Acta Psychologica 164 (2016) 144–150

Contents lists available at ScienceDirect

Acta Psychologica journal homepage: www.elsevier.com/locate/actpsy

Retrieving autobiographical memories: How different retrieval strategies associated with different cues explain reaction time differences Tugba Uzer Department of Psychology, ZiyaGokapl Cad. No: 48, TED University, Cankaya, Ankara 06420, Turkey

a r t i c l e

i n f o

Article history: Received 24 April 2015 Received in revised form 6 January 2016 Accepted 8 January 2016 Available online xxxx Keywords: Generative retrieval Emotion Autobiographical memory

a b s t r a c t Previous research has shown that memories cued by concrete concepts, such as objects, are retrieved faster than those cued by more abstract concepts, such as emotions. This effect has been explained by the fact that more memories are directly retrieved from object versus emotion cues. In the present study, we tested whether RT differences between memories cued by emotion versus object terms occur not only because object cues elicit direct retrieval of more memories (Uzer, Lee, & Brown, 2012), but also because of differences in memory generation in response to emotions versus objects. One hundred university students retrieved memories in response to basiclevel (e.g. orange), superordinate-level (e.g. plant), and emotion (e.g. surprised) cues. Retrieval speed was measured and participants reported whether memories were directly retrieved or generated on each trial. Results showed that memories were retrieved faster in response to basic-level versus superordinate-level and emotion cues because a) basic-level cues elicited more directly retrieved memories, and b) generating memories was more difficult when cues were abstract versus concrete. These results suggest that generative retrieval is a cue generation process in which additional cues that provide contextual information including the target event are produced. Memories are retrieved more slowly in response to emotion cues in part because emotion labels are less effective cues of appropriate contextual information. This particular finding is inconsistent with the idea that emotion is a primary organizational unit for autobiographical memories. In contrast, the difficulty of emotional memory generation implies that emotions represent low-level event information in the organization of autobiographical memory. © 2016 Elsevier B.V. All rights reserved.

1. Introduction The concreteness effect refers to the finding that words representing concrete concepts (e.g. book, chair) are processed and recognized faster than words representing more abstract information (e.g. freedom, intelligence) in many different cognitive tasks, such as reading, naming, priming, word association, lexical decision, and laboratory memory tasks (Bleasdale, 1987; Bransford & McCarrell, 1974; DeGroot, 1989; Holcomb, Kounios, Anderson, & West, 1999; Kieras, 1978; Kounios & Holcomb, 1994; Levy-Drori & Henik, 2006; Paivio, 1986, 1991; Richardson, 2003; Schwanenflugel, 1991; Schwanenflugel & Shoben, 1983; Strain, Patterson, & Seidenberg, 1995; West & Holcomb, 2000; Whaley, 1978). These studies demonstrate that concrete words are processed faster than abstract words because concrete information has more contextual (Schwanenflugel & Shoben, 1983; contextual availability theory) and pictorial (Paivio, 1986; dual-coding theory) support than abstract information. A concreteness effect has also been observed in memory for real-life events. For example, many autobiographical memory researchers

E-mail address: [email protected].

http://dx.doi.org/10.1016/j.actpsy.2016.01.004 0001-6918/© 2016 Elsevier B.V. All rights reserved.

(Anderson & Conway, 1993; Berntsen & Rubin, 2002; Brown & Schopflocher, 1998a,b; Conway, 1990; Conway & Bekerian, 1987; Fitzgerald, 1980; Fitzgerald & Shifley-Grove, 1999; Larsen & Plunkett, 1987; Robinson, 1976; Rubin, 1982, 2000; Rubin & Berntsen, 2003; Rubin & Schulkind, 1997a,b; Schlagman, Kliegel, Szhulz, & Kvavilashvili, 2009; Schlagman & Kvavilashvili, 2008; Wagenaar, 1986) have used cue word methods to study retrieval of autobiographical memories. In these studies, the nature of the cues presented to participants was manipulated, and the time required to retrieve cue-related personal memories was measured. Some of these studies compared memories cued by concrete concepts, such as object names (e.g., BOOK), to more abstract concepts, such as emotion labels (e.g., HAPPY; Conway & Bekerian, 1987; Fitzgerald, 1980; Larsen & Plunkett, 1987; Robinson, 1976; Uzer et al., 2012), and demonstrated that concrete terms lead to faster retrieval than abstract terms. The ease-of-retrieval account (Conway & Bekerian, 1987; Larsen & Plunkett, 1987; Robinson, 1976) is commonly used to explain RT differences between emotion-induced memories and memories cued by object names. The ease-of-retrieval account is based on two assumptions: 1) autobiographical memories are mostly generated, and 2) generation is easiest when concrete cues access related memories compared to abstract cues, such as emotional states, that must be reframed or

T. Uzer / Acta Psychologica 164 (2016) 144–150

elaborated on before related memories can be accessed. In other words, retrieving memories in response to emotional cues take longer than retrieving memories in response to object names because generating memories in response to emotion cues is more difficult and time consuming than generating memories in response to object cues (Conway & Bekerian, 1987; Larsen & Plunkett, 1987; Robinson, 1976). Recently, Uzer et al. (2012) proposed an alternative dual-strategies approach to explain RT differences in cued-retrieval studies. The dualstrategies approach agrees that autobiographical memories are sometimes generated. However, it also proposes that autobiographical memories can be directly retrieved, and that direct retrieval is much more common than previously suggested. The dual-strategies approach also argues that average RTs are a frequency-weighted blend of both fast responses, which occur when a memory is directly recalled, and slow responses, which occur when generation is required. In three experiments, Uzer et al. (2012) decomposed this cue-type effect by showing that participants used more direct retrieval when they were cued with objects than when they were cued with emotions. They concluded that a) the prevalence of directly retrieved memories in cue-word tasks along with those retrieved involuntarily in real-life (Ball & Little, 2006; Berntsen, 1996, 1998, 2007, 2009, 2010; Berntsen, Staugaard, & Sørensen, 2013; Rasmussen & Berntsen, 2011; Rasmussen, Ramsgaard, & Berntsen, 2015; Berntsen & Hall, 2004; Mace, 2005, 2006, 2007, 2010; Schlagman & Kvavilashvili, 2008; Schlagman, Kvavilashvili, & Schulz, 2007; Staugaard & Berntsen, 2014) imply that event memories are mostly pre-stored and b) event memories are more likely to be indexed by concrete information than abstract concepts, such as feelings. At the same time, this research suggests that generative retrieval is an additional cue generation process in which the person search for another cue or set of cues that would trigger one of these pre-stored event representations (Addis, Knapp, Roberts, & Schacter, 2012; Conway, 2009; Morton, Hammersley, & Bekerian, 1985; Norman & Bobrow, 1979; Uzer et al., 2012; Uzer & Brown, under review; Whitten & Leonard, 1981; Williams & Hollan, 1981). Reiser, Black, and Abelson's (1985) directed search model also proposes a similar cue generation process for retrieving autobiographical memories. This model is based on the idea that retrieval is more like a re-understanding process, and when people retrieve an event, they first find a context that includes the target event, and then specify features that discriminate the target event from other experiences in that context. Based on verbal protocol data collected from Yale undergraduate students, Reiser et al. (1985) showed that activities are primary search contexts, because they include most of the information required to retrieve appropriate information. Similarly, Williams and Hollan (1981) described three steps in retrieving specific memories: 1) finding a context, 2) searching within the context, and 3) verifying. Williams and Hollan (1981) proposed that when retrieving an event or object, people first try to find possible contexts (e.g. activity, location) associated with the provided cue. Once they come up with a context, they continue to search for additional information within that context, and finally they verify whether the recovered information is appropriate or not. According to Williams and Hollan (1981), retrieval follows this cycle until the person finds an appropriate response. There are also other retrieval models which propose that retrieval consists of three iterative stages (Burgess & Shallice, 1996; Conway & Pleydell-Pearce, 2000; Norman & Bobrow, 1979; Williams et al., 2007). In the first stage, a retrieval cue is elaborated; in the second stage, relevant information is accessed from long-term memory; and in the final stage the retrieved information is evaluated to decide whether it satisfies current task demands. If it does not fulfill the demands, a second cycle is initiated to generate new set of cues. Uzer and Brown (under review) argue that each event we experience is composed of some basic components (e.g., the person, location and object of the event) and that each event is represented by a node in which each core component is combined together. Each event node is also indexed by these event components (see Barsalou, 1988;

145

Brown, Hansen, Lee, Vanderveen, & Conrad, 2012; Conway, 2009; Linton, 1986; Morton et al., 1985; Shimamura, 2014 for similar arguments). There is always some level of activation in an event node accumulating from the event components, or from other associated event nodes. Access to an event representation occurs when activation of the event node exceeds certain threshold (e.g., J.R. Anderson, 1993). For instance, when there is a match between the cue(s) in the environment and the encoded event, activation in the node exceeds threshold and involuntary retrieval occurs (Berntsen, 2012). Within this framework, direct retrieval in a word cueing task can be considered as a process in which memory is accessed in the first cycle (see above). On the other hand, in generative retrieval, memories are recalled on subsequent cycles. Uzer and Brown (under review) used the term “generative” rather than “reconstructive” to describe these extended retrieval process. This was due to differentiate cue generation approach (Addis et al., 2012; Burgess & Shallice, 1996; Conway, 2009; Morton et al., 1985; Norman & Bobrow, 1979; Reiser et al., 1985; Uzer et al., 2012; Uzer & Brown, under review; Whitten & Leonard, 1981; Williams & Hollan, 1981) from the constructive approach (Conway & Pleydell-Pearce, 2000) in explaining memory generation. Dual-strategies account agrees with cue generation approach and argues that cues that will provide access to a pre-stored memory that meets the task requirements are produced during generative retrieval (Uzer & Brown, under review). In other words, direct retrieval and generative retrieval represent a continuum of a single retrieval process by which one of the pre-stored event representations are accessed with (i.e., generative retrieval) or without (i.e., direct retrieval) extra effort.1 Constructive approach, on the other hand, is different from the cue generation approach. Constructive view states that “autobiographical memories are not stored in long-term memory, but rather are constructed on the basis of knowledge sampled from the autobiographical knowledge base”. According to constructive retrieval model, memories are “temporary or transitory mental representations that only exist in the context of some specific processing episode” (Conway, 1996, p. 76). On this view, “a specific autobiographical memory is a pattern of activation across the indexes of the autobiographical knowledge base conjoined with the retrieval model used to shape that pattern” (Conway & Pleydell-Pearce, 2000, p. 274). The indexes used to create the event representation and the ones retrieved during the extended search process are considered identical in this model. In other words, constructive approach defines an autobiographical memory as “all the knowledge accessed in a whole set of retrieval cycles” (Conway, 1996, p. 77). During these retrieval cycles, knowledge is accessed hierarchically from the autobiographical knowledge base which contains information at three levels of specificity (i.e., life-time periods, general events and eventspecific knowledge (ESK); Conway & Pleydell-Pearce, 2000). In other words, when a cue is provided, it creates a pattern of activation which starts from a life-time period and ends in the associated ESK. Once this pattern of activation is established and meets the task demands, the target memory has been constructed. The present study tested whether RT differences between memories cued by emotion versus object cues reflect not only the fact that object cues directly retrieve more memories (Uzer et al., 2012), but also differences in memory generation between responses to emotions and objects. Based on the directed search model, we argue that directing memory search based on affective features requires more steps than directing memory search based on other features (e.g. features associated with objects). This is because emotion indirectly refers to an action, 1 Note, some other retrieval models (e.g., directed search model, retrieval model proposed by Williams and Hollan) take cue generation approach and proposes that memory retrieval is more like a cue generation process in which people generate additional cues that provide contextual information to recall the target event rather than an event construction where memories are reconstructed by retrieving information from a hierarchically organized autobiographical knowledge base. Different from the dual-strategies account, these retrieval models do not acknowledge the existence of direct retrieval but assumes that memories are always generated on several cycles.

146

T. Uzer / Acta Psychologica 164 (2016) 144–150

place, or person (Linton, 1986; Robinson & Clore, 2002a). More specifically, remembering angry experiences requires specifying something you were angry about, and then focusing on a particular occasion of being angry. As Spencer (1899) argued, representations of anger cannot be achieved without imagining, and relying on some circumstances to produce that emotion. In contrast, object terms more directly specify the context (Schwanenflugel & Shoben, 1983) where an appropriate experience can be found. Hence, directing memory search based on object cues requires one less step (see also Robinson, 1976). Understanding how emotional cues are used to search memory is important because it would change our understanding of how personal events are represented in memory, and explain how emotion is integrated in memory representations. There is some evidence that emotion is a primary component of autobiographical memory representations. For example, Conway (1998) reported that emotion words are more likely to spontaneously trigger autobiographical memories than generic concepts such as “furniture” or “presents”. Diary studies also indicate that emotional memories are recalled easier than neutral memories (Brewer, 1988; Wagenaar, 1986). Alternatively, other studies have demonstrated that emotions represent low-level event information in the organization of autobiographical memory. For instance, there are some studies reporting weak relationships between emotion and autobiographical memory retrieval (Linton, 1982, 1986; Pillemer, Goldsmith, Panter, & White, 1988; Rubin & Schulkind, 1997b). In the present experiment, basic-level (e.g. tree, orange), superordinate-level (e.g. plant, fruit), and emotion (e.g. frustrated, surprised) word cues were used to elicit autobiographical memories. We predicted that basic-level cues would elicit more directly retrieved memories than superordinate-level and emotion cues. This prediction is based on the assumption that direct retrieval requires a strong association between a cue and an event memory (Berntsen, 2009, 2012; Rasmussen & Berntsen, 2009), and that objects representing basiclevel categories are more likely associated with event memories than objects from superordinate-level categories and emotions. Proponents of the constructionist position also recognize the existence of direct retrieval. However, they assume that the direct retrieval (i.e., directly forming a distinct and stable pattern) of autobiographical memories is quite uncommon (Conway & Pleydell-Pearce, 2000; also see Addis et al., 2012; Anderson & Dewhurst, 2009; Conway, et al., 1999; Eade, Healy, Williams, Chan, Crane & Barnhofer, 2006; Haque & Conway, 2001; Hauer, Wessel, Geraerts, Merckelback, & Dalgleish, 2008; Williams et al., 2007). Thus, constructive view would predict that direct retrieval rates should be low regardless of the cue type. Regarding generative retrieval, based on the dual strategies account (Uzer et al., 2012; Uzer & Brown, under review), the directed search model (Reiser et al., 1985), and contextual availability theory (Schwanenflugel & Shoben, 1983), we argue that generating memories in response to emotion and superordinate-level cues will take the same amount of time because specifying an appropriate memory context for both emotion and superordinate-level cues requires more steps. In contrast, generating memories should be faster in response to basic-level versus superordinate-level and emotion cues because basic-level categories more directly refers to an action, place, thing, or context where a memory could be retrieved. Constructionist position would not necessarily predict any RT difference among generated memories cued by basic-level, superordinate-level and emotion cues. This is because any one of these categories does not necessarily provide an advantage over others in accessing life-time periods, general events or event specific knowledge. 2. Method 2.1. Participants One hundred undergraduates (60 females, median age = 18; 40 males, median age = 18) received course credit for participation.

Participants were tested individually in a procedure that took approximately 30 min.

2.2. Procedure Participants were presented with 3 words from basic-level categories (orange, hammer, dog, potato, tree, and doll), 3 words from superordinate-level categories (fruit, tool, animal, vegetable, plant, and toy), and 3 words from emotion categories (surprised, satisfied, and frustrated),2 and were asked to recall an autobiographical memory for each cue. Cues were presented in a random order. If the participant was shown a word from a basic-level category (e.g. orange), s/he was not presented with that word's superordinate associate (e.g. fruit); instead, s/he received a different word's superordinate associate (e.g. toy). In other words, the same subject did not receive both the basic-level word and its superordinate-level partner (e.g. orange-fruit) or vice versa (e.g. vegetable-potato). Each word was presented with equal frequency across participants. On each trial, participants were asked to recall a specific, cue-related autobiographical memory that happened at least one week ago. A specific event memory had the following qualities: 1) the event happened at a specific time and location, 2) it occurred only once, and 3) it lasted no longer than a day. Participants were also warned to avoid providing repeated or general event memories. Participants were instructed to press the SPACEBAR as soon as an appropriate memory came to mind. Pressing the SPACEBAR stopped the RT timer, and the strategy-menu display was presented. The menu included the general question “How did you retrieve this memory?” and two response options: “The memory came to mind almost immediately” and “I had to actively search to find the memory”. Participants pressed either the “Z” or “N” key (counterbalanced across participants) to indicate that the memory had been retrieved without apparent effort, implying direct retrieval, or that memory retrieval was effortful and required active search, implying generation. After responding to the strategy menu, participants typed a brief description of the event. If no appropriate memory came to mind within 90 s, the trial was terminated, and the participant was asked to initiate a new trial. The distinction between direct and generative retrieval was explained to participants in detail. Specifically, participants were told that direct retrieval refers to occasions in which memories come to mind with little or no effort, while generative retrieval occurs when memories have to be actively searched for to be retrieved.3

3. Results The frequency of direct and generative retrieval as a function of cue type is presented first. We also compared retrieval speed of direct versus generative retrieval. Next, we report RTs for direct and generative retrieval as a function of cue type to determine whether RT differences between emotion and object (basic-level) cues denotes both more directly retrieved memories in response to object cues (Uzer et al., 2 The study of comparing memories cued by basic level, superordinate level, and emotion words was first developed and conducted by Tugba Uzer, Peter J. Lee and Norman R. Brown at the University of Alberta Psychology Lab. The present study is a modified version which was undertaken by using different cue words and different sample. In selecting words, following criteria were achieved a) basic-level cues are good representatives of their super-ordinate levels, b) there is no big difference among superordinate level words in terms of abstractness, and c) there is no big difference between superordinate level words and emotion terms in terms of abstractness. Additionally, most of the superordinate and basic level terms in the present study have been frequently used in previous studies (Markman & Wisniewski, 1997; Murphy & Brownell, 1985; Murphy & Wisniewski, 1989; Rosch, Mervis, Gray, Johnson, & Boyes-Braem, 1976; Wisniewski & Murphy, 1989). 3 Uzer et al. (2012) manipulated strategy-reporting formats between subjects in Experiment 3. The format used in one condition, was identical to the one just describe; the format in the other was asking whether participants had used additional information during recall (i.e., generative retrieval), or not (i.e., direct retrieval). This manipulation had little effect on performance. Therefore, it wasn't included as a factor in the present study.

T. Uzer / Acta Psychologica 164 (2016) 144–150

2012) and a difference in the memory generation between responses to emotion and object cues. The analyses described below are based on 864 responses. Fifteen time-out trials and 21 observations with RTs equal to or greater than the 2.5 standard deviations from their strategy-type by cue-type group means were excluded from these analyses. 3.1. Strategy prevalence and cue type The strategy reports indicated that direct retrieval (49.7%) was as common as generative retrieval (50.3%), and that basic-level cues (57.8%) produced more direct retrieval than superordinate-level (42.9%) and emotion (47.9%; Fig. 1) cues. Direct retrieval rates were similar to those reported in previous studies (Harris, O'Connor, & Sutton, 2015; Uzer et al., 2012; Uzer & Brown, under review). We applied a logistic linear mixed-effects (LME) analysis to assess the effect of cue type on retrieval strategy. In this analysis, cue type (basic-level, superordinate-level, and emotion) was entered as an independent variable and retrieval type was coded as a dependent variable. Participants and items (cue words and personal cues) were random variables. Results indicated that basic-level cues led to a significantly higher rate of direct retrieval than superordinate-level (b = 0.65; SE = 0.27; z = 2.41; p b .05) and emotion (b = 0.44; SE = 0.31; z = 2.05; p b .05) cues. 3.2. Cue type, retrieval strategy, and RT First, RTs were log-transformed. Then, LME models were fit using cue type (basic-level, superordinate-level, and emotion), and retrieval strategy (direct and generative) as fixed factors, and participants and cues as random factors. All main effects and interactions were examined using a stepwise variable elimination method (see Uzer et al., 2012 for details). Mean RTs are presented in the figures with 95% confidence intervals. This analysis revealed three main results. First, basic-level cues (M = 8.86 s; 95% CI [7.57, 10.14]) produced memories significantly faster than superordinate-level (M = 12.91 s; 95% CI [11.01, 14.81]; b = 3.21; SE = 1.57; t = 2.15; p b .05) and emotion (M = 12.35 s; 95% CI [10.65, 14.05]; b = 3.58; SE = 1.60; t = 2.47; p b .05); Fig. 2). Second, direct retrieval (M = 4.47 s; 95% CI [4.03, 4.90]) was significantly faster than generative retrieval (M = 18.11 s; 95% CI [16.51, 19.72]; b = 12.06; SE = 1.39; t = 8.67; p b .05; Fig. 3). This point is also illustrated in Fig. 4, which presents the frequency and cumulative RT distributions for each retrieval strategy.

Fig. 1. Percentage of direct retrievals by cue type.

147

Third, directly retrieved responses were equally fast regardless of whether participants were responding to basic-level (M = 3.99 s; 95% CI [3.31, 4.67]), superordinate-level (M = 4.54 s; 95% CI [3.68, 5.40]; b = 0.57; SE = 2.09; t = 0.28; p N .05) or emotion (M = 4.99 s; 95% CI [4.23, 5.77]; b = 1.14; SE = 2.06; t = 0.56; p N .05) cues. However, RT for generated memories was faster for basic-level (M = 15.51 s; 95% CI [13.04, 17.98]) cues than superordinate-level (M = 19.20 s; 95% CI [16.27, 22.13]; b = 2.36; SE = 0.05; t = 2.42; p b .05) and emotion (M = 19.12 s; 95% CI [16.33, 21.90]; b = 2.09; SE = 0.05; t = 2.29; p b .05) cues. There was no RT difference between memories generated in response to superordinate-level and emotion cues (b = 0.02; SE = 0.05; t = 0.41; p N .05; see Fig. 5). These results indicate that the relationships between cue type and rate of direct retrieval and between retrieval type and RT account for interactions between cue type and RT. Specifically, aggregate RT was slower for emotion and superordinate-level cues compared to basiclevel cues because 1) emotion and superordinate-level cues yielded less direct retrieval than basic-level cues, and 2) memory generation in response to emotion and superordinate-level cues was slower than memory generation in response to basic-level cues. In summary, consistent with prior research (Harris et al., 2015; Uzer & Brown, under review; Uzer et al., 2012), the present experiment demonstrated that direct retrieval is common in response to basic-level cues, and, as predicted, basic level cues are more likely to trigger direct retrieval than superordinate-level and emotion cues. In addition, the RT pattern was comparable to Uzer and Brown (under review), and Uzer et al. (2012); regardless of cue type, RTs were faster when participants indicated that they directly retrieved the memory than when they did not. However, this experiment also showed that this RT difference occurred in part because emotion-cued memories were generated more slowly than object-cued (i.e. basic-level cues) memories. 4. Discussion In the present study, we measured the time to retrieve autobiographical memories in response to basic-level, superordinate-level, and emotion cues. We also collected self-reported retrieval strategies. Replicating previous research (Harris et al., 2015; Jeunehomme & D'Argembeau, 2015; Uzer et al., 2012; Uzer & Brown, under review), we found that direct retrieval was as common as generative retrieval, and that directly retrieved memories were recalled faster than generated memories. In addition, as predicted, we found that direct retrieval was more common in response to object (basic-level) cues than superordinate-level and emotion cues, and memories triggered by emotion cues and superordinate-level cues were generated more slowly than memories triggered by object (basic-level) cues. Therefore, overall RTs were slower when cues were more abstract (i.e. emotion and superordinate-level cues).

Fig. 2. Mean reaction times by cue type.

148

T. Uzer / Acta Psychologica 164 (2016) 144–150

Fig. 3. Mean reaction times by retrieval strategy type.

The following conclusions can be drawn from these findings. First, the prevalence of direct retrieval in these word-cue tasks and involuntary memory literature (Ball & Little, 2006; Berntsen, 1996, 1998, 2007, 2009, 2010; Berntsen & Hall, 2004; Mace, 2005, 2006, 2007, 2010; Schlagman & Kvavilashvili, 2008; Schlagman, et al., 2007) strongly imply that there are enduring event representations in autobiographical memory (Barsalou, 1988; Uzer et al., 2012; Uzer & Brown, under review). These findings also argue against the view that autobiographical memories must be (re)constructed from retrieval cues and fragments of associated event-specific knowledge (Conway & PleydellPearce, 2000). In other words, consistent with previous research (Harris et al., 2015; Jeunehomme & D'Argembeau, 2015; Uzer et al., 2012; Uzer & Brown, under review) direct retrieval data support dualstrategies account and disconfirm constructive view's notion that direct retrieval is infrequent. Second, the finding that direct retrieval is more common when cues are concepts representing basic-level categories versus more abstract concepts, such as superordinate-level categories and emotions, is consistent with the view that personal memories are typically indexed by concrete information versus abstract concepts, such as feelings (Uzer et al., 2012). Finally, consistent with dual-strategies account, the generative retrieval time data demonstrate that generative retrieval is a back-up strategy in which people continue to generate contextual cues until they find an appropriate memory, and more abstract concepts, such as emotions and superordinate-level categories, are less effective in triggering memories because they do not directly refer to an activity, place, or person. Specifying contexts associated with emotions and/or superordinate-level categories requires additional steps, and these additional steps increase retrieval time, and participants often report them as generative retrievals. Note, having more difficulty in generating memories in response to emotion words such as “happy” and cues from super-ordinate level categories such as “furniture” than cues from basiclevel categories such as “chair” is not predicted by constructive retrieval approach since “chair” is not necessarily more efficient than “happy” or “furniture” to access a particular life-time period and to form a “stable pattern of activation in the autobiographical knowledge base”.

Fig. 5. Mean reaction times by cue type and retrieval strategy type.

In sum, the present data indicate that people can often directly retrieve memories, they generally rely on a generative strategy when the cue fails to directly trigger a memory (Uzer et al., 2012; Uzer & Brown, under review), and that generative retrieval involves a cue generation cycle in which people generate cues which provide a context that includes the target event. These findings are completely predicted by dual-strategies account which proposes that a-) autobiographical memories can be both directly retrieved and generated and b-) generative retrieval is a cue generation process in which people keep producing contextual cues until they find an appropriate memory. The following concurrent verbal protocol from previous research (Uzer et al., 2012) illustrates an example of a cue generation for the emotion word “satisfied”. Protocol content: What might be satisfying? Food can be satisfying, homework can be satisfying, satisfied with work I guess? So then work, work situations, satisfied…okay (the person remembered the event). On the other hand, the results of the present project are inconsistent with constructive approach which a) assumes that generative retrieval is the most common form of retrieval and b) describes it as a reconstruction process during which knowledge represented at different levels of the autobiographical knowledge base is combined to form a memory. Other retrieval models (e.g., directed search model, William and Hollan's model) that propose cue generation as a generative retrieval process but do not address direct retrieval cannot account for all of the results either. To conclude, this research is concerned with memory generation in response to concrete concepts such as basic-level categories versus abstract concepts, such as superordinate-level categories and emotions. Our results suggest that generative retrieval is a cue generation process in which additional cues are produced to provide contextual information that includes the target event. Memories are retrieved more slowly

Fig. 4. Reaction time frequency distributions (top row) and cumulative distributions. (bottom row) type.

T. Uzer / Acta Psychologica 164 (2016) 144–150

in response to emotions in part because emotions, at least when they are cued by labels, serve as less effective cues to appropriate contextual information. This is inconsistent with the idea that emotion is a primary component in organizing autobiographical memories. In contrast, the difficulty of emotional memory generation implies that emotions represent low-level event information in autobiographical memory. If emotion is not a primary dimension, why do many studies find a profound effect of mood on autobiographical memory (Christianson & Safer, 1996)? One possibility is that linguistic representations of emotions (words) are not very effective in eliciting the emotional experience itself (see also Schulkind & Woldorf, 2005). There is also another possibility that could be discussed here. If emotion is an important component in organizing autobiographical memories, then there might be a large number of memories associated with an emotional term. Previous research (Berntsen, et al., 2013; Vannucci, Pelagatti, Hanczakowski, Mazzoni, & Rossi Paccani, 2014) on involuntary memories demonstrated that cue overload (i.e., when there is a large number of memories associated with a cue) decreased the occurrence of involuntary memories. Therefore, one reason for the difference between basic-level words and emotion words in the nature of retrieval might be due to cue overload. In other words, because there might be so many memories associated with general emotions no specific memory is likely to come directly into mind in response to emotion cues. Testing of cue overload hypothesis in this type of voluntary retrieval task would be important to understand why some cues are associated with different retrieval strategies.4 In addition, using concurrent verbal protocols to investigate the particular strategies people use when emotional words versus emotions themselves are cues would be a fruitful direction for future research and would help us understand how cues are used to generate memories.

References Addis, D. R., Knapp, K., Roberts, R. P., & Schacter, D. L. (2012). Routes to the past: Neural substrates of direct and generative autobiographical memory retrieval. NeuroImage, 59, 2908–2922. Anderson, J. R. (1993). Rules of the mind. Hillsdale, NJ: Earlbaum. Anderson, S. J., & Conway, M. A. (1993). Investigating the structure of autobiographical memory. Journal of Experimental Psychology. Learning, Memory, and Cognition, 20, 1063–1087. Anderson, R. J., & Dewhurst, S. A. (2009). Remembering the past and imagining the future: Differences in event specificity of spontaneously generated thought. Memory, 17(4), 367–373. Ball, C. T., & Little, J. C. (2006). A comparison of involuntary autobiographical memory retrievals. Applied Cognitive Psychology, 20(9), 1167–1179. Barsalou, L. W. (1988). The content and organization of autobiographical memories. In U. Neisser, & E. Winograd (Eds.), Remembering reconsidered: Ecological and traditional approaches to the study of memory (pp. 193–243). Cambridge: Cambridge University Press. Berntsen, D. (1996). Involuntary autobiographical memory. Applied Cognitive Psychology, 10, 455–460. Berntsen, D. (1998). Voluntary and involuntary access to autobiographical memory. Memory, 6, 113–141. Berntsen, D. (2007). Involuntary autobiographical memories: Speculations, findings and an attempt to integrate them. In J. H. Mace (Ed.), Involuntary memory (pp. 20–50). Malden: Blackwell. Berntsen, D. (2009). Involuntary autobiographical memories: An introduction to the unbidden past. Cambridge: Cambridge University Press.

4 As noted above, we argue that a higher rate of directly retrieved memories evoked by basic-level cues than emotion cues imply that autobiographical memories are indexed by concrete concepts rather than abstract concepts. Previous research (Harris et al., 2015; Jeunehomme & D'Argembeau, 2015; Uzer & Brown, under review; Uzer et al., 2012) also confirmed that this is a robust finding. However, some studies on involuntary memory (Berntsen, et al., 2013; Vannucci et al., 2014) also suggest that cue overload is one factor which decreases the frequency of memories that come to mind spontaneously. Considering the similarities between involuntary memories in real-life and direct retrieval in laboratory studies (Schlagman & Kvavilashvili, 2008) we wanted to acknowledge the possibility of cue overload in our data. We believe that testing whether the rate of directly retrieved memories cued by emotions is lower than the rate of directly retrieved memories cued by objects is due to cue overload or not might be a fruitful research direction to better understand whether autobiographical memories are indexed by concrete nouns or emotions.

149

Berntsen, D. (2010). The unbidden past: Involuntary autobiographical memories as a basic mode of remembering. Current Directions in Psychological Science, 19, 138–142. Berntsen, D. (2012). Spontaneous recollections: Involuntary autobiographical memories as a basic mode of remembering. In D. Berntsen, & D. C. Rubin (Eds.), Understanding autobiographical memory: Theories and approaches (pp. 209–310). Cambridge: Cambridge University Press. Berntsen, D., & Hall, N. M. (2004). The episodic nature of involuntary autobiographical memories. Memory & Cognition, 32, 789–803. Berntsen, D., & Rubin, D. C. (2002). Emotionally charged autobiographical memories across the lifespan: The recall of happy, sad, traumatic and involuntary memories. Psychology and Aging, 17, 636–652. Berntsen, D., Staugaard, S. R., & Sørensen, L. M. T. (2013). Why am I remembering this now? Predicting the occurrence of involuntary (spontaneous) episodic memories. Journal of Experimental Psychology: General, 142(2), 426–444. Bleasdale, F. A. (1987). Concreteness-dependent associative priming: Separate lexical organization for concrete and abstract words. Journal of Experimental Psychology: Learning, Memory, and Cognition, 13(4), 582–594. Bransford, J. D., & McCarrell, N. S. (1974). Cognition and the symbolic processes. In W. Weimer, & D. Palermo (Eds.), A sketch of a cognitive approach to comprehension: Some thoughts on what it means to comprehend (pp. 189–230). Hilsdale, NJ: Erlbaum. Brewer, W. F. (1988). Memory for randomly sampled autobiographical events. In U. Neisser, & E. Winograd (Eds.), Remembering reconsidered: Ecological and traditional approaches to the study of memory (pp. 21–90). New York: Cambridge University Press. Brown, N. R., & Schopflocher, D. (1998a). Event clusters: An organization of personal events in autobiographical memory. Psychological Science, 9, 470–474. Brown, N. R., & Schopflocher, D. (1998b). Event cueing, event clusters, and the temporal distribution of autobiographical memories. Applied Cognitive Psychology, 12, 305–319. Brown, N. R., Hansen, T. G. B., Lee, P. J., Vanderveen, S. A., & Conrad, F. G. (2012). Historically-defined autobiographical periods: Their origins and implications. In D. Berntsen, & D. C. Rubin (Eds.), Understanding autobiographical memories: Theories and approaches (pp. 160–180). Cambridge: Cambridge University Press. Burgess, P. W., & Shallice, T. (1996). Confabulation and the control of recollection. Memory, 4, 359–411. Christianson, S. A., & Safer, M. A. (1996). Emotional events and emotions in autobiographical memories. In D. C. Rubin (Ed.), Remembering our past: Studies in autobiographical memory (pp. 218–243). Cambridge, MA: Cambridge University Press. Conway, M. A. (1990). Associations between autobiographical memories and concepts. Journal of Experimental Psychology. Learning, Memory, and Cognition, 16, 799–812. Conway, M. A. (1996). Autobiographical memories and autobiographical knowledge. In D. C. Rubin (Ed.), Remembering our past: Studies in autobiographical memory (pp. 67–93). Cambridge, England: Cambridge University Press. Conway, M. A. (1998). Conceptual representation of emotions: The role of autobiographical memories. In K. J. Gilhooly, M. T. G. Keane, R. H. Logie, & G. Erdos (Eds.), Lines of thinking. Vol. 2. (pp. 133–143). Chichester, U.K.: Wiley. Conway, M. A. (2009). Episodic memories. Neuropsychologia, 47, 2305–2313. Conway, M. A., & Bekerian, D. B. (1987). Organization in autobiographical memory. Memory and Cognition, 15, 119–132. Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical memories in the self-memory system. Psychological Review, 107, 261–288. Conway, M. A., Turk, D. J., Miller, S. L., Logan, J., Nebes, R. D., Meltzer, C. C., & Becker, J. T. (1999). A positron emission tomography (PET) study of autobiographical memory retrieval. Memory, 7, 679–703. DeGroot, A. M. B. (1989). Representational aspects of word imageability and word frequency assessed through word association. Journal of Experimental Psychology: Learning, Memory, and Cognition, 15(5), 824–845. Eade, J., Healy, H., Williams, J. M. G., Chan, S., Crane, C., & Barnhofer, T. (2006). Retrieval of autobiographical memories: The mechanisms and consequences of truncated search. Cognition & Emotion, 20, 351–382. Fitzgerald, J. M. (1980). Sampling autobiographical memory reports in adolescents. Developmental Psychology, 16, 675–676. Fitzgerald, J. M., & Shifley-Grove, S. (1999). Memory and affect: Autobiographical memory distribution and availability in normal adults and recently detoxified alcoholics. Journal of Adult Development, 6, 11–19. Harris, C. B., O'Connor, A. R., & Sutton, J. (2015). Cue generation or memory construction in direct and generative autobiographical memory. Consciousness and Cognition, 33, 204–216. Haque, S., & Conway, M. A. (2001). Sampling the process of autobiographical memory construction. European Journal of Cognitive Psychology, 13, 529–547. Hauer, B. J. A., Wessel, I., Geraerts, E., Merckelbach, H., & Dalgleish, T. (2008). Manipulating autobiographical memory specificity in adults reporting childhood sexual abuse. Journal of Abnormal Psychology, 117, 444–453. Holcomb, P. J., Kounios, J., Anderson, J. E., & West, W. C. (1999). Dual-coding, contextavailability, and concreteness effects in sentence comprehension: An electrophysiological investigation. Journal of Experimental Psychology: Learning, Memory, and Cognition, 25(3), 721–742. Jeunehomme, O., & D'Argembeau, A. (2015). Prevalence and determinants of direct and generative modes of production of episodic future thoughts in the word cueing paradigm. The Quarterly Journal of Experimental Psychology. Advanced Online Publication. http://dx.doi.org/10.1080/17470218.2014.993663. Kieras, D. (1978). Beyond pictures and words: Alternative information-processing models for imagery effects in verbal memory. Psychological Bulletin, 85, 532–554. Kounios, J., & Holcomb, P. J. (1994). Concreteness effects in semantic processing: ERP evidence supporting dual-coding theory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 20(4), 804–823.

150

T. Uzer / Acta Psychologica 164 (2016) 144–150

Larsen, S. F., & Plunkett, K. (1987). Remembering experienced and reported events. Applied Cognitive Psychology, 1, 15–26. Levy-Drori, S., & Henik, A. (2006). Concreteness and context availability in lexical decision tasks. American Journal of Psychology, 119(1), 45–65. Linton, M. (1986). Ways of searching the contents of the memory. In D. C. Rubin (Ed.), Autobiographical memory (pp. 50–67). New York: Cambridge University Press. Linton, M. G. (1982). Transformations of memory in everyday life. In U. Neisser (Ed.), Memory observed: Remembering in natural contexts (pp. 77–91). New York: Freeman. Mace, J. H. (2005). Priming involuntary autobiographical memories. Memory, 13, 874–884. Mace, J. H. (2006). Episodic remembering creates access to involuntary conscious memory: Demonstrating involuntary recall on a voluntary recall task. Memory, 14, 917–924. Mace, J. H. (2007). Involuntary memory: Concept and theory. In J. H. Mace (Ed.), Involuntary memory (pp. 1–19). Oxford, England: Blackwell. Mace, J. H. (2010). The act of remembering the past: an overview. In J. H. Mace (Ed.), The act of remembering: Toward and understanding of how we recall the past (pp. 3–10). Oxford; UK: Wiley-Blackwell. Markman, A. B., & Wisniewski, E. J. (1997). Similar and different: The differentiation of basic-level categories. Journal of Experimental Psychology: Learning, Memory, and Cognition, 23, 54–70. Morton, J., Hammersley, R. H., & Bekerian, D. A. (1985). Headed records: A model for memory and its failures. Cognition, 20(1), 1–23. Murphy, G. L., & Brownell, H. H. (1985). Category differentiation in object recognition: Typicality constraints on the basic category advantage. Journal of Experimental Psychology: Learning, Memory, and Cognition, 11, 70–84. Murphy, G. L., & Wisniewski, E. J. (1989). Categorizing objects in isolation and in scenes: What a superordinate is good for. Journal of Experimental Psychology: Learning, Memory, and Cognition, 15, 572–586. Norman, D. A., & Bobrow, D. G. (1979). Descriptions: An intermediate stage in memory retrieval. Cognitive Psychology, 11, 107–123. Paivio, A. (1986). Mental representations: a dual coding approach. Oxford University Press: New York. Paivio, A. (1991). Dual coding theory—Retrospect and current status. Canadian Journal of Psychology, 45(3), 255–287. Pillemer, D. B., Goldsmith, L. R., Panter, A. T., & White, S. H. (1988). Very long-term memories of the first year in college. Journal of Experimental Psychology. Learning, Memory, and Cognition, 14, 709–715. Rasmussen, A. S., & Berntsen, D. (2009). The possible functions of involuntary autobiographical memories. Applied Cognitive Psychology, 23, 1137–1152. Rasmussen, A. S., & Berntsen, D. (2011). The unpredictable past: Spontaneous autobiographical memories outnumber autobiographical memories retrieved strategically. Consciousness and Cognition, 20(4), 1842–1846. Rasmussen, A. S., Ramsgaard, S. B., & Berntsen, D. (2015). Frequency and Functions of Involuntary and Voluntary Autobiographical Memories Across the Day. Psychology of Consciousness: Theory, Research and Practice, 2, 185–205. Reiser, B. J., Black, J. B., & Abelson, R. P. (1985). Knowledge structures in the organization and retrieval of autobiographical memory. Cognitive Psychology, 17, 89–137. Richardson, J. T. E. (2003). Dual coding versus relational processing in memory for concrete and abstract words. European Journal of Cognitive Psychology, 15(4), 481–509. Robinson, J. A. (1976). Sampling autobiographical memory. Cognitive Psychology, 8, 578–595. Robinson, M. D., & Clore, G. L. (2002a). Belief and feeling: Evidence for an accessibility model of emotional self-report. Psychological Bulletin, 128, 934–960. Rosch, E., Mervis, C. B., Gray, W. D., Johnson, D. M., & Boyes-Braem, P. (1976). Basic objects in natural categories. Cognitive Psychology, 8, 382–439. Rubin, D. C. (1982). On the retention function for autobiographical memory. Journal of Verbal Learning and Verbal Behavior, 21, 21–38. Rubin, D. C. (2000). The distribution of early childhood memories. Memory, 8, 265–269.

Rubin, D. C., & Berntsen, D. (2003). Life scripts help to maintain autobiographical memories of highly positive, but not highly negative, events. Memory & Cognition, 31, 1–14. Rubin, D. C., & Schulkind, M. D. (1997a). The distribution of autobiographical memories across the lifespan. Memory & Cognition, 25, 859–866. Rubin, D. C., & Schulkind, M. D. (1997b). Distribution of important and word-cued autobiographical memories in 20-, 35-, and 70-year-old adults. Psychology and Aging, 12, 524–535. Schlagman, S., & Kvavilashvili, L. (2008). Involuntary autobiographical memories in an outside the laboratory: How different are they from voluntary autobiographical memories? Memory & Cognition, 36, 920–932. Schlagman, S., Kvavilashvili, L., & Schulz, J. (2007). Effects of age on involuntary autobiographical memories. In J. Mace (Ed.), Involuntary memory (pp. 87–112). New York: Blackwell. Schlagman, S., Kliegel, M., Szhulz, J., & Kvavilashvili, L. (2009). Differential effects of age on involuntary and voluntary autobiographical memory. Psychology and Aging, 24, 397–411. Schulkind, M. D., & Woldorf, G. (2005). Emotional organization of autobiographical memory. Memory & Cognition, 33(6), 1025–1035. Schwanenflugel, P., & Shoben, E. (1983). Differential context effects in the comprehension of abstract and concrete verbal materials. Journal of Experimental Psychology. Learning, Memory, and Cognition, 9, 82–102. Schwanenflugel, P. J. (1991). The psychology of word meanings. In P. J. Schwanenflugel (Ed.), Why are abstract concepts hard to understand? (pp. 223–250). Erlbaum: Hillsdale, NJ. Shimamura, A. P. (2014). Remembering the past: Neural substrates underlying episodic encoding and retrieval. Current Directions in Psychological Science, 23, 257–263. Spencer, H. (1899). The principles of psychology. New York: D. Appleton. Staugaard, S. R., & Berntsen, D. (2014). Involuntary memories of emotional scenes: The effects of cue discriminability and emotion over time. Journal of Experimental Psychology: General, 143(5), 1939–1957. Strain, E., Patterson, K., & Seidenberg, M. S. (1995). Semantic effects in single word naming. Journal of Experimental Psychology. Learning, Memory, and Cognition, 21(5), 1140–1154. Uzer, T., & Brown, N. R. (under review). The effect of cue content on retrieval from autobiographical memory. (under review). Uzer, T., Lee, P. J., & Brown, N. R. (2012). On the prevalence of directly retrieved autobiographical memories. Journal of Experimental Psychology: Learning, Memory, and Cognition, 38, 1296–1308. Vannucci, M., Pelagatti, C., Hanczakowski, M., Mazzoni, G., & Rossi Paccani, C. (2014). Why are we not flooded by involuntary autobiographical memories? Few cues are more effective than many. Psychological Research. http://dx.doi.org/10.1007/s00426-0140632-y (Advance online publication). Wagenaar, W. A. (1986). My memory: A study of autobiographical memory over six years. Cognitive Psychology, 18, 225–252. West, W. C., & Holcomb, P. J. (2000). Imaginal, semantic, and surface-level processing of concrete and abstract words: And electrophysiological investigation. Journal of Cognitive Neuroscience, 12(6), 1024–1037. Whaley, C. P. (1978). Word–nonword classification times. Journal of Verbal Learning and Verbal Behavior, 17(2), 143–154. Whitten, W. B., & Leonard, J. M. (1981). Directed search through autobiographical memory. Memory and Cognition, 9, 566–579. Williams, J. M. G., Barnhofer, T., Crane, C., Hermans, D., Raes, F., Watkins, E., et al. (2007). Autobiographical memory specificity and emotional disorder. Psychological Bulletin, 133, 122–148. Williams, M. D., & Hollan, J. D. (1981). Process of retrieval from very long-term memory. Cognitive Science, 5, 87–119. Wisniewski, E. J., & Murphy, G. L. (1989). Superordinate and basic category names in discourse: A textual analysis. Discourse Processes, 12, 245–261.