What is the self? As an academic psychologist who studies self-concept, I recently gave this question some serious thought while writing a paper with my graduate students over the holidays, and I was surprised at how difficult it is to answer this seemingly simple question.
To be clear, there are many facets of the self that are under the purview of others ranging from philosophers to anthropologists to clergy. But as a psychologist whose research speciality is "the self," it's stunning to think about how frequently we throw around terms like self-esteem, self-concept, self-enhancement, and self-control yet never spell out exactly what this "self" is in each of those terms.
One answer: The self is memory
As a psychologist, my first answer to this age-old question is that the self is memory. Our mind, its contents, and (perhaps as importantly) the organization of those contents are key for understanding what the self is.
We all take for granted our ability to rattle off our names, first loves, core personality traits, home address, and daily work routine quickly and accurately without any effort. We know our current goals, we know our past histories, and our memory is the maestro that conducts a concert involving our past and present as we pursue our goals into some anticipated hypothetical future.
Our social connections are woven into our self-concept, and there is considerable research that shows important interpersonal relationships, group memberships, and social roles are strongly associated with the self in memory. In addition, there is a lot of work that indicates our goals are an integral part of our self-concepts and that we often compare our current state to these goal selves to assess our progress (or lack thereof) in achieving our aspirations and meeting our obligations.
But still not a complete answer
Everything described above (and more) represents my standard, stock answer to the question of "what is the self." In fact, my colleagues and I have published many papers that echo the themes above. And they are important themes. Yet, recent events in my life have led me to revisit this pat response.
Dealing with loved ones suffering from Alzheimer's dementia has led me to think a bit more comprehensively about the issue of what constitutes the self. At first blush, observing people suffering from significant memory loss would seem to simply reaffirm the stock answer. Indeed, losing one's memory leads to irrevocable changes in one's personality, behavior, and ability to achieve goals. This is certainly true, but it is also too simplistic of an analysis.
Evidence from the neuroscience literature
Some of the most interesting, relevant work I have encountered lately has involved studies of patients who have experienced significant memory disruptions (e.g., amnesias, Alzheimer's dementia). Findings from these neuropsychological studies suggest that self-knowledge is the product of multiple, interacting systems involving general and self-specific memory, and different systems of memory for one's past history of specific events (episodic memory) and for summary information not tethered to particular episodes from one's life (semantic memory).
For instance, Klein and Gangi (2010) discuss several patients who have experienced significant memory disruptions. In a number of cases, although episodic memory for oneself can be compromised, patients can often accurately report on their own personalities. For example, although a woman suffering from Alzheimer's dementia may not be able to remember any times when she's acted in a gregarious and outgoing fashion, she may be quite aware (and be accurate in her knowledge) that she is extraverted, and her sense of her extraversion conforms to multiple, independent indicators of this personality trait (e.g., reports of her friends and family).
Interestingly, these neurological case studies suggest that although one's awareness of overarching trait summaries may be accurate, the updating of these summaries often becomes "stuck" at a point where one's memory becomes impacted. Further, it appears that one's own general summary knowledge of their own traits and attributes is unique to the self -- many patients can accurately report on their own personalities (at least, their own premorbid qualities) but reveal great difficulty in describing the personalities of very close others (e.g., one's own daughter). In short, it appears that summary knowledge for the self such as an awareness of one's own abstract qualities is initially based upon episodic memories but is ultimately represented in a separate store from other social knowledge that is less subject to compromise in many cases.
A number of caveats should be acknowledged at this point. First, case studies are idiosyncratic and not experiments, thus any conclusions are subject to a variety of alternative accounts. Yet, when we see an array of striking dissociations in memory loss patients (e.g., amnesiacs, dementia) where episodic memory for the self is compromised but abstract knowledge for the self is not (while both forms of knowledge are compromised for others), it suggests that the "self is memory" answer is more nuanced and complex. Our sense of self may be the product of our memories, but even when episodic knowledge is compromised, some self-knowledge persists and can remain accurate (at least for a time).
Although I am more certain than ever that a "self is memory" answer is accurate, I am also increasingly convinced that it is an incomplete answer when one assumes that "memory" is one, large, homogenous system. Recent findings from the neuroscience literature is beginning to paint a picture of the self as a product of complex, interrelated systems of memory. Although adding new complexity to the issue, this work is now beginning to shed important light on age-old questions about the nature of the self and its function in guiding human behavior.