevent ourselves
from enhancing our personal narratives with elements that smooth over
discontinuities and discrepancies in our self-constitution. This view
is in good accordance with approaches to the narrative self developed
in social psychology. The statement of Fivush (1998) characterizes the
basic assumptions generated in social psychology with respect to the interrelations
of the narrative self and episodic autobiographical memory. The generation
of continuity along the axis of one's life and the generation of a more
or less coherent self-image are supposed to be accomplished by means of
storytelling in social contexts. It is some kind of conversational remembering
(Edwards & Potter, 1991) of personal past experiences which is assumed
to allow for the unfolding of the extended self. The extended self is
constantly re-interpreted by the narrative process which establishes coherence
with respect to at least four dimensions: temporal coherence, spatial
coherence, thematic coherence, and social cultural coherence. From this
approach, social psychologists generally conclude that both the narrative
self and the personal life history are fictional. However, one can argue
that this conclusion does not take into consideration the role of the
proper functioning of episodic autobiographical memory as a necessary
condition for non-fictional aspects of the narrative self. Firstly, episodic
autobiographical memory is involved in providing a sense of time: It allows
for mental time travel (Tulving & Markowitsch, 1998), that is, it
provides us with the ability to mentally travel back to our personal past
experiences. Secondly, it has a direct link into emotion processing. Thus,
it comprises the emotional evaluation of personal life experiences with
respect to our selves. Additionally, Tulving (1995) proposed that episodic
autobiographical memory is associated with a specific kind of consciousness
which he named "autonoetic consciousness". The notion of autonoetic consciousness
depicts the form of self-awareness typically accompanying episodic autobiographical
memory retrieval and which is assumed to be contingent on brain structures
of the prefrontal cortex. Studies of amnesic patients have repeatedly
demonstrated that specific impairments of episodic autobiographical memory
- in particular those resulting from damage to areas of the prefrontal
cortex and limbic structures - can lead to changes in the patients' personality
and self-concepts, to a deficient ability of temporal sequencing and emotion
processing, or even to an inability to develop and maintain some kind
of stable self-referential cognition. Obviously, impaired functioning
of episodic autobiographical memory affects our ability to generate a
more or less coherent self-structure or self-image which is extended in
time as it includes a sense of a continuity between our past and our future.
In a recent functional neuroimaging study, Vogeley et al. (2001) aimed
at elucidating the neural mechanisms of theory of mind and taking a self-perspective
using functional magnetic resonance imaging (fMRI). Increased neural activity
associated with taking a self-perspective was found in the medial prefrontal
cortex (the frontal pole) and in the right temporoparietal junction area.
Our own fMRI data revealed activation of the frontal pole to be related
to the retrieval of positively valenced autobiographical memories. Can
we assume activation of this area during the recollection of positively
valenced personal past episodes (compared to the retrieval of negatively
valenced memories) to be associated with taking a self-perspective, as
well? We then would have to ask why positively valenced memories should
be stronger associated with taking a self-perspective than negatively
valenced memories. One highly speculative explanation would suggest that
the subjects' identification with positive autobiographical memories is
per se stronger than identification with negative ones. Possibly, remembering
positively valenced personal past experiences might be associated with
an immediate, pre-reflective, and non-observational access to the self.
In contrast, access to negatively valenced autobiographical memories might
be not as immediate as access to positive ones but rather could imply
a conscious act of reflection associated with some kind of an observer
perspective leading to a less pronounced formation of a self-perspective.
Based on our functional neuroimaging data we can assume that negative
autobiographical memories are kept more distant than positive ones. Interestingly,
additional increased neural activity associated with positively valenced
autobiographical memories was observed in limbic regions - in particular
in the amygdala and the hippocampus - and in the retrosplenial cortex.
These activations are supposed to indicate a higher level of emotional
re-experiencing for positive than for negative memories, thus, supporting
the view that recollection of positively valenced autobiographical memories
is associated with an immediate and emotional pre-reflective access to
the self whereas recollection of negative memories are rather kept distant.
The formation of some kind of observational distance during retrieval
of negative memories might be suggested to prevent the pronounced specification
of a self-perspective as seen during the recollection of positively valenced
personal past experiences. Finally, I want to discuss the interrelations
between the minimal ('core') self and the narrative ('autobiographical')
self. Damasio (1999) captured some of the difficulties one has to face
while trying to find suitable expressions for the interrelations between
the minimal and the narrative self. These difficulties are primarily due
to complexities characterizing the interconnections among episodic autobiographical
memory and the narrative self. Episodic memory obviously is indispensable
for the construction of the extended narrative self. However, episodic
autobiographical memory is subject to constant remodeling under the influence
of factors that include both an individual's acquired and innate dispositions.
Additionally, episodic autobiographical memory is affected by the narrative
process itself and the social and cultural environments an individual
lives in. Nonetheless, the registration of episodic autobiographical memory
as my memory of myself is contingent on a minimal but constantly reiterated
sense of self that I recognize, without error, as myself. In some respects
this may depend on narrowly defined, embodied capabilities and feelings.
In other regards, we have to take into account that the core features
of the minimal self might be constantly re-interpreted by the narrative
process, as well. In an essay on the biology of consciousness, Damasio
(1998) refers to these issues in terms of a distinction between a "core
consciousness" and an "extended consciousness" which has highly similar
implications than that between the minimal self and the narrative self.
According to Damasio's approach, the extended consciousness also depends
on the core consciousness. However, both forms of consciousness principally
perform precisely the same mental operations: They place mental contents
into an individual perspective, confer ownership to such contents, and
give the owner the sense of an ability to act on them. Obviously, the
conceptions of the self developed by different researchers in a variety
of different research disciplines are characterized by a number of differences
and similarities. However - as I attempted to illustrate in my talk -
these conceptions coincide with regard to several main structural and
theoretical implications of the self. Thus, interdisciplinary research
between philosophers, psychologists, and neuroscientists might open new
ways for the development of deeper insights and more sophisticated conceptions
of self-referential information processing.
|
Navigation
Seitenanfang, Arbeitseinheit 14,
Abteilung für Psychologie,
Universität Bielefeld, Suchen
|