Abstract
Autobiographical memory (AM) is memory for personally relevant events in one’s own past. It entails our memories of the place of the experience (e.g., where the event occurred), the when of the experience in terms of both conventional time (e.g., day of the week) and time in one’s own life story (e.g., in what life period the event occurred), and the emotions associated with the experience. AM is comprised of two different but related types of memory: semantic memory (SM) and episodic memory (EM). Semantic AM is memory for facts about one’s self; episodic AM is memory for past personally experienced events. Individuals with autism spectrum disorder (ASD) typically have impaired AM.
The authors of this study investigated how typical adolescents and adolescents with ASD represented and retrieved personality traits associated EMs and attributed mental state knowledge about personality traits to themselves and others. The authors hypothesized that compared with typically developing (TD)-matched controls, children with ASD would (a) show increased difficulties retrieving semantic and episodic AMs, with fewer specific EMs and more prompts required for memory retrieval; (b) show reduced introspection and mentalizing abilities in relation to knowledge of their own and a close other’s internal personality traits (mental states), but not external personality traits (behaviors); and (c) given that autobiographical and mental state knowledge are hypothesized to be related and contribute to one’s self-concept, there would be an atypical relationship between AM retrieval and internal mental state attribution.
Participants in the study consisted of 48 individuals with IQs above 70 between the ages of 11 and 18 years—24 with ASD and 24 TD controls matched on the basis of gender, age, IQ, and receptive vocabulary. The researchers developed novel AM tasks to assess personality traits and associated EM in relation to different aspects of the self. Self-perceived introspection and mentalizing abilities were assessed both for introspection and mentalizing ability in relation to knowledge of both internal and external personality traits.
Assessments
Semantic Episodic Autobiographical Memory Task (SEAM Task)
This task was designed to assess the retrieval of semantic personality traits and episodic AMs in response to five personality trait cues (family, school, happy, sad, in general). Participants were first asked to generate personality traits by considering what type of person they were in response to the cue. They were then asked to recall an EM that reflected them being that type of person. Participants were encouraged to generate as many personality traits as possible and select the one they considered to be most representative. If participants initially produced a general memory, they were prompted to recall a specific example (e.g., “Can you think of a specific time when that happened?”), with children given one prompt during memory retrieval to maximize the amount of information generated (e.g., “Can you tell me anything else about that?”). Responses were audio recorded for transcription.
To assess semantic autobiographical knowledge, the number of personality traits produced was recorded.
To assess episodic autobiographical knowledge, each memory was scored for memory type (specific or general) in accordance with previously reported criteria. Specific memories were defined as an event particular to 1 day (e.g., sixth birthday), with general memories sub-classified as either categoric or extended. Categoric memories referred to multiple occurrences of an event (e.g., going to Scouts). Extended memories referred to a single event that occurred over an extended period of time (e.g., holiday in France). The number of items for which participants required prompts prior to specific memory retrieval was recorded, as was the total number of memories where children generated additional information following a general memory prompt. The number of memories that included references to sensory details and emotional details was recorded. Sensory details were defined as information perceived by the senses (e.g., “I sniffed the flowers”). Emotional details were defined as references to affective responses (e.g., “I felt happy, it made me laugh”).
Self-Knowledge Interview Task (SKI Task)
This task was developed to assess who has the authority on knowledge of one’s self (self-knowledge) and who has the authority on knowledge of a comparison individual (other-knowledge). Self- and otherknowledge were assessed for six internal personality traits (ill, tired, sad, relaxed, daydreaming, what type of person they are) and six external personality traits (dressed smartly, good at sums, running fast, reading well, eating lots of sweets, good at tidying up). For all questions, level of knowledge was rated using a 10-point Likert-type scale (0 = low knowledge and 10 = high knowledge).
Participants were first asked to identify a comparison individual. For self-knowledge questions, participants were asked to rate their own expertise: “How well do you know when you [feel . . . /are . . .?]” and then the comparison individual’s expertise: “How well does < comparison individual > know when you [feel . . . /are . . .?].” For other-knowledge questions, participants were asked to rate the comparison individual’s expertise: “how well does < comparison individual > know when they [feel . . . /are . . .?]” and then their own expertise: “how well do you know when < comparison individual > [feels . . . /is . . .?].” To ensure participants were rating expertise on the basis of their or the comparison individual’s knowledge, they were asked to give a reason for each rating: “How do you know when . . .?”
The SKI task generated four internal and four external knowledge scores:
SK-self (participant’s knowledge of themselves),
SK-other (comparison individual’s knowledge of the participant),
OK-self (participant’s knowledge of comparison individual), and
OK-other (comparison individual’s knowledge of themselves).
For each scale (SK-self, SK-other, OK-self, and OK-other), scores were out of 60 and reflected a summation of the child’s level of knowledge ratings (out of 10) to the six internal or external personality trait cues. For the internal questions, SK-self assessed introspection (i.e., knowledge of own internal mental states), SK-other assessed second-order mentalizing ability (i.e., mentalizing about the other person’s ability to attribute internal mental states to them), OK-self assessed perceived mentalizing ability (i.e., evaluation of participant’s ability to attribute internal mental states to the other person), and OK-other assessed mentalizing ability (i.e., attribution of internal mental state knowledge to the other person).
Results
Compared with TD controls, a selective pattern of impaired and preserved performance was identified across the domains of AM (semantic and episodic) and the self (introspection/mentalizing abilities and internal/external self-knowledge), with evidence of atypical relationships between these cognitive constructs for young people with ASD.
Children with ASD recalled significantly fewer personality traits than TD controls. In contrast, they recalled a typical number and type of EMs, though they required more initial prompts to facilitate memory retrieval and generated fewer memories containing emotional and sensory information.
Children with ASD rated themselves as knowing significantly less about the comparison individual’s mental states than TD controls.
The ASD group rated the comparison individual as having more knowledge about their behaviors than themselves (SK-self < SK-other, but not their mental states). In contrast, the TD group rated the comparison individual as having a similar level of knowledge about their behaviors as themselves (SK-self = SK-other). Both groups attributed the comparison individual as knowing more about their own behaviors than themselves (OK-self < OK-other).
Discussion
For TD adolescents, a significant relationship was found between semantic personality traits and episodic AMs, introspection, and mentalizing abilities. For adolescents with ASD, this relationship was not observed. This supports the idea that different aspects of AM and the self may dissociate during development. Individuals with ASD perceived themselves as knowing significantly less about other people’s internal mental states than did TD young people. Somewhat unexpected, however, was the finding that adolescents with ASD perceived someone close to them (typically a family member) as knowing more about their own (ASD participants’) behaviors than themselves, which was in contrast to the comparison group. These findings indicate a distinction between the representation of internal and external personality traits for young people with ASD, such that they perceive themselves as knowing less about their own behaviors than someone close to them. During typical development, a distinction has been proposed between the objective evaluation of behaviors and subjective measurement of mental states, with the former requiring comparative judgments and the latter requiring qualitative judgments. This suggests that for young people with ASD, introspection abilities may vary as a function of objective or subjective evaluation processes. Perceived knowledge of physical (behavioral) aspects of the self may be more impaired than perceived knowledge of psychological (mental state) aspects of the self due to reduced confidence and/ or competence in making comparative judgments about themselves relative to others. Exploring subjective and objective evaluation processes using experimental paradigms for young people with ASD would be of relevance to these discussions.
These study findings suggest that although adolescents with ASD are able to reflect and use their own experiences to inform their understanding of themselves and other people, the way in which they derive knowledge about themselves is atypical. Because there is a bidirectional relationship between AM and the self, the authors speculate that difficulties with introspection may also contribute to the atypical specification of AMs and poorer use of the self as an effective memory organizational principle. These difficulties may affect the social and directive functions of AM and may contribute to the problems persons with ASD have in forming social relationships and using past experiences to guide future behaviors and manage change.