Elsevier

Consciousness and Cognition

Volume 55, October 2017, Pages 91-105
Consciousness and Cognition

By-passing strategic retrieval: Experimentally induced spontaneous episodic memories in 35- and 46-month-old children

https://doi.org/10.1016/j.concog.2017.08.001Get rights and content

Highlights

  • We show how spontaneous memories can be induced experimentally in young children.

  • Age showed no systematic effect on spontaneous episodic recollection.

  • In contrast, age affected the control questions requiring strategic retrieval.

  • The results challenge the tenet that episodic memory develops late in childhood.

Abstract

Most parents have experienced their preschool child having spontaneous episodic memories, that is, verbally reported memories of past events that come to the child almost out of the blue. Until recently such memories had only been observed outside the lab. By means of a new paradigm we report experimentally induced spontaneous memories of a unique event experienced one week earlier in 35- and 46-month-old children (N = 110). At the first visit, half of the children experienced a Teddy event and the other half experienced a Game event. At the second visit the children’s spontaneous utterances were recorded while waiting. The results revealed that the children talked spontaneously about the unique event experienced previously. Age showed no systematic effect on spontaneous episodic recollection, but there was a clear effect of age on subsequent control questions requiring strategic retrieval. The results support the idea of involuntary episodic remembering being a developmentally early achievement.

Introduction

The predominant method used when examining verbally reported memories in children is to simply ask the children about their previous experiences (Hayne, Scarf, & Imuta, 2015). When asked directly about previously experienced events, young children often have a hard time retrieving and talking about specific episodes from their past (e.g., Dahl et al., 2015, Simcock and Hayne, 2002). In order to respond to such questions, children will have to engage in a deliberate and strategic retrieval process. Deliberate and strategic retrieval requires executive control and involves the use of frontal lobes, which matures relatively late in the ontogenesis (e.g., Johnson, 2005). Thus, for young children, the process involved and required when responding to direct questions about past events, is by itself a cognitively demanding achievement which, all things equal, may have detrimental effects on the children’s memory performance.

However, strategic recall is not the only way in which memories of past events come to mind. Evidence from healthy adults has shown that memories frequently come to mind involuntarily, or spontaneously, that is, without any deliberate attempts to recall the episode, but often in response to concrete environmental cues (Berntsen, 1996, Berntsen, 2009). Involuntary episodic memories differ from voluntary episodic memories by not involving neural activity in prefrontal areas associated with retrieval effort (Hall et al., 2014) and by involving shorter retrieval time (e.g., Berntsen et al., 2013, Schlagman and Kvavilashvili, 2008), both consistent with the idea of relatively effortless retrieval. At the same time, such involuntary memories appear to be highly cue dependent (Berntsen, 1996, Berntsen, 2009). In addition involuntary memories are more frequently about specific episodes and have been found to have more emotional impact, and some studies find them to be more vivid than voluntary memories (for a review, see Berntsen, 2010). The notion of spontaneous memories as used here is adopted from the research on involuntary memories in adults.

We define spontaneous memories as (i) verbally produced, (ii) socially unprompted, and (iii) environmentally cued (Krøjgaard, Kingo, Dahl, & Berntsen, 2014). ‘Socially unprompted’ means that the memories do not occur as a result of prompts of any kind, as for instance explicit or implicit questions or demand characteristics directed at the child. Because spontaneous retrieval is based primarily on associative mechanisms and hence less dependent on executive functions and mature frontal lobes, (a) spontaneous retrieval is assumed to be less cognitively demanding than strategic retrieval, and, following the same logic, (b) spontaneous memories may be more prevalent in children than in adults, and (c) is likely to be developmentally earlier than strategic remembering (Berntsen, 2009, Berntsen, 2012). In the following we briefly present the existing evidence on spontaneous memories in children (for a more detailed review, see Krøjgaard et al., 2014).

We recently reported experimental evidence of spontaneous recollections of a repeated, non-specific event in 46 months old children (Krøjgaard et al., 2014). However, no systematic research has been conducted on children’s spontaneous recollections of unique episodes in their past. The literature is also lacking systematic examination of the effects of age on spontaneous versus strategic episodic memories.

Although previous work is scarce, at times spontaneous memories as defined here have appeared in studies examining children’s memory in general. Spontaneous memories have for instance been observed in a number of studies using unstructured or semi-structured methodologies, usually diary studies (Ashmead and Perlmutter, 1980, Hudson, 1990, Nelson, 1989, Nelson and Ross, 1980, Reese, 1999, Todd and Perlmutter, 1980). As an illustrative example of a spontaneous memory, Todd and Perlmutter (1980, p. 82) reported the following:

[…] the parents of one of the three-year-olds reported that while watching a commercial that displayed a bottle of honey, the child said she had liked the chocolate stuff her mother used to give her but not the yellow. The mother felt this could only refer to the chocolate syrup and honey she used to put in the child’s bottle when she was a baby.

A number of experimentally based studies have collected children’s ‘spontaneous’ verbal references to their previous visits to a lab, while their non-verbal memories were tested by means of the elicited imitation paradigm (Bauer, Kroupina, Schwade, Dropik, & Wewerka, 1998; Bauer, van Abbema, Wiebe, Cary, Phill, & Burch, 2004; Bauer et al., 2002, Bauer and Wewerka, 1995). To illustrate, in some of the studies by Bauer and colleagues the children ‘spontaneously’ reported task-relevant material from the previous lab visits during the first of two delayed tests (Bauer et al., 1998, Bauer and Wewerka, 1995, Bauer et al., 2002, Bauer et al., 2004). However, because the setting involved encouragements to remember the previous task, as the children were explicitly asked to recall the motor actions of the events (i.e., “You can make a windmill with this stuff. Show me how to make a windmill”), these verbal productions were prompted and would therefore not fulfill the criteria used here for spontaneous memories.

Krøjgaard et al. (2014) presented evidence of experimentally induced spontaneous memories of a repeated event. Forty-six-month-old children were brought back to the lab where they had previously experienced memorable events, and while waiting in front of the props for the experimenter to return from a “phone call”, their possible spontaneous utterances were recorded. The rationale behind being able to induce spontaneous memories from specifically these previous encounters in the lab (and not any other experiences) was based on two assumptions: First, by bringing the children back to the exact same lab, with the same props, the same experimenter etc., there was a substantial overlap between the recall situation and the to-be-remembered episode which is known to facilitate memory (Hayne, 2004, Newcombe et al., 2007). Second, the cues involved were not only overlapping; they were highly distinct. This strategy was consistent with the notion of cue overload stating that “The probability of recalling an item declines with the number of items subsumed by its functional retrieval cue” (Watkins & Watkins, 1975, p. 442). Thus, the likelihood of a cue providing access to a given target memory depends on the extent to which this cue is uniquely associated with the target. Its strength declines to the extent it is associated with other memories as well. This principle has shown its relevance in relation to involuntary episodic memories in adults (e.g., Berntsen et al., 2013) and in non-human primates (Martin-Ordas, Berntsen, & Call, 2013). Based on this principle, the most important factor is not the number or range of overlapping features between a present context and a remembered event, but rather whether the available cues referred specifically to the to-be-remembered episode, and not to other episodes. The recordings from the waiting period revealed that the 46-month-olds spontaneously produced significantly more mnemonic material relative to an age-matched control group (Krøjgaard et al., 2014).

Whereas the study by Krøjgaard et al. (2014) demonstrated that spontaneous memories could be induced experimentally, the study had a number of limitations: First, only 46-month-olds were examined; hence we do not know whether spontaneous memories could be induced in even younger children. Second, the props were fully visible during the test, whereby online reasoning could be a confounder. Third, because the children in the experimental group had visited the lab three times before, the to-be-remembered event was not unique and therefore did not fulfill a commonly used criterion for episodic memory. In addition, the multiple previous visits may have made the children in the experimental group more comfortable during the test relative to controls that had never been in the lab before. Fourth, the previous visits concerned memory tasks. Thus, the possibility remains that although the children were never asked or prompted while waiting in front of the props, some of the children in the experimental group may have guessed that this study was about memory too.

In order to pursue the investigation of experimentally induced spontaneous recall further, we here tested the children’s memories for a single distinct event, contrasted age groups and improved the experimental procedure used by Krøjgaard et al. (2014).

In the present study we attempted to induce spontaneous memories in 35- and 46-month-old children, and relate the results from spontaneous recall to the results from strategic and deliberate recall obtained in subsequent control questions. Thus, when examining different age groups, we may see a change in the pattern of results across age and mode of retrieval. Forty-six-month-old children were chosen, because this was the age of the children participating in the study we set out to improve (Krøjgaard et al., 2014). The 35-month-old children were chosen in order to see, if spontaneous recall could be induced in substantially younger children, that is, in children below the age of 3.

All children visited the lab twice with a one-week retention interval. At the first visit the children were shown one out of two possible interesting and highly distinct events: either a Teddy Event involving mechanical teddies that could sing, clap and wiggle their ears; or a Game Event involving two funny games in which the children won medals. The props for the events resided in two differently looking, opaque and locked boxes. One week later the children returned to the lab, and while waiting for the experimenter to return to the test room, the children’s possible spontaneous utterances were recorded. In addition, after the recollection of spontaneous utterances, the children were asked control questions requiring strategic recall.

We expected the 46-month-old children to produce more spontaneous utterances concerning the event they had been exposed to at the first visit, relative to the event they had not seen before. Since no previous studies have examined experimentally induced spontaneous recall in 35-month-old children, our expectations could only be speculative. Nonetheless, since spontaneous recall is less demanding than strategic recall, we expected the 35-month-old to also demonstrate spontaneous recollections of the staged event, potentially at the same level as the older children. In contrast, because the control questions would require strategic and deliberate recall, which involves prefrontal processes to monitor the search, we expected the 35-month-old children to have more problems with the control questions relative to their older peers.

Section snippets

Participants

A total of 110 children participated in the study. Sixty 35-month-olds (22 female, Mage = 35.62 months, SD = 0.51; range: 35.00–37.73 months) and fifty 46-month-olds (30 female, Mage = 46.75 months, SD = 1.15; range: 45.13–49.63 months). In each age group the children were randomly assigned to one of two conditions, either the Teddy Condition or the Game Condition. Three additional children were tested but not included in the analyses (experimental error: 2; technical error: 1). The children were

Discussion

The results from the present study replicate and expand on previous findings on experimentally induced spontaneous memories (Krøjgaard et al., 2014). The results were clear, systematic, and in accordance with the proposed hypotheses: When returning to the lab (T2), the children in both age groups talked spontaneously and reliably about the specific event they had been exposed to at the encoding session (T1) one week earlier. Spontaneous retrieval was generally not influenced by age, indicating

Acknowledgement

This research was supported by the Danish National Research Foundation (Grant DNRF89) and Velux Fonden (Grant 10386). We would like to thank Tirill Fjellhaugen, Trine Ravn Andersen, Sidsel Tornhøj Sørensen, and Anne Nørgaard for help with recruiting the participants and scoring the data. We would also like to thank the children who participated in this study and their parents for letting them participate.

References (42)

  • N.S. Newcombe et al.

    Development of episodic and autobiographical memory: A cognitive neuroscience perspective

  • D.H. Ashmead et al.

    Infant memory in everyday life

    New Directions for Child and Adolescent Development

    (1980)
  • P. Bauer et al.

    If memory serves, will language? Later verbal accessibility of early memories

    Development and Psychopathology

    (1998)
  • P.J. Bauer et al.

    The onset of childhood amnesia in childhood: A prospective investigation of the course and determinants of forgetting of early-life events

    Memory

    (2013)
  • P.J. Bauer et al.

    Props, not pictures, are worth a thousand words: Verbal accessibility of early memories under different conditions of contextual support

    Applied Cognitive Psychology

    (2004)
  • P.J. Bauer et al.

    Journal of Cognition and Development

    (2002)
  • S.R. Beck et al.

    Children's thinking about counterfactuals and future hypotheticals as possibilities

    Child Development

    (2006)
  • D. Berntsen

    Involuntary autobiographical memories

    Applied Cognitive Psychology

    (1996)
  • D. Berntsen

    Involuntary autobiographical memories: An introduction to the unbidden past

    (2009)
  • D. Berntsen

    The unbidden past: Involuntary autobiographical memories as a basic mode of remembering

    Current Directions in Psychological Science

    (2010)
  • D. Berntsen

    Spontaneous recollections: Involuntary autobiographical memories are a basic mode of remembering

  • Cited by (23)

    • Spontaneous verbal recall: A new look at the mechanisms involved in episodic memory retrieval in young children

      2022, Developmental Review
      Citation Excerpt :

      Although the to-be-remembered event was repeated and hence not a unique event as is typically required when investigating episodic memory, the results from Leichtman (2006) made promise that spontaneous memories could in principle be examined in a controlled lab setting, and that spontaneous recollections appeared early in life. In a recent series of experiments Krøjgaard and colleagues have examined spontaneous recollections of unique, one-off experiences, in an experimental setup (Hjuler et al., 2021; Jensen et al., 2022; Krøjgaard, et al., 2014; Krøjgaard et al., 2017; Sonne et al., 2019, 2020, 2021, in press). The reasoning behind these experiments was based on insights from Berntsen et al.’s experimental studies with both adults (Berntsen et al., 2013; Staugaard & Berntsen, 2014) and apes (Martin-Ordas et al., 2013)—all demonstrating how distinct cues, uniquely related to a specific event, were powerful elicitors of spontaneous recollections.

    • Real-time assessment of looking time at central environmental cues for spontaneous recall in 35-month-olds

      2021, Cognitive Development
      Citation Excerpt :

      Interrater agreement was high: 98.7 %, range 88.7–100 %. As for the previous studies using this design (Krøjgaard et al., 2017; Sonne et al., 2019, in press), the spontaneous utterances were coded by two measures: a Word List and a Coding Scheme. A predefined Word List (see Table 1) was used to code the video clips from both the first and the second visit.

    View all citing articles on Scopus
    View full text