In this note I review Wilson and Ross (2003). I critique Wilson and Ross in light of Arbuthnott, Arbuthnott and Thompson (2006); Schacter (1996); and Bechara, Tranel, Damasio, Adolphs, Rockland and Damasio (1995). I then attempt to discern if there is a role in Wilson and Ross’ schematic for attachment theory as developed by Ainsworth and Bowlby (1991).
Summary of Wilson and Ross
Autobiographical memory and the self. Wilson and Ross (2003) explore the interaction between one’s autobiographical memory and one’s construction of a sense of self. Since they do not explicitly define either, one must turn elsewhere to attempt to discern their meaning. Autobiographical memory “refers to the store of memories of events that have happened to an individual” (Rubin, 2006). Although philosophers and psychologists have argued for centuries about what the self is, recent speculation is to the effect that there “may not be a single thing to be described.” Rather, the self “is a multiplicity of related, yet separable, processes and contents” (Klein & Gangi, 2010, p. 2). These include “episodic memories of one’s life events;” “semantic summary representations of one’s personality traits;” “semantic knowledge of facts about one’s life;” “an experience of continuity through time;” “a sense of personal agency and ownership;” “the ability to self-reflect” (self-consciousness); and the “physical self” (one’s body) (Klein & Gangi, 2010, p. 2).
Wilson and Ross’ (2003) basic theory is that there is a reciprocal relationship between one’s autobiographical memory and one’s sense of self, in that “people’s recollections influence their self-views and vice versa” (Wilson & Ross, 2003, p. 137). Given this bivalency, one’s current self-appraisals should influence how one reconstructs one’s past; and one’s past in turn should influence one’s current conception of whom one is.
It is important to understand Wilson and Ross’ framework clearly. One exists at a current moment in space and time. At that moment, one has a certain concept of one’s self. One performs an act of reflection, traveling backwards in time to a discrete temporal moment or interval. One then characterizes that past interval, based on how one conceives of oneself now, in the present. The past interval then in turn affects one’s present characterization of one’s current identity. The moment of the now, of course, itself constantly is moving forward in time; as I write this sentence, it is later than it was when I started writing this paragraph.
For clarity, I divide Wilson and Ross’ (2003) argument into two distinct propositions. The first is that sense of self affects autobiographical memory; and the second is that autobiographical memory affects sense of self. I will examine each, followed by a critique of both.
Sense of self affects autobiographical memory. As to their first proposition, Wilson and Ross (2003) characterize the “received view” as premised on a “consistency bias,” in that one who believes one’s personality is firmly established tends to think of one’s past as a stable progression of events through time. Although not cited by Wilson and Ross, this idea is based on earlier work by Maylor, Chater and Brown (2001). Wilson and Ross attempt to refute Maylor et al.’s theory by arguing one’s concept of one’s identity changes. In fact, it changes for the better. Across various samples, “people reported their past selves to be inferior to their present self” (Wilson & Ross, 2003, p. 139).
There is some expansion to this process, in that one does not experience time as a discrete sequence of instants, each of equal duration. Rather, one’s phenomenological perception of time is elastic and enlarges in response to positive self-evaluation. Thus, “to protect their current [positive] self-regard, people are motivated to feel farther from past failings than from achievements, even when calendar time does not differ (Wilson & Ross, 2003, p. 141). Older memories of past failures are “less vivid” and typically phrased in the third person, as opposed to pleasant memories, which typically are phrased in the first person (Wilson & Ross, 2003, p. 144).
This facilitates a process of reattribution. One can assign past transgressions to a former self – an “old” me – thereby relieving oneself of personal responsibility and blending in more smoothly with one’s social environment, or coping with it with improved dexterity and facility. It also has considerable psychological utility, because “people’s moods and reports of life satisfaction improve when they recall pleasant personal experiences and worsen when they remember distressing personal episodes” (Wilson & Ross, 2003, p. 145).
Autobiographical memory affects sense of self. As to their second proposition, Wilson and Ross (2003) assert, “autobiographical memories influence current self-views” (Wilson & Ross, 2003, p. 142). It is somewhat more confusing to understand just how this works. Wilson and Ross describe a previous study they had conducted, in which a sample group was presented with a time-line spanning many years from “now” to the distant past. Participants were asked to situate a specific pleasantly-remembered event somewhere along the time line. The time-line then was re-scaled to span only recent years, and participants again were asked to situate the remembered event. These two trials then were repeated, only this time participants were asked to situate a specific unpleasant remembered event somewhere along both lines.
Wilson and Ross found that people situated the event much closer to “now” when the time-line spanned only a few years, as opposed to when it spanned many years. Further, using an undisclosed psychometric measure, Wilson and Ross found people who situated themselves closer to failures tended to have less-favorable current self-appraisals, and vice versa. On this basis, Wilson and Ross concluded that remembered outcomes affect present self-regard.
Problems with Wilson and Ross
Wilson and Ross (2003) do not explore several correlates to their theory, which tend to invalidate or at least severely challenge it. Some of these issues are raised by, variously, Arbuthnott et al. (2006); Schacter (1996); and Bechara et al. (1995), among others. I will review six of these issues.
Negative self-concept. In their first proposition – that sense of self affects autobiographical memory – Wilson and Ross (2003) only consider the situation where one’s concept of self improves on reflection. What happens in the reverse situation – that is, when one’s concept of self deteriorates? Wilson and Ross assume one in effect becomes progressively more content with one’s sense of self and that as one does so, one feels closer to past achievements and further away from past failures. It is entirely possible, though, that one might view one’s former self in a better light than one’s current self. This should entail that one would feel closer to one’s past failings and further away from one’s past achievements.
It is not difficult to imagine a situation where one’s present recollection of an early memory might shift in one direction (positive) at one point in time, then shift in the opposite direction (negative) at another. This process might occur in real time, during a spontaneous and perhaps non-conscious moment of reflection. It also might transpire, more deliberately, during the therapeutic process. In a kind of pretest-posttest, one subject, pre-experimental research design, a clinician might record a patient’s early autobiographical memories. Then, later in therapy, the content of those early autobiographical memories, together with their associated emotional affect, might be reconsidered. This comparison might reveal nuanced shifts, subtle changes in details, extraneous commentary, and other perspectives, which then can be assessed. They might demonstrate a transition towards a more positive self-concept; or, a deterioration. Something like this takes place during Adlerian therapy. Wilson and Ross, however, do not explore this hypothetical.
Although they characterize it somewhat differently, Arbuthnott et al (2006) do so. Rather than speculating about the human condition in general, their approach is to focus more on specific clinical psychopathologies. They develop several examples. (1) Persons with affective disorders may be unable to differentiate autobiographical memory (recalling its semantic content) from the associated emotions it provokes. The memory’s affective content in turn may exacerbate, or retard, its impact. This in turn may cause various interference effects, such as “emotion-regulation difficulties will selectively interfere with specific memories more than general memories” (Arbuthnott et al., 2006, p. 68). (2) A person prone to dissociate may have difficulty “connecting the dots” of discrete temporal locations in order to form a coherent memory of a past event. (3) Persons with major depressive disorder might be more likely to ruminate over past failures than those with better coping or problem-solving skills, which also “creates interference in a memory retrieval task” (Arbuthnott et al., 2006, p. 68). (4) Persons with who have experienced PTSD or childhood trauma may find these comprise a “stronger part of their self-concept and had more negative influence on their current life” (Arbuthnott et al., 2006, p. 59). In this respect, Arbuthnott et al.’s account is more nuanced than that of Wilson and Ross (2003), who simply attempt to describe the phenomenological nature of human experience in general – an initiative which, because it is not context-sensitive, seems less likely to succeed.
Reverse causation. Wilson and Ross’ (2003) second proposition – that autobiographical memories influence current self-appraisals – lacks coherency. The methodology of their research protocol, with its dueling time-lines, is suspicious. It does not really prove their second proposition is true. More significantly, it is impossible for one to reconstruct a stand-alone, independently verifiable, third-party perspective on a previous event in one’s life. Under Wilson and Ross’ first proposition, concept of self thoroughly permeates autobiographical memory. This being so, how is it possible for the self (assuming the validity of this construct in the sense Wilson and Ross intend) to reorganize one’s past in a way that makes it intelligible as an unbiased source of recollection, and as a causal influencer? It is difficult to see how one can do so, free from the influence of one’s current concept of self. Wilson and Ross’ reasoning is circular because there is no independent narrative structure.
Salience. Wilson and Ross (2003) also fail to consider the relative salience of events. It stands to reason that a particularly impactful event would accrue greater topographic prominence in the conceptual landscape of one’s thoughts. An incident of early childhood trauma, for example, might be better-remembered, despite being in the distant past; as might an early childhood achievement, such as winning an academic or athletic competition. Because of their relative prominence, their persistency, and their effect on one’s personality at the time, these impactful events may be less subject to temporal elasticity in the manner Wilson and Ross propose.
Arbuthnott et al. (2006) address this issue by distinguishing between three types of autobiographical information – lifetime periods, general events, and event-specific knowledge. Lifetime period knowledge “organizes personal memories according to major segments or transitions in one’s life (Arbuthnott, et al., p. 57). General event knowledge is “knowledge of personal events and categories of events that are of shorter duration and narrower focus” (Arbuthnott, et al., p. 57). Event-specific knowledge is an experiential “re-evoking of sensory and affective experiences that occurred during the original event” (Arbuthnott, et al., p. 58). These snips overlap, recombine, and need not present themselves in a specific temporal sequence. Sometimes they even become conflated, or cross the border into the imagined. The source memory may not correspond in many significant particulars with the actual event. Wilson and Ross should have evaluated these possibilities, all of which suggest widely varying differential outcomes.
Multiple recollections. Wilson and Ross (2003) rely on an overly-simplified schematic. For Wilson and Ross, one’s present self, comprising an ensemble of ascriptive predicates, simply reflects on a single previous event. Recollection of previous events, however, is far more complicated. One seldom focuses on a particular, discrete incident. Rather, one constantly flits back and forth between the immediate and the distant past, often without rhyme or reason (and certainly this often happens during the process of therapy, for example, “free association” in a Freudian sense). Not only do a multiplicity of events present themselves for contemplation, but they also do not organize into a neat temporal sequence. Rather, their location on a chronological time-line is bi-directional and even jumbled. There are intervening, over-lapping, multi-cause-provoking and multi-effect-producing events. Sometimes they might be teleologically oriented, sometimes causally, or sometimes completely random. It seems unlikely this subjective, stream-of-consciousness assignment of events to temporal phases or stages ever could be duplicated over successive trials.
Arbuthnott et al. (2006) correctly identify this lability. They state:
[M]ental objects are not simply located in whole from a vast store of past experiences, but rather are constructed to satisfy the needs of current goals and contexts … In therapy, even theoretical modalities that concentrate on the present, such as cognitive behavioral therapy, rely on client’s recollections about therapeutically targeted behaviors or complaints. (p. 55)
Arbuthnott et al. (2006) continue: “Two recollections of the same event are seldom identical, as different occasions usually involve different goals and different cues with which to probe long-term memory” (Arbuthnott et al. p. 62). They hypothesize there are two different types of retrieval processes. Direct retrieval is activated by a cue or prime, which is a “current sensory experiences that accesses a similar phenomenal sensation in the event-specific knowledge base” (Arbuthnott et al., p. 63). Intentional retrieval is a reconstructive process. “A rememberer will first access general or lifetime period knowledge, and use the retrieved information to generate more specific cues to the target event” (Arbuthnott et al., p. 64). This account seems to better accord with actual experience, than what Wilson and Ross propose.
Developmental processes. A related variable, not considered by Wilson and Ross (2003), is the issue of how memories change with the developmental process. Wilson and Ross assume one’s ability to retrieve autobiographical memory remains constant with age. Schacter (1996) convincingly demonstrates this is not so. Schacter’s particular concern is the coherency of memories of persons who are older adults. He evocatively describes an informal experiment conducted by Pat Potter, an artist from Alabama, who selectively altered old photographs to correspond with one’s present recollections of the incident they portray (Schacter, 1996, p. 281). The substantive propositional content of those autobiographical memories is considerably altered as faces, places and situations change.
This being so, it would be incorrect to assume the autobiographical memory of persons who are older is based predominantly on “forgetting.” Schacter reviews studies, which show persons who are older, have considerable neural plasticity. There is not as much cortical and hippocampal neuron loss as once was surmised. Schacter cites previous studies using brain imaging techniques such as PET scans, which showed that retrieval of autobiographical memory occurs in the frontal lobes. Both persons who are older and those who are younger have similar levels of hippocampal activity. Schacter concludes the abilities of persons who are older to retrieve autobiographical memories vary “widely across different situations, ranging from perfectly normal to significantly impaired” (Schacter, 1996, p. 283). Absent degenerative conditions, semantic knowledge is retained; “our abilities to call on our enormous networks of facts and associations are generally well preserved” (Schacter, 1996, p 291). In some instances, one’s ability to reminisce and recall very old experiences (those occurring in adolescence or early adulthood) actually is enhanced, a phenomenon Schacter calls a “reminiscence bump” (Schacter, 1996, p. 298). While procedural memory (novel cognitive skills, motor skills) is somewhat impaired, implicit memory (using cues and primes to facilitate memory retrieval) remains intact. Schacter concludes the ability of persons who are older to retrieve and narrate autobiographical memories plays an important role in promoting community and social cohesion. “The need to preserve memories across intergenerational time … is a fundamental human imperative” (Schacter, 1996, p. 305). The continuing development of new technology, such as the Internet, facilitates one’s ability to undertake this task.
Mood and affect. Wilson and Ross (2003) base their entire theory on a conscious process of self-reflection. There are, however, many other factors that influence not only the semantic content but also the temporal organization of recalled events, and the ease with which they are brought to fore. Generally these might be characterized as “unconscious” mediators or moderators such as those, which are the subject of psychoanalysis. One’s characterization of past events may be subject to a number of other variables, all the way from complex personality traits to something as simple as one’s mood and affect, Arbuthnott et al. (2006).
The role of mood and its interplay with cognition particularly is important to the encoding and retrieval of declarative knowledge, such as autobiographical memory. Wilson and Ross (2003) do not consider these neuropsychological variables. A study, current when Wilson and Ross wrote, is Bechara et al. (1995). Bechara et al. examined the roles played by the amygdala and the hippocampus in coding and retrieving memory. The amygdala regulates emotional response and conditioning. The hippocampus contributes to the establishment and access of declarative knowledge. Bechara et al. studied three persons with different brain lesions. In person (1), the amygdala was damaged but the hippocampus was intact; in person (2), the hippocampus was damaged but the amygdala was intact; and in person (3), both the hippocampus and the amygdala were damaged. Each person underwent several trials to determine whether he or she could acquire conditioned autonomic responses, pairing various audio and visual stimuli. Each person subsequently was asked several questions about the stimuli (e.g. “what was it”). Person (1) was unable to acquire a conditioned response, but was able to answer facts about the stimuli. Person (2) acquired the conditioned response, but was unable to supply factual information about the pairing, i.e. the opposite of Person (1). Person (3) was unable to do either. On this basis, Bechara et al. concluded the amygdala was crucial to couple “exteroceptive” stimuli (originating outside of the body) with “interoceptive” stimuli (originating inside the body, such as somatic and affective states). While Bechara et al.’s methodological approach is subject to critique on several grounds, it sets forth a basic and still-useful framework for the interaction between emotion and cognitive variables in the coding and recall of autobiographical memories.
Bechara et al. (1995) present several challenges for Wilson and Ross (2003). Neuropsychological events, such as those described by Bechara et al., clearly cause conscious processes such as autobiographical memory; there is nowhere else for them to come from. Yet, the mechanism of just how this occurs remains unclear. Neuropsychological processes are “multiply realizable” because they have the potential to cause different cognitive states (Putnam, 1967; Fodor, 1974). This leads to an “explanatory gap” between them (Levine, 1983). Wilson and Ross do not even attempt to reconcile these opposing points of view. If the doctrines of multiple realizability and the explanatory gap are true, it would not be possible to distinguish positive memories from negative ones, or for that matter the substantive propositional content of any memories, to begin with.
The problem becomes even more complex when one transitions from the cognitive to the affective. Broadly read, Bechara et al. imply the hippocampus affects autobiographical memory, and the amygdala affects its associated emotional affect. If Bechara et al. are correct, and if there is a disconnect in the relationship between cognition and the hippocampus (as postulated by the doctrines of multiple realizability and the explanatory gap), then any disconnect between affect and the amygdala will be even more difficult to discern. Current work in the neuroscience of memory identifies other important aspects of this process such as outcome uncertainty, which dramatically affects the adaptation of internal timing mechanisms to the temporal characteristics of the environment (Jazayeri & Shadlen, 2010); and factors such as retrieval rates, relative recency effects, relative lag-recency effects, and issues of contiguity and scale similarity (Moreton & Ward, 2010).
I also will briefly examine the relationship between Wilson and Ross (2003) work and attachment theory. Attachment theory considers the relationship between a child’s emotional development and the caliber and quality of its relationships with initial caregivers (predominantly, the child’s mother). The child requires consistency and sensitivity in these relationships in order to develop secure cognitive and affective models for adult life (Ainsworth & Bowlby, 1991). If there is lack of early maternal interaction, or the child’s attachment bond with his or her mother is disrupted, then subsequent personality development can be adversely affected.
Bowlby argued that attachment had evolutionary utility; “the establishment of a deep and tenacious bond to the mother is an instinctual system that enhances the infant’s chances for survival” (Mitchell & Black, 1995, p. 136). In this respect, Bowlby was fundamentally different from Sigmund Freud, who held children went through stages of psycho-sexual development (the oral, anal, phallic, latent, and genital), driven by innate sexual drives, the repression of which resulted in the development of neuroses. Anna Freud, who was Sigmund Freud’s daughter, succeeded him. She thought that poor parenting and environmental deprivation were the chief sources of childhood trauma. The mother primarily is a need-gratifying object (who in turn has needs that need to be gratified), and the child can be separated from the mother, so long as some other competent caregiver replaces her (Makari, 2008).
Ainsworth supported Bowlby’s theories experimentally with field research into the child-rearing practices of Ugandan women. Bowlby’s and Ainsworth’s research led them to hypothesize various infant styles of attachment – the anxious-avoidant, secure, ambivalent-resistant and disorganized/disoriented. All are based on the mother’s interpretation of and response to infant behavioral signals. Ainsworth developed what she called a “strange situation” test, which puts the child in various stressful situations where a stranger replaces the primary care giver. The child’s response to these tests is a predictor of subsequent adult personality development. Based on Bowlby’s and Ainsworth’s work, Mary Main, Erik Hesse and Ruth Goldwyn developed an “adult attachment interview,” used to assess the relative impact of these and other factors.
Wilson and Ross (2003) ignore attachment theory. If attachment theory is true, then the stability and consistency of a child’s early relationship with his or her primary caregiver affects how the child encodes early memories, together with their substantive propositional content. It also should affect the ease with which they subsequently are retrieved, and the emotional or affective valence with which they are associated.
Although not discussed by attachment theorists, this makes sense from a neurological perspective, and should be validated by the types of experiments performed by Bechara et al. (1995). Brain activity is lateralized in the left hemisphere and the right hemisphere. Each hemisphere executes different, specialized functions. Normal right-handed persons use the left hemisphere to undertake cognitive activities such as language (including narration, interpretation) and analysis (rational thought, abstract reasoning, problem-solving). They use the right hemisphere for feelings, emotions, intuition, and creativity. (These are of course generalizations.)
According to neurophysiologists such as Eric Kandel (2007), the infant’s right hemisphere develops first. It dominates information processing from birth for up to 36 months thereafter. It develops “procedural” or “implicit memory” for non-verbal exchanges (as opposed to “explicit” or “declarative” memory, which has semantic or propositional content). It therefore should be the one most responsive to the kinds of non-verbal cues transmitted between mother and baby, identified by Bowlby and Ainsworth.
Two hypotheses should follow. First, the mother also should deploy right brain functioning in order to communicate with the baby and maximize the likelihood of successful interactions. Second, infants who have undergone trauma during the right-brain developmental period are unlikely to remember much of what occurred. The reason why is they lack sufficient left-brain capacity. Only with deep analysis and simulation or even reenactment of the actual aversive situation will they be able to articulate their implicit memories and in effect convert them into explicit ones (1). If Bowlby and Ainsworth are right, then the types of attachment relationships they identify should develop in the evolving milieu of the child’s right-brain environment. Initially they will be non-representational, non-semantic and non-propositional. Bowlby and Ainsworth do not contend that during the baby’s critical developmental period, attachment style is an actual cognitive, self-conscious or reflective mental state. They are not even unconscious or non-conscious; it would be more accurate to describe them as pre-conscious. They are precipitates of the emotional interaction between mother and child. As part of the child’s operational reality, they affect not only the child’s early development, as demonstrated in behavior patters and behavioral reactions – for example, where the child goes, and what the child does, when the child is hurt and needs soothing. They also in turn reflect the child’s later adult personality. It is at that point they also become amenable to articulation and reflection, as disclosed (for example) by the adult attachment interview, among other psychometrics. The child’s types of instinctive or intuitive feelings, and the attachment relationships they engender, correspond exactly to the development of the right brain. This reconciles Bowlby and Ainsworth with the actual neurology of the brain processes involved. Wilson and Ross (2003) do not consider any of this, even though it crucially affects the theory of autobiographical memory they develop.
Personal Application of Wilson and Ross (2003)
In conclusion, I would like to state that I have a number of interesting and suggestive early autobiographical memories. I have lead a busy life, full of static arising out of sequences of disconcerting events (though no more disconcerting or eventful than anybody else’s, considered qua event). I do not claim these are in any way unique; they just happen to be the personal ones that I have experienced. I do not doubt they have been subject to the attenuation process Wilson and Ross (2003) describe. Most noteworthy, the entire propositional content of my early memories has been compressed into a series of brief snapshots, much like colored slides or thin strips of celluloid. The emotional affect associated with the original experience has been attached to that file, riding along like a kind of appendage. Unlike Wilson and Ross, however, neither the memories, nor their correlative emotional lenses, have been redefined over time. In fact, their relative stability has enabled them to serve as reference points, or check points, throughout the confusing amalgam of activity that characterizes one’s temporal existence. I can take them out, flip through them like a photo album, and then put them back in the box from whence they came. I attribute this durability primarily to the original memory’s impact, and salience, in my conceptual scheme and view of the world.
The Scottish philosopher David Hume believed there was no such thing as a self, only a disconnected sequences of perceptions and impressions with no constancy or invariability. We confusedly amalgamate these together and derive a sense of continuous personal identity. We further hypothesize there is some sort of entity, which undertakes this activity (Hume, 1888/1968). Hume’s concept of the self has been critiqued on thousands of occasions since and it is beyond the scope of this essay even to summarize the various arguments and counter-arguments that have been adduced in support of, or against, his theory. Although they do not discuss it, Wilson and Ross (2003) would have to enroll themselves in the list of Hume defenders. I will observe, however, that as applied to me, Hume is not quite right. The reason why is because I have experienced a finite number of what might be characterized as “anchor events,” my memory of which has remained more-or-less stable and continuous, and provided me with consistent points of reference over time, despite the plasticity and variegation of other recollections.
(1) This is just a hypothesis on my part; I cannot find any references in the literature to this connection.
Ainsworth, M. & Bowlby, J. (1991). An ethological approach to personality development. American Psychologist, 46, 331 – 341.
Arbuthnott, K., Arbuthnott, D. & Thompson, V. (2006). The mind in therapy – Cognitive science for practice. Mahwah, NJ: Lawrence Erlbaum Associates.
Bechara, A., Tranel, D., Damasion, H., Adolphs, R., Rockland, C. & Damasio, A. Double dissociation of conditioning and declarative knowledge relative to the amygdala and hippocampus in humans. Science, 269(5227), 1115 – 1118. doi: 10.10.1126/science.7652558
Fodor, J. (1974). Special sciences: Or the disunity of science as a working hypothesis. Synthese, 28, 97 – 115.
Gaddis, J. (2005). The cold war. New York, NY: Penguin.
Hume, D. (1888/1968). A treatise of human nature. Oxford, U.K.: Oxford University Press.
Jazayeri, M. & Shadlen, M. Temporal context calibrates interval timing. Nature Neuroscience, 13(8), 1020 – 1038. doi: 10.1038/nn.2590
Kandel, E. (2007). In search of memory: The emergence of a new science of mind. New York, NY: Norton & Co.
Klein, S. & Gangi, C. (2010). The multiplicity of the self: Neuropsychologic evidence and its implications for the self as a construct in psychological research. Annals of the New York Academy of Sciences, 1191(1), 1 – 15. doi: 10.1111/j.1749-6632.2010.05441.x
Levine, J. (1983). Materialism and qualia: The explanatory gap. Pacific Philosophical Quarterly, 64, 354 – 361.
Makari, G. (2008). Revolution in mind – The creation of psychoanalysis. New York, NY: HarperCollins.
May, K. (2002). Golden state, golden youth – The California image in popular culture, 1955 – 1966. Chapel Hill, NC: University of North Carolina Press.
Maylor, E., Chater, N. & Brown, G. (2001). Scale invariance in the retrieval of retrospective and prospective memories. Psychonomic Bulletin & Review, 8(1), 162 – 167.
Mitchell, S. & Black, M. (1995). Freud and beyond – A history of modern psychoanalytic thought. New York, NY: Basic Books.
Moreton, B. & Ward, G. (2010). Time scale similarity and long-term memory for autobiographical events. Psychonomic Bulletin & Review, 17(4), 510 – 515. doi:10.3758/PBR.17.4.510
Putnam, H. (1967). Psychological predicates. In W. Capitan & D. Merrill (Eds.). Art, mind and religion (pp. 37 – 48). Pittsburgh, PA: University of Pittsburgh Press.
Reed, T. (2004). At the abyss – An insider’s history of the Cold War. New York, NY: Ballantine.
Rubin, D. (2006). Autobiographical memory. In Nadel, L. (Ed.). Encyclopedia of cognitive science. Hoboken, NJ: Wiley. doi: 10.1002/0470018860.s00646
Schacter, D. (1996). Searching for memory. New York, NY: Basic Books.
Wilson, A. & Ross, M. (2003). The identity function of autobiographical memory: Time is on our side. Memory, 11(2), 137 – 149. doi:10.1080/09658210244000324