Semantic-to-autobiographical memory priming is ubiquitous

Abstract

It is now well established that the activation of semantic memories leads to the activation of autobiographical memories. Studies have shown that semantic processing of words or pictures primes autobiographical memories on voluntary and involuntary autobiographical memory tasks (the Crovitz cue-word task and the vigilance task). Known as semantic-to-autobiographical memory priming, our goal in the current study was to demonstrate the ubiquitous nature of this form of priming by showing that a wide variety of stimuli will prime involuntary autobiographical memories on the vigilance task. In Experiment 1, semantic-to-autobiographical priming was obtained on the vigilance task following the processing of sounds (e.g., the sound of bowling) and spoken words (e.g., the word bowling). In Experiment 2, semantic-to-autobiographical priming was observed on the vigilance task following tactile processing (e.g., the objects ball, glasses) and visual word processing (e.g., the words ball, glasses). In Experiment 3, semantic-to-autobiographical priming was observed on the vigilance task following the processing of videos (e.g., videos of a marching parade) and visual word processing (e.g., the word parade). The results of these experiments support the idea that semantic-to-autobiographical activations occur across a wide variety of stimuli (e.g., linguistic, perceptual). The results also further support the idea that semantic-to-autobiographical memory priming may play an important role in the production of involuntary memories in everyday life. Additional implications (for priming theory and autobiographical memory functions) are discussed.

Introduction

Throughout the course of a day, one processes massive amounts of information from many different sources (e.g., reading a newspaper or a book, watching television, engaging in conversation, etc.). Conway has argued that such routine general information processing continuously activates memories in the autobiographical knowledge base (e.g., Conway, 2001; Conway, 2005). Mace and colleagues (e.g., Mace, 2010; Mace et al., 2019) named this process semantic-to-autobiographical memory priming, and they, along with Conway (2005), have argued that occasionally these activations become conscious where they are experienced as involuntary autobiographical memories (see reviews of involuntary memories in Berntsen, 2009, and Mace, 2007). Mace et al. (2019) proposed a working model for semantic-to-autobiographical memory priming that is based on notions of spreading activation. They argued that semantic access involves both a spread to related generic knowledge (e.g., as in semantic priming models; see Anderson, 1983; Collins & Loftus, 1975), as well as related autobiographical knowledge (see also Mace, 2010). Thus, as one encounters stimuli, both relevant semantic and autobiographical information is activated, and most of the time this information remains unconscious, where it may aid in concept comprehension (Baars, 1988; Mace et al., 2019), but occasionally the autobiographical information may also surface as involuntary memories when one has contact with cues that index the primed memories in some way (e.g., Mace & Hidalgo, 2022).

To illustrate the simplest form of this process, imagine that one processes the concept dog (e.g., sees one in a movie). This processing event will activate a pool of related autobiographical memories, consisting of both general autobiographical memories (e.g., I had a dog when I was little; my sister has dogs; experientially based images of dogs, e.g., my dog; etc.), as well as specific autobiographical memories (i.e., episodic memories, e.g., taking my dog for a walk; playing with my sister’s dogs, etc.). Given that such autobiographical knowledge is likely to remain activated for a significant period of time (days to perhaps weeks, e.g., Mace, 2005; Mace & Petersen, 2020; Mace & Hidalgo, 2022), such activations have the potential of entering consciousness as involuntary memories when cues (or primes) cause further activation in the system. For example, if one hears a dog bark or sees someone walking or playing with a dog, these cues could induce the spontaneous retrieval of any of the associated autobiographical memories (e.g., I had a dog when I was little; playing with my sister’s dogs).

Early evidence of semantic-to-autobiographical activations can be found in Conway (1990), where autobiographical memory retrieval was shown to be faster when it followed semantic primes. Conway’s (1990) original study focused on the relations between concepts and autobiographical memories, whereas more recent work focused on the role of semantic-to-autobiographical priming in the production of involuntary memories on laboratory tasks (e.g., Mace et al., 2019; Mace & Unlu, 2020). In the first of these, Mace et al. (2019) presented participants with a list of common words (e.g., music, sports, summer), where they were to rate the words on the familiarity of their meaning. Following this task, participants were engaged in an involuntary memory task (i.e., the vigilance task; Schlagman & Kvavilashvili, 2008),¹ where primed participants showed more involuntary memories involving primed content (e.g., autobiographical memories about music, sports, summer) than control participants (see also related work in Ball, 2007, and Sheldon et al., 2020).² Also using the vigilance task, Mace and Unlu (2020) replicated these results, while also showing that words embedded in sentences (e.g., Jim likes sports), as well as images representing concepts (e.g., an image of American football), had equally primed the production of involuntary memories.

As these studies had established some of the fundamentals of semantic-to-autobiographical memory priming, subsequent work examined prime repetition and the time course of priming. Examining prime repetition, Mace and Kruchten (2023) presented primes one or three times, and they found that primed involuntary memory production on the vigilance task was significantly greater following three repetitions. They argued that the results suggest that stimulus (or concept) repetition may be a factor in determining if routine semantic-to-autobiographical memory activations transform into conscious autobiographical memories. Thus, the more one experiences a particular prime, the greater the likelihood that its unconsciously activated autobiographical knowledge can enter consciousness, where one becomes aware of such content and it is experienced as a spontaneous memory (see Barzykowski and colleagues’ awareness threshold; e.g., Barzykowski & Staugaard, 2016; Barzykowski, Niedźwieńska, & Mazzoni, 2019a). Relatedly, Mace and Hidalgo (2022) argued that the duration of semantic-to-autobiographical activations may also be a factor in conscious involuntary memory production. If, for example, such activations only lasted for minutes to hours, then the probability that they will result in the conscious retrieval of involuntary memories may be low compared to activation durations of days to weeks, as one should be more likely to encounter multiple subsequent retrieval cues and primes in the latter case. Obtaining evidence supporting the long-term nature of semantic-to-autobiographical memory priming, Mace and Hidalgo (2022) found semantic-to-autobiographical priming on the vigilance task did not diminish after a delay of 7 days compared to a delay of several minutes. Mace and Hidalgo (2022) further argued that semantic-to-autobiographical priming may last weeks or even months, as priming in the 7-day condition was as strong as it was in the immediate condition (see Coane & Balota, 2009, who reported long-term semantic priming).

Our goal in the present study was to further advance the semantic-to-autobiographical memory priming paradigm by demonstrating its ubiquitous nature. As noted, Conway (2005) argued that a wide variety of generic stimuli cause unconscious activations of autobiographical memories. However, with one exception (Experiment 2, Mace & Unlu, 2020, where picture primes were used), semantic-to-autobiographical priming has only been demonstrated with verbal stimuli (i.e., words in isolation or in a sentence). Thus, in this study we set out to show that semantic-to-autobiographical priming results from a wider variety of stimuli, further establishing the ubiquity claim.

To accomplish this objective, we employed three experiments that assessed the effects of semantic-to-autobiographical priming on involuntary memory production in the vigilance task. The primes consisted of a variety of nonverbal stimuli, along with verbal stimuli, mostly for comparison purposes. Experiment 1 utilized sounds (e.g., airplane, lawnmower, alarm clock) and spoken words as primes (e.g., airplane, lawnmower, alarm clock). Here, we wanted to show that auditory nonverbal stimuli, as well as auditory verbal stimuli primed autobiographical memory production, extending semantic-to-autobiographical priming to auditory perception. Experiment 2 utilized objects in vivo (e.g., ball, gloves, cellphone), where participants processed such objects by touch, with an eye toward extending semantic-to-autobiographical priming to tactile perception. Experiment 3 utilized videos (e.g., people running) as primes, with an eye toward extending semantic-to-autobiographical priming to complex visual contexts (e.g., the perception of action). Both Experiment 2 and Experiment 3 also included verbal primes (i.e., printed words, e.g., ball, running) that matched the nonverbal primes.

Our main interest in these experiments was to establish the ubiquity of semantic-to-autobiographical priming. In addition, we reasoned that the results would also serve one additional purpose, further linking generic priming to involuntary memory production. As reviewed, our studies on semantic-to-autobiographical priming have mainly focused on how this form of priming may influence the production of involuntary memories. In those studies, we have shown repeatedly that semantic-to-autobiographical priming does influence involuntary memory production (Mace et al., 2019; Mace & Unlu, 2020), that its influence can be long lasting (Mace & Hidalgo, 2022), and that it does indeed cause involuntary memory production (Mace & Kruchten, 2023). Consequently, we have argued that semantic-to-autobiographical priming has a significant role in everyday involuntary memory production (see a similar priming account in Kvavilashvili & Mandler, 2004). For this study, we reasoned that this role for generic priming would be furthered if the present set of experiments demonstrated that multiple different stimulus types prime involuntary memories, as this would expand its range of influence.

Beyond these considerations, we reasoned that the results of this study may also have implications for the role of autobiographical memory in concept comprehension (e.g., Mace et al., 2019), as well as uniquely position semantic-to-autobiographical priming among priming phenomena. In the first case, we have suggested that autobiographical memory activations may occur during stimulus processing because they play an integral role in the comprehension of stimuli. Thus, as one perceives a stimulus (e.g., dog), personal knowledge and experiences with a stimulus (e.g., my dog spot, walking spot) are activated along with generic semantic knowledge (e.g., barks, chases squirrels), and such activations, together, form the basis for understanding.

Regarding the unique position for semantic-to-autobiographical priming, this form of priming may be the only (or one of few) priming phenomena that shows such wide flexibility in that multiple different stimuli can result in equivalent priming within the same target domain. Of further theoretical interest here are the reasons for stimulus universality in semantic-to-autobiographical priming. Conway (2005) appears to suggest that stimulus universality occurs merely because the autobiographical memory system is highly sensitive to cues. In line with our argument above, we believe that the stimulus universality may be a function of autobiographical memory’s role in concept comprehension. In other words, it occurs because personal experiences with stimuli are part of the context for their understandings. This perspective for stimulus universality is more consistent with the notion that autobiographical memory and semantic memory are more interdependent than originally conceived (e.g., Greenberg & Verfaellie, 2010; Renoult et al., 2012). Regardless of which view ultimately proves correct, demonstrating stimulus universality in semantic-to-autobiographical priming should add it to an already large catalog of unconscious cognitive processes (e.g., Kihlstrom, 1987; Reber, 1993). We discuss these various implications in greater detail in the General discussion.

Experiment 1

Experiment 1 contained two priming groups (sound and auditory word) and one control group (letter decision). Participants in the sound-priming group listened to sounds (e.g., a lawnmower), participants in the auditory-word group listened to the words for each of those sounds (e.g., lawnmower), and participants in the letter-decision group decided if letter strings (e.g., ABC, LXA) were alphabetical. To assess the influence of priming on autobiographical memory content, participants were subsequently engaged in the vigilance task (Schlagman & Kvavilashvili, 2008).

Regarding predictions and potential implications, we predicted that the sound-priming group would show more autobiographical memories related to the primed content (e.g., a personal memory involving lawnmower) than control, and that such findings would indicate that auditory stimuli (i.e., non-speech-based sounds) also cause autobiographical memory access and priming. For the auditory-word group, we also predicted that it would show more autobiographical memories related to the primed content than the control group. Semantic-to-autobiographical priming following word processing has been demonstrated on numerous occasions with visually presented words (e.g., Mace et al., 2019; Mace & Unlu, 2020). In those studies, we reasoned that semantic-to-autobiographical priming would generalize to spoken words, thus indicating that this form of priming would occur at a practical level when one is listening to or engaged in conversation. Accordingly, we reasoned that significant priming in the auditory word group would confirm that generalization. Finally, we predicted that the sound-priming group would not differ from the auditory-word priming group, and that such findings would indicate that both forms of stimuli have similar priming power.

Method

Participants

The participants were 105 undergraduate students from Eastern Illinois University, who participated in exchange for course credit. Fifty-seven of the participants were females, and 48 were males, with an age range of 18–22 years (M = 18.91 years). Participants were randomly assigned to the groups, with equal numbers (35) serving in each group, and similar gender and age compositions (sound priming, M = 19.00 years, 17 males and 18 females; auditory-word priming, M = 19.03 years, 14 males and 21 females; control, M = 18.71 years, 17 males and 18 females). We used G*Power to help us determine sample sizes. Effect sizes in our previous semantic-to-autobiographical priming papers have tended to be varied (η_p²s, .11–.31; ds, 0.98–1.09; Mace et al., 2019; Mace & Unlu, 2020; Mace & Hidalgo, 2022; Mace & Kruchten, 2023). The smallest effect size from the two papers (η_p² = .11, Experiment 1 in Mace & Unlu, 2020), combined with power set at .80 and an alpha level at .05, yielded a total of 84 participants (28 per group × 3 groups), whereas the largest effect size (η_p² = .31, Experiment 3 in Mace et al., 2019) yielded a total of 27 participants (9 per group × 3 groups). According to the Central Limit Theorem, if a sample size is relatively large, consisting of 30 or more individuals, the shape of its sampling distribution can be considered approximately normal, regardless of the shape of the original population distribution. Thus, we reasoned that a total sample size between 90 and 105 (i.e., group sizes ranging from 30 to 35) would be more than sufficient. We used this same calculation for determining sample sizes in the next two experiments.

Materials and design

The stimuli used in the priming phase for the sound and auditory-word priming groups consisted of 30 sounds or 30 spoken words. The sounds were taken from publicly available sound effects websites (https://epicstockmedia.com/; https://www.tunepocket.com/; https://pixabay.com/sound-effects/;https://www.zapsplat.com/sound-effect-categories/). They were selected on the basis of their recognizability and their correspondence with common concepts (e.g., bowling, dog). The sounds represented the most common sounds associated with the concepts (e.g., a bowling ball rolling down an alley and hitting pins, a dog barking). The commonality of the primes was determined by the investigators intuitively, while their recognizability was determined by a small group of pre-experimental participants (n = 6). These individuals listened to the sounds, and all were able to accurately identify the 30 primes. The words were selected to match the sound stimuli (e.g., bowling, dog; see Appendix A for the complete list of sound and word stimuli). In addition to these stimuli, there were eight abstract stimuli included in the sound stimuli list, and eight non-English words included in the word stimuli list (the words people or table in Azerbaijani, Croatian, German, Hungarian, Polish, Portuguese, and Turkish). These items were added to legitimize the nominal purpose of the task, to recognize sounds or spoken words. The stimuli for the filler task phase (control) consisted of 38 non-word, letter strings (e.g., EFG). The strings ranged from three to five letters, with one-half arranged in alphabetical order (e.g., EFG), one-half not (e.g., BCDZ). The sound stimuli were loaded into a single Windows Media Player file, and they ranged in duration from 5 to 15 s, with 5-s intervals in between them. The words were spoken by an adult female at a normal conversation pace, and they were loaded into a single Windows Media Player file with 5-s intervals in between them. The order of the priming stimuli was randomized, with the stipulation that associated items (e.g., cat, dog) did not follow one another and the abstract sounds or non-English words were spaced evenly throughout the list. The stimulus order of the sound and word lists were identical. The stimuli for the filler task were printed and presented in booklet form.

The stimuli for the vigilance task consisted of 512 slides containing either horizontal or vertical lines, each with a unique word phrase (e.g., going to work, growing a garden) embedded in the center of the slide. 477 of the slides contained horizontal lines, 35 contained vertical lines. Four of the slides were used as practice and were unrelated to the primes. Forty-two of the embedded phrases contained words that overlapped with sound or word priming stimuli (e.g., barking dog, children playing, parking a car, with a few primes appearing in more than one phrase, e.g., getting a dog), 67 could be seen as associated with the primes (e.g., sitting in the backseat, visiting an animal shelter), whereas most (399) appeared to be unrelated (e.g., reaching for the stars, growing a garden). The vast majority of the slides were either neutral or positive (484, e.g., starting a conversation, getting ready, going for a drive, reaching for the stars, enjoying the view, having fun), with a much smaller number that could be perceived as negative (28, e.g., missed opportunity, eating awful food, hurting your foot, falling down, shivering in the cold, being exhausted).

Procedure

All participants were tested individually. Sound and auditory-word priming group participants were first engaged in their respective priming task, followed by the vigilance task. Participants in the control group were engaged in the letter-string task, followed by the vigilance task. Each task was presented to participants as a separate study, and the time lag between the priming or filler task and the memory task was approximately 5 min on average. In the sound-priming task, participants were told that they were being involved in a study on sound recognition. They were instructed to listen to the sounds and indicate “yes” if they recognized what a sound was, “no” if they did not. Once they understood the task, the audio file of sounds was played for them, and they were to indicate “yes” (using the letter “Y”) or “no” (using the letter “N”) on an answer sheet that was numbered 1–38 with blank lines for each response. In the auditory-word priming task, participants were told that they were being involved in a study on word recognition. They were instructed to listen to the words and indicate “yes” if they had recognized a word, “no” if they did not. Once they understood the task, the audio file of spoken words was played for them, and they were to indicate “yes” (using the letter “Y”) or “no” (using the letter “N”) on the same type of answer sheet as used for sound-priming participants. In the control group’s filler phase, participants were told the study concerned alphabetical order recognition. They were instructed to view the letter strings printed in the booklet, and decide if they were in alphabetical order, answering “yes” or “no” by circling one of the two responses printed next to each letter string. In the vigilance task, all participants were told that they were being involved in a study on concentration. They were told they would see slides with either horizontal or vertical lines, and they were to say “yes” out loud whenever the slides contained vertical lines. They were further told that the slides would also contain phrases, but they were to ignore them. To ensure their understanding of the instructions, they then received four practice slides, one containing vertical lines. Once it was clear that they understood, they were further instructed that it was possible that they may experience task-unrelated spontaneous thoughts or memories (the concept of involuntary memories was explained to them), and if they experienced one or the other, they were to click the mouse and record them in a booklet that contained sheets of lined paper. Once this aspect was understood, the task commenced. On trials where participants clicked the mouse, an instruction screen would pop up reminding them to record their thoughts or memories in the booklet, clicking the mouse again when finished to return to the vigilance program. The slides were presented randomly for 1.5 s on a computer screen via the SuperLab (version 4.5) software. Following the vigilance task, all participants were asked to read through their booklet, marking entries as either spontaneous thoughts or memories. They were further instructed on the differences between general (i.e., more abstract autobiographical memories, such as “I went to London in 2005”) and specific autobiographical memories (i.e., episodic memories, such as remembering “having a flat tire while driving across the Tower Bridge”), and they were asked to mark their memories as general or specific.

Categorization method

The content of the autobiographical memories for both the control and priming groups was read by two independent judges, who were blind to the hypotheses. The judges were looking for memories that clearly were personal memories involving the concepts presented in the priming phase. For example, personal memories involving, car, bowling, dog, and so forth (e.g., I remember when I first got my dog; I remember when we won the bowling tournament). Memories were deemed as primed (conceptually overlapping) memories only in cases where the content of the memory clearly involved the content of items from the priming list, as in the examples given. Memories were deemed primed if they involved the explicit (e.g., I remember when we won the bowling tournament) or implicit description of a concept (e.g., I remember a time when I threw gutter balls almost all night, a clear reference to a bowling memory), as conceptual overlap was the critical factor. In no event were memories deemed primed if the primes were used as tangential references (e.g., I remember we spent three weeks in Florida, visiting Disney Land, …, and I think we traveled by car on that trip). In cases where memories contained overlapping content from two (or more) of the target items, the memories were only counted once in the scoring process. Disagreements between the judges were settled through discussion. The Kappa statistic indicated high concordance between the judges on the categorizations, K = .93, SE = .02.

Results and discussion

An examination of the total number of spontaneous memories and thoughts reported on the vigilance task showed that the sound-priming group reported an average of 17.63 (SD = 14.62, scores ranged from 1 to 56) memories and 2.57 (SD = 3.94, scores ranged from 0 to 15) thoughts on the vigilance task, the auditory-word priming group reported an average of 18.48 (SD = 18.01, scores ranged from 1 to 72) memories and 1.86 (SD = 3.24, scores ranged from 0 to 15) thoughts, and the control group reported an average of 12.14 (SD = 9.88, scores ranged from 0 to 33) memories and 2.62 (SD = 3.07, scores ranged from 0 to 13) thoughts. Two participants reported no memories (both in control). Two independent-samples analysis of variance (ANOVA) tests (one conducted on the memories, the other on thoughts) found no significant differences among the groups in either the total memories or thoughts reported (memories, F(2, 102) = 1.95, MSE = 212, p = .14, η_p² = .04 [effect size 95% CI: < .01, .12]; thoughts, F(2, 102) = 0.54, MSE = 12, p = .55, η_p² = .01[effect size 95% CI: < .01, .06]). To further confirm the null results from frequentist analyses, two separate Bayesian ANOVAs (one conducted on the memories, the other on thoughts) were conducted using BayesFactor 0.9.12 package (Rouder et al., 2012) in R 4.1.1 (https://www.r-project.org) with default priors. These Bayesian statistics also indicated no group differences in memories and thoughts (memories, BF [Bayes Factor] = 0.43; thoughts, BF = 0.14). Both Bayes Factors are calculated as the ratio between the likelihood of the data given the model versus that of the null model. Note that a BF > 3 is conventionally considered as evidence favoring the alternative hypothesis (Dienes, 2021; Keysers et al., 2020; Lee & Wagenmakers, 2013). Moreover, a Box-Cox power family of transformation was applied to total memories, total thoughts that were positively skewed. Specially, the maximum-likelihood estimation was used to optimize the tuning power parameter, λ, in a way that the distribution of the transformed data has the largest similarity to normality. When zero or negative values occur in the data, a second parameter, 𝛾, was also estimated to reduce transformation bias and then added to make all values positive before the data can be handled by the Box-Cox transformation (Box & Cox, 1964; Fox & Weisberg, 2019; Hawkins & Weisberg, 2017). Another two ANOVA tests with transformed data again found no significant group differences in memories and thoughts (memories, F(2, 102) = 2.35, MSE = 7.11, p = .10, η_p² = .04 [effect size 95% CI: < .01, .13]; thoughts, F(2, 102) = 2.86, MSE = 11.86, p = .06, η_p² = .05 [effect size 95% CI: < .01, .14]). Finally, of the memories reported, 63–65% were specific memories (63% control group; 65% sound and auditory-word priming groups), with the balances being general autobiographical memories. A chi-square test found no differences among the groups in specific/general memory generation, χ² (2) = 0.61, p = .74, φ = 0.02. Thus, these results show that the groups were roughly equivalent in memory type generation, and in total spontaneous thought or memory generation, though the priming groups trended higher in spontaneous memory generation, a trend that has been observed in all semantic-to-autobiographical priming studies (e.g., see Mace & Unlu, 2020; Mace & Hidalgo, 2022; and also discussions in Mace & Unlu, 2020; Mace & Kruchten, 2023).

Turning to the priming data, we first examined the stimulus recognition rates on the priming task for both priming groups. This analysis showed participants in both priming groups indicating that they had recognized all of the priming stimuli, while they rarely indicated recognition of the foil stimuli (i.e., abstract sounds or non-English words, with both categories receiving less than 1% recognition). Next, we analyzed the vigilance task memory data for semantic-to-autobiographical memory priming. Table 1 shows the results of this analysis for all three groups. The results represent the proportion of autobiographical memories (both specific and general memories) found to involve the concepts presented in the priming phase. As is evident in the table, both priming groups had more memories involving primed content than the control group. These data were subjected to a one-way independent-samples ANOVA, which indicated a significant effect of semantic-to-autobiographical priming, F(2, 102) = 6.61, MSE = 0.03, p = .002, η_p² = .12 (effect size 95% CI: .02, .23). Follow-up with Fisher's least significant difference (LSD) statistic showed that both priming groups differed from the control group, but not from one another (LSD = .08), thereby confirming the presence of roughly equivalent priming in both priming groups. The above results were consistent with those from a Bayesian ANOVA and post hoc Bayesian t-tests (BF_ANOVA = 16.97; BF_{control vs. sound} = 14.38, BF_{control vs. word} = 9.20, BF_{sound vs. word} = 0.26. The post hoc t-tests were uncorrected but the scaling factor for fixed effects was set to a larger value of 1/sqrt(2) instead of a smaller value of 0.5 to make the t-tests more conservative (Oberauer & Eichenberger, 2013).

Table 1 Mean proportions of autobiographical memories (both specific and general) involving primed concepts in the control group, sound, and auditory-word priming groups

Consistent with previous work (Mace & Hidalgo, 2022), we further analyzed the presence of primes in the data, and the cue type distributions among the primed memories (i.e., overlapping, containing primed content; associated, related to primed content; or unrelated cues, contenting no primed content or relations to primed material).³ The cue type analysis sought to determine the distribution of the cue types in the primed memories, whereas the examination of primes in the omnibus data sought to determine how many of the primes appeared in the data. The first analysis showed that all of the primes (see Appendix A) appeared in the primed memories of the two priming groups, while all but four (alarm clock, laughing, skateboard, and wedding) appeared in the control group data. The cue type analysis showed that all three cue types were present in the primed and control data (sound priming, overlapping, 57%, associated, 25%, unrelated, 18%; word priming, overlapping, 63%, associated, 24%, unrelated, 13%; control, overlapping, 48%, associated, 18%, unrelated, 34%). A chi-square test found no differences between the priming groups (χ² (2) = 2.09, p = .35, φ = 0.07), while the control group was found to differ from the priming groups (χ² (4) = 16.45, p =.002, φ = 0.19). Thus, these results show (1) that the data set involved all or nearly all of the primes, indicating a reasonable correspondence between the primes and autobiographical memories, and (2) that the primed memories were cued by all cue types, consistent with our hypothesis, though there was greater weight with overlapping cues, as one should expect (e.g., the encoding specificity principle, Tulving & Thomson, 1973).

In summary, both the sound and auditory-word priming groups exhibited significant semantic-to-autobiographical priming. Finding this effect with sound stimuli suggests that such stimuli cause autobiographical memory activations, and these activations may subsequently surface in consciousness as involuntary memories. While obtaining priming with auditory word stimuli was expected and predicted by previous studies with similar visually presented verbal stimuli (e.g., Mace & Hidalgo, 2022; Mace & Unlu, 2020), the finding reported here confirms conclusions in those studies, which argued that semantic-to-autobiographical priming was likely to result from auditory word perception. This finding suggests, as argued here and elsewhere (e.g., Mace & Unlu, 2020), that semantic-to-autobiographical priming occurs when one is listening to or engaging in conversation. Finally, given apparent equivalence between the sound and word priming groups, it seems that both forms of stimuli are equally capable of causing autobiographical memory access and spontaneous memory production.

Experiment 2

In Experiment 2, we sought to extend the semantic-to-autobiographical priming hypothesis to tactile perception. Tactile-priming participants were given objects to recognize by touch, and then they were treated to the vigilance task. A visual-word-priming group was employed for comparison purposes, and a control group was also employed to assess priming in both groups.

Our predictions, and their implications, were similar to those put forward in Experiment 1. We predicted that the tactile-priming group would show more priming than control, and that this finding would extend the semantic-to-autobiographical priming hypothesis to tactile perception, with respect to autobiographical memory access and involuntary memory production. Additionally, as in Experiment 1 and Mace and Unlu (2020), we predicted that the word-priming group would not differ from the tactile group, and that this would show that tactile stimuli (or processes) can be as powerful as verbal stimuli in the priming of autobiographical memories.

Method

Participants

The participants were 93 undergraduate students from Eastern Illinois University, who participated in exchange for course credit. Fifty-six of the participants were females, and 37 were males, with an age range of 18–28 years (M = 18.74 years). Participants were randomly assigned to the groups, with equal numbers (31) serving in each group, and similar gender and age compositions (tactile priming, M = 18.77 years, 11 males and 20 females; visual-word priming, M = 18.94 years, 13 males and 18 females; control, M = 18.50 years, 13 males and 18 females). Regarding sample sizes, we used the method put forth in Experiment 1 as our guide.

Materials and design

The stimuli used in the priming phase for the tactile and word-priming groups consisted of 26 objects or 26 printed words (see Appendix B for the complete list). The objects were selected on the basis of their recognizability and commonality (e.g., key, toothbrush). As in Experiment 1, we judged the commonality of the objects intuitively, and we presented the 26 objects to pre-experimental participants (n = 7), who all were able to accurately identify the objects. The words were selected to match the tactile stimuli. In addition to these stimuli, there were eight uncommon objects included in the tactile stimuli list (e.g., tuning fork, sound-level meter), and eight non-English words included in the word stimuli list (same as Experiment 1). These items were added to legitimize the nominal purpose of the tasks, to recognize objects or words. The words were printed to be presented in booklet form. The stimuli for the filler task controls were the same as in Experiment 1. The order of the priming stimuli was randomized, with the stipulation that associated items (e.g., glove, hat) did not follow one another and the uncommon objects or non-English words were spaced evenly throughout the list. The stimulus order of the object and word lists were identical.

The stimuli for the vigilance task consisted of the 512 slides used in Experiment 1, with some changes made to the phrases to accommodate the priming stimuli (e.g., wearing a watch). Thirty-two of the embedded phrases contained words that overlapped with priming stimuli (e.g., getting new glasses, drinking from a bottle, with a few primes appearing in more than one phrase, e.g., new tennis shoes, running sneakers), 17 could be seen as associated with the primes (e.g., changing the channel; keeping track of time), whereas most (459) appeared to be unrelated (e.g., reaching for the stars, growing a garden). All other characteristics of the slides were the same (i.e., number of practice slides, number of slides containing horizontal or vertical lines, valence composition).

Procedure

All participants were tested individually. The procedure for the filler task phase in the control group was identical to Experiment 1. For the tactile-priming group, participants were told that they were engaging in a study on object recognition through touch. Participants wore a sleeping mask blindfold, which completely occluded their vision. Once blinded folded, the experimenter placed objects in their hands one at a time. Participants were instructed to feel the object and to call out “yes” if they recognized the object, “no” if they did not. Once they responded, the object was taken from them and they were then given another object. Participants chose which hand they wanted the object placed in and the recognition process was done at their pace. In no event did participants see the objects, with the exception of the debriefing phase, if they requested to view them. For the word-priming group, participants were told that they were engaging in a study on word recognition. They were instructed to read the words on the printed form, deciding if they recognize them, circling “yes” or “no” options that was printed next to each item. Participants performed this task at their own pace. As in Experiment 1, the vigilance task immediately followed the priming or filler phases, and procedures for it, and the post-task thought and memory classification phase, were identical to Experiment 1.

Categorization method

Using the same criteria specified for Experiment 1, the judges examined the memories of the priming groups and the control group for content overlap. As in Experiment 1, the judges were blind to the hypotheses. The Kappa statistic indicated high concordance between the judges, K = .89, SE = .01.

Results and discussion

An examination of the total number of spontaneous memories and thoughts reported on the vigilance task showed that the tactile-priming group reported an average of 13.58 (SD = 10.11, scores ranged from 1 to 44) memories and 2.10 (SD = 3.25, scores ranged from 0 to 14) thoughts on the vigilance task, the word-priming group reported an average of 13.03 (SD = 7.46, scores ranged from 2 to 32) memories and 2.13 (SD = 2.93, scores ranged from 0 to 12) thoughts, and the control group reported an average of 12.03 (SD = 10.04, scores ranged from 0 to 47) memories and 2.03 (SD = 3.21, scores ranged from 0 to 12) thoughts. One participant reported no memories in the control group. Two independent-samples ANOVA tests (one conducted on the memories, the other on thoughts) found no significant differences between the groups in either the total memories or thoughts (memories, F(2, 90) = 0.22, MSE = 86.32, p = .80, η_p² < .01 [effect size 95% CI: < .01, .04]; thoughts, F(2, 90) < 0.01, MSE = 9.81, p = .99, η_p² < .01[effect size 95% CI: < .01, < .01]). Again, Bayesian statistics indicated no group difference in memories and thoughts (memories, BF [Bayes Factor] = 0.12; thoughts, BF = 0.10). Another two ANOVA tests with transformed data also found no significant group differences in memories and thoughts (memories, F(2, 90) = 0.57, MSE = 0.33, p = .57, η_p² = .01 [effect size 95% CI: < .01, .07]; thoughts, F(2, 90) = 0.13, MSE = 0.56, p = .88, η_p² < .01 [effect size 95% CI: < .01, .03]). Of the memories reported, 66–69% were specific memories (66%, control group; 67%, tactile group; 69%, word group), with the balances being general autobiographical memories. A chi-square test found no differences among the groups on these memory characteristics, χ² (2) = 0.81, p = .67, φ = 0.03. Thus, these results show that the groups were roughly equivalent in memory type generation, and in total spontaneous thought or memory generation, though the priming groups trended higher in spontaneous memory generation, as in Experiment 1 and elsewhere (e.g., Mace & Hidalgo, 2022; Mace & Unlu, 2020).

Turning to the priming data, we first examined the stimulus recognition rates on the priming task for both priming groups. This analysis showed participants in both priming groups indicating high recognition of the priming stimuli (100% in the word group, 99.4% in the tactile group, with a few participants failing to recognize either medicine bottle, earbuds, button, ruler, or remote). The recognition of foil stimuli was quite low (less than 1%, with no recognition of non-English words, and five participants indicating recognition of the tuning fork). Next, we analyzed the vigilance task memory data for semantic-to-autobiographical memory priming. Table 2 shows the results of this analysis for all three groups. The results represent the proportion of autobiographical memories (both specific and general memories) found to involve the concepts presented in the priming phase. As is evident in the table, both priming groups had more memories involving primed content than the control group. These data were subjected to a one-way independent-samples ANOVA, which indicated a significant effect of semantic-to-autobiographical priming, F(2, 90) = 6.00, MSE = 0.02, p = .004, η_p² = .12 (effect size 95% CI: .01, .24). Follow-up with Fisher's LSD statistic showed that both priming groups differed from the control group, but not from one another (LSD = .08), thereby confirming the presence of roughly equivalent priming in both priming groups, as in Experiment 1. The above results were consistent with those from a Bayesian ANOVA and post hoc Bayesian t-tests (BF_ANOVA = 10.55; BF_{control vs. tactile} = 11.84, BF_{control vs. word} = 10.77, BF_{tactile vs. word} = 0.26).

Table 2 Mean proportions of autobiographical memories (both specific and general) involving primed concepts in the control group, tactile, and word priming groups

As in Experiment 1, we further analyzed the presence of primes in the data, and the cue type distributions among the primed memories (i.e., overlapping, associated, or unrelated cues).⁴ The first analysis showed that 16–17 of the 26 primes (62%, word group; 65%, tactile group) appeared in the primed memories of the two priming groups, while nine of the primes (35%) appeared in the control group data. The cue type analysis showed that all three cue types were present in the primed data (tactile priming, overlapping, 72%, associated, 24%, unrelated, 4%; word priming, overlapping, 62%, associated, 34%, unrelated, 4%), but only two types appeared in the control data (overlapping, 63%, associated, 37%). A chi-square test found no differences between the priming groups and control, χ² (4) = 3.06, p = .55, φ = 0.09. The first analysis showed that approximately two-thirds of the primes were found in the priming data, unlike Experiment 1 which showed all to be present. Perhaps this indicates that these primes are associated with fewer autobiographical memories and therefore less likely to trigger them (e.g., imagine how many memories one may have involving buttons). The cue type analysis was more similar to the analysis done in Experiment 1, though there was much less involvement with unrelated cues, perhaps also owing to a smaller number of available memories.

In summary, both the tactile and word-priming groups exhibited significant semantic-to-autobiographical priming. Observing the priming effect with tactile stimuli suggests that the processes of tactile perception also cause access to autobiographical memories, and these memories may also subsequently surface in consciousness as involuntary memories. Although semantic-to-autobiographical priming following visual-word processing has been demonstrated many times (e.g., Mace & Hidalgo, 2022; Mace & Unlu, 2020), its purpose here was to allow for comparison to the tactile-priming group, and here the groups did not appear to differ from one another. This finding suggests that tactile stimuli are as a powerful as word stimuli in the priming of autobiographical memories, as suggested for a similar comparison in Experiment 1. We should note that while there was less involvement of the primes in the data relative to Experiment 1, this should not be seen as a function of tactile processes, but instead the relation of the primes to autobiographical memories, as both the priming groups produced equivalent priming. We discuss this point further in the General discussion.

Experiment 3

Mace and Unlu (2020) found semantic-to-autobiographical priming following the processing of picture primes. In Experiment 3, we sought to extend semantic-to-autobiographical priming to more complex visual processing, viewing action/activity scenes. Participants in a video priming group viewed videos involving various actions and activities (e.g., a parade marching, individuals swimming or running), followed by the vigilance task. As in Experiment 2, a visual-word priming group was employed for comparison purposes, and a control group was also employed to assess priming in both groups. We predicted that the video priming group would show more priming involving the primed concepts (e.g., swimming, running) than the control group, and that video priming would be similar in magnitude to word priming.

Method

Participants

The participants were 102 undergraduate students from Eastern Illinois University, who participated in exchange for course credit. Sixty-two of the participants were females, and 40 were males, with an age range of 18–26 years (M = 19.36 years). Participants were randomly assigned to the groups, with equal numbers (34) serving in each group, and similar gender and age compositions (video priming, M = 19.62 years, 13 males and 21 females; word priming, M = 19.00 years, 14 males and 20 females; control, M = 19.47 years, 13 males and 21 females). Again, we followed the sample size guideline described for Experiment 1.

Materials and design

The stimuli used in the priming phase for the video and word priming groups consisted of 41 videos or 41 printed words (see Appendix C for the complete list). The videos were taken from publicly available website (YouTube), and they were selected on the basis of their recognizability and commonality (e.g., scenes of graduation, a parade). Most of the videos were of actions/activities (e.g., riding a bike), but a few contained potentially ambiguous actions (i.e., a cat walking, scanning scenes of a lake or mountain, see Appendix C). The videos contained no text, their audio tracks were muted, and they varied in length from 1 to 9 min). As in Experiments 1 and 2, we judged the commonality of the videos intuitively, and we presented them to pre-experimental participants (n = 6), who all were able to ascribe actions to the videos that comported well with the scenes. The words were selected to match the videos (e.g., graduation, parade, cat, lake, and mountain). In addition to these stimuli, there were eight uncommon scenes included in the video list (i.e., abstract images, or astronomical images, e.g., nebulas), and eight non-English words included in the word stimuli list (same as Experiments 1 and 2). These items were added to legitimize the nominal purpose of the tasks, to recognize scenes or words. The videos were loaded into the Microsoft PowerPoint program, with each video occupying a single, numbered slide (numbered 1–49). The words were printed to be presented in booklet form, as in Experiment 2. The stimuli for the filler task controls were the same as in Experiments 1 and 2. The order of the priming stimuli was randomized, with the stipulation that associated items (e.g., winter, storm) did not follow one another and the uncommon scenes or non-English words were spaced evenly throughout the list. The stimulus order of the videos and word lists were identical.

The stimuli for the vigilance task consisted of the 512 slides used in Experiment 1. Forty-nine of the embedded phrases contained words that overlapped with priming stimuli (e.g., going for a swim, growing a garden, with a few primes appearing in more than one phrase, e.g., swimming in a pond, riding in a car, parking a car), 101 could be seen as associated with the primes (e.g., going for a drive, planting flowers), whereas most (358) appeared to be unrelated (e.g., making a new friend, finding strength). All other characteristics of the slides were the same as previous experiments (i.e., number of practice slides, number of slides containing horizontal or vertical lines, valence composition).

Procedure

All participants were tested individually. The procedure for the filler task phase in the control group was identical to Experiments 1 and 2, whereas the procedure for the word priming phase in the word group was identical to Experiment 2. For the video-priming group, participants were told that they were engaging in a study on activity recognition. They were instructed to watch each video by scrolling through the PowerPoint slides and indicate “yes” if they recognized the action in the video, “no” if they did not. Once they understood the task, the PowerPoint program was started for them in slideshow mode. Participants worked through the slides at their own pace, and they indicated their recognition responses by circling “yes” or “no” on an answer sheet that was numbered 1–49, with each number corresponding to the number on the PowerPoint slide. As in Experiments 1 and 2, the vigilance task immediately followed the priming or filler phases, and procedures for it, and the post-task thought and memory classification phase, were identical to Experiments 1 and 2.

Categorization method

Using the same criteria specified for Experiment 1, the judges, both blind to the hypotheses, examined the memories of the priming groups and the control group for content overlap (e.g., memories about hiking, swimming, running, etc.). The Kappa statistic indicated high concordance between the judges, K = .91, SE = .01.

Results and discussion

An examination of the total number of spontaneous memories and thoughts reported on the vigilance task showed that the video-priming group reported an average of 14.94 (SD = 14.78, scores ranged from 2 to 72) memories and 2.62 (SD = 2.60, scores ranged from 0 to 32) thoughts on the vigilance task, the word-priming group reported an average of 15.50 (SD = 14.23, scores ranged from 1 to 64) memories and 2.38 (SD = 3.96, scores ranged from 0 to 28) thoughts, and the control group reported an average of 11.38 (SD = 9.59, scores ranged from 0 to 43) memories and 2.44 (SD = 2.63, scores ranged from 0 to 11) thoughts. Two participants reported no memories in the control group. Two independent-samples ANOVA tests (one conducted on the memories, the other on thoughts) found no significant differences among the groups in either the total memories or thoughts (memories, F(2, 99) = 0.99, MSE = 171.01, p = .38, η_p² = .02 [effect size 95% CI: < .01, .09]; thoughts, F(2, 99) = 0.05, MSE = 9.80, p = .95, η_p² < .01 [effect size 95% CI: < .01, .01]). Again, Bayesian statistics indicated no group difference in memories and thoughts (memories, BF [Bayes Factor] = 0.20; thoughts, BF = 0.09). Another two ANOVA tests with transformed data also found no significant group differences in memories and thoughts (memories, F(2, 99) = 1.20, MSE = 1.19, p = .31, η_p² = .02 [effect size 95% CI: < .01, .10]; thoughts, F(2, 99) = 0.95, MSE = 3.30, p = .39, η_p² = .02 [effect size 95% CI: < .01, .09]). Of the memories reported, 63–67% were specific memories (63%, control group; 64%, video group; 67%, word group), with the balances being general autobiographical memories. A chi-square test found no differences among the groups on these memory characteristics, χ² (2) = 1.77, p = .41, φ = 0.04. As in Experiments 1 and 2, these results show that the groups were roughly equivalent in memory type generation, and in total spontaneous thought or memory generation, with the priming groups trending higher in spontaneous memory generation, once again.

Turning to the priming data, we first examined the recognition rates on the priming task for both priming groups. This analysis showed participants in both priming groups indicating high recognition of the priming stimuli (100% in the word group, 96% in the video group). The recognition of foil stimuli was quite low (less than 10% in the video group, with recognition occurring mostly with astronomical videos, and no recognition of non-English words in the word group). Next, we analyzed the vigilance task memory data for semantic-to-autobiographical memory priming. Table 3 shows the results of this analysis for all three groups. The results represent the proportion of autobiographical memories (both specific and general memories) found to involve the concepts presented in the priming phase. As is evident in the table, both priming groups had more memories involving primed content than the control group. These data were subjected to a one-way independent-samples ANOVA, which indicated a significant effect of semantic-to-autobiographical priming, F(2, 99) = 7.78, MSE = 0.04, p < .001, η_p² = .14 [effect size 95% CI: .03, .25]. Follow-up with Fisher's LSD statistic showed that both priming groups differed from the control group, but not from one another (LSD = .10), thereby confirming the presence of roughly equivalent priming in both priming groups, as in Experiments 1 and 2. The above results were consistent with those from a Bayesian ANOVA and post hoc Bayesian t-tests (BF_ANOVA = 41.87; BF_{control vs. video} = 19.09, BF_{control vs. word} = 29.61, BF_{video vs. word} = 0.26).

Table 3 Mean proportions of autobiographical memories (both specific and general) involving primed concepts in the control group, video, and word priming groups

As in Experiments 1 and 2, we further analyzed the presence of primes in the data, and the cue type distributions among the primed memories (i.e., overlapping, associated, or unrelated cues).⁵ The first analysis showed that 39 of the 41 primes (95%) appeared in the primed memories of the two priming groups (sailing and traffic were absent from both), while 35 of the primes (85%) appeared in the control group data (sailing and traffic, along with fishing, winter, woods, and vacuuming were absent from control). The cue type analysis showed that all three cue types were present in the primed and control data (video priming, overlapping, 41%, associated, 35%, unrelated, 24%; word priming, overlapping, 43%, associated, 34%, unrelated, 23%; control, overlapping, 30%, associated, 29%, unrelated, 41%). A chi-square test found no differences between the priming groups (χ² (2) = 0.24, p = .88, φ = 0.02), while the control group was found to differ from the priming groups (χ² (4) = 19.49, p <.001, φ = 0.16). Similar to Experiment 1, the first analysis showed that nearly all of the primes were found in the priming data, while a slightly smaller number of them appeared in the control group’s data. The cue type results were similar to the same analysis done in Experiment 1.

In summary, both the video and word priming groups exhibited significant semantic-to-autobiographical priming. Thus, we have replicated semantic-to-autobiographical priming following linguistic processing and have also extended it to video/scene processing. The novel finding suggests that semantic-to-autobiographical priming occurs as a result of complex visual processing, such as viewing events and actions in everyday life, as well as many similar routine processes (e.g., watching television). As with Experiments 1 and 2, the novel video priming results expands the range of semantic-to-autobiographical priming, and because video priming did not appear to differ from word priming, one could argue that the two stimulus forms produce equivalent priming.

General discussion

Mace and Unlu (2020) showed semantic-to-autobiographical priming following word, sentence, and picture processing. They argued that their results supported the idea that generic priming is ubiquitous, as it is likely to involve multiple different stimuli and processes. The current study was a continuation of that line of reasoning. In three experiments, we have shown that three types of novel stimuli also cause semantic-to-autobiographical priming, further supporting a fundamental premise of stimulus universality in semantic-to-autobiographical priming. Experiment 1 found semantic-to-autobiographical priming following the perception of sounds and spoken words. As noted, the sound priming results suggest that auditory perception of common, recognizable sounds causes autobiographical memory activation and possible subsequent involuntary memory production. Experiment 2 found semantic-to-autobiographical priming following a very different form of cognitive processing, tactile perception, suggesting that autobiographical memory access and production result from these processes as well. Finally, Experiment 3 found semantic-to-autobiographical priming as a result of video processing, suggesting that scene and activity perception also cause autobiographical memory access and production.

The results of these three experiments add to the inventory of stimuli that are known to be involved in semantic-to-autobiographical priming. Because the record now shows that a wide variety of stimuli can prime autobiographical memories, we believe there is now broad support for the view that all manner of stimuli can cause autobiographical memory activations, and such activations can occasionally enter consciousness as spontaneous memories (e.g., Conway, 2005;Mace et al., 2019 ; Mace & Unlu, 2020). Also, as all of the novel stimuli used here were non-verbal, it can now be stated with greater confidence that semantic-to-autobiographical memory priming is not limited to verbal stimuli and linguistic processes. Further, because the priming effects from these stimuli did not differ from their comparable verbal stimuli, one can say that verbal and non-verbal stimuli appear to be equally powerful. Although the scores for tactile priming in Experiment 2 were lower relative to Experiments 1 and 3, it seems that this was a function of the associations that these primes had with autobiographical memories, and not the tactile processes themselves. That is, because one may have fewer available autobiographical memories for some of the stimuli used in Experiment 2 compared to Experiment 3 (e.g., button or ruler vs. school or pet), the likelihood of their surfacing as involuntary memories is diminished. While this observation may seem more intuitive than objective, it is, however, supported by main and supplementary findings in Experiment 2. As noted earlier, both word and tactile priming produced equivalent effects, and each showed the same level of involvement of the primes in the data, thus showing that the reduced scores in the tactile group were a function of the stimuli used and not the processes.

To our knowledge, semantic-to-autobiographical priming is the only priming paradigm to show that a wide variety of different stimuli produce equivalent priming effects. For example, in perceptual word priming (e.g., word completion tasks, see Tulving et al., 1982), priming does not occur or is greatly diminished across stimulus types (e.g., words vs. pictures, see Weldon, 1991), or even across sensory modalities with the same stimulus types (e.g., words read vs. words heard, e.g., Blaxton, 1989; see reviews in Roediger III & McDermott, 1993). While conceptual priming and semantic priming paradigms (e.g., category instance generation, lexical decision) have demonstrated more flexibility than perceptual word priming (see McNamara, 2005; Roediger III & McDermott, 1993), these paradigms have not as yet demonstrated equivalent priming from a wide array of different stimuli. Thus, in semantic-to-autobiographical priming, significantly different stimulus forms produce equivalent priming, suggesting a common underlying factor. We believe that the common factor across stimulus and modality forms in semantic-to-autobiographical priming is concept processing and apprehension.

Relatedly, we have hypothesized that semantic-to-autobiographical priming may be reflective of the role that autobiographical knowledge has in the comprehension of concepts (Mace et al., 2019; Mace & Unlu, 2020). It was argued that autobiographical memory access occurs along with semantic memory access because the former is involved in, or is integral to, concept comprehension. Thus, in this view, autobiographical memory access occurs because personal experiences shapes or colors how we understand or perceive concepts, or it serves as part of the fundamental basis for comprehension (see similar notions of how unconscious contexts shape conscious experience in Baars, 1988, as well as other similar roles for autobiographical knowledge in Barsalou, 1999, 2008; Schank, 1999; and the relations between semantic memory and autobiographical memory in Greenberg & Verfaellie, 2010; Mace et al., 2019; Renoult et al., 2012). While we favor this view of semantic-to-autobiographical priming, we also note that there are alternative explanations. As noted, Conway (2005) appears to suggest that ubiquitous autobiographical memory access occurs simply because the autobiographical memory system is highly (or overly) sensitive to all manner of cues. Consistent with this notion, diary studies of everyday involuntary memories have shown that all manner of stimuli and processes can trigger autobiographical memories (e.g., Ball et al., 2007; Berntsen, 1996; Berntsen & Hall, 2004; Mace et al., 2015; Schlagman et al., 2007). Thus, showing here that multiple different stimuli prime autobiographical memories either supports the sensitivity to cues account, or it supports the view that autobiographical knowledge is involved in stimulus recognition. As the data do not allow us to distinguish between these two views, future work should direct its attention to this aspect of semantic-to-autobiographical priming. Should such work favor the functional account, such an outcome would further cement that view that autobiographical and semantic memory are interdependent.

While the results may be equivocal on this aspect of autobiographical memory function, they otherwise appear to offer clear implications for everyday involuntary remembering, as have numerous other studies on semantic-to-autobiographical priming (e.g., Mace et al., 2019; Mace & Hidalgo, 2022; Mace & Unlu, 2020). In those studies, it was argued that semantic-to-autobiographical priming has a significant role in involuntary memory production, as it may frequently be a source of it. Because the current study has significantly expanded the range of stimuli involved in semantic-to-autobiographical priming, it appears that this view is further supported and expanded. The rationale here appears to be fairly straightforward. Because a wide range of stimuli and processes appear to cause autobiographical memory access, and such cognitive processing is continuous, it seems reasonable to assume that this process is often the cause of involuntary memories, when they do occur. This is not to say that other variables do not play a role in involuntary remembering (e.g., cue characteristics; Berntsen et al., 2013), but to say that given the enormity and frequency of semantic-to-autobiographical priming, it should have a fairly large role in this process. We should note that this conclusion does not (and should not) lead to the prediction that involuntary remembering should be continuous (see Barzykowski, Radel, et al., 2019b; Vannucci et al., 2015), as there are theoretically a universe of constant unconscious processes that do not enter consciousness (see readings on the cognitive unconscious and unconscious contexts, respectively; e.g., Baars, 1988; Kihlstrom, 1987; Reber, 1993). Indeed, there may often be numerous factors present before primed autobiographical memories become conscious. For example, there may need to be significant stimulus repetition, and priming may often interact with other variables (e.g., cue characteristics; Berntsen et al., 2013).

The primes used in these experiments, along with the verbal and picture primes used in other studies (e.g., Mace & Unlu, 2020), cover a wide set of possible processing circumstances in everyday life in which semantic-to-autobiographical priming is likely to occur. While there are just a few remaining processes that have not been examined in the semantic-to-autobiographical priming paradigm (e.g., olfactory perception), we believe that the ubiquity principle predicts that priming should also occur in the remaining untested areas, so long as the processes (or stimuli) have associations with autobiographical memory. Diary studies of everyday involuntary memories may serve as a useful guide here (e.g., Berntsen & Hall, 2004; Ball et al., 2007). As noted, they have shown that a very wide range of different types of experiences serve as cues for involuntary memories. For example, among the untested prime set, these studies have shown that involuntary memories are cued by tastes and smells, internal bodily/physiological (e.g., feeling pain or cold) and emotional states. While these cues have not been shown to have a large presence in involuntary remembering (e.g., Ball et al., 2007), some have otherwise been shown to be powerful elicitors of autobiographical memories (odors and emotions; e.g., Bower, 1981; Chu & Downes, 2000, 2002; Sheldon & Donahue, 2017). It is for this reason, and their established association with autobiographical memory, that we are predicting that they, too, are likely to be involved in semantic-to-autobiographical priming, and future studies should direct their attention to them.

However, it is important to note the power that verbal cues appear to have in the elicitation of involuntary memories. Some studies have shown that verbal cues form a significant proportion of the cues that have been observed to elicit involuntary memories in everyday life (e.g., Mace, 2004, 2005; Mace et al., 2015). Interestingly, for the purposes of this study, Mazzoni, Vannucci, and Batool (2014) have shown that verbal cues in the vigilance task elicit more involuntary memories than pictorial cues, suggesting the former is a more powerful elicitor of involuntary memories. While verbal primes in this study, as well as in Mace and Unlu (2020), were not shown to be superior to other primes, perhaps for the reasons indicted above, it may be of interest to future studies to compare semantic-to-autobiographical priming in the vigilance task using verbal and pictorial cues. It would be interesting to see if under these circumstances semantic-to-autobiographical priming differs across the cue types.

Another interesting possibility for future studies concerns another potential relationship between semantic memory and autobiographical memory; that is, autobiographical-to-semantic priming (Mace, 2010). In this iteration of the priming paradigm, one would examine if the contents of autobiographical memories prime semantic memories. We believe that based on the priming literature on conceptual priming and semantic priming (see McNamara, 2005; Roediger III & McDermott, 1993), autobiographical memories would be shown to prime semantic memories. Obtaining evidence for this relationship would have a number of theoretical implications (e.g., on the relationship between the two memory systems, e.g., Rubin, 2022), including reinforcing the notion that autobiographical memory has a role in concept comprehension.

Before closing, we should note two limitations in the present study. The first concerns the cue types (overlapping, associated, and unrelated) in the vigilance task. As in other studies (e.g., Mace & Hidalgo, 2022), our analysis of these cues showed that overlapping and associated cues outstripped the unrelated category significantly, with the emphasis on overlapping. Theoretically, this finding makes sense, however, we may not have realized the full magnitude of this difference, as unrelated cues outnumbered the two other categories in the vigilance task. Thus, we may have observed a larger difference if there were an equal number of cues in each category. At the very least, a more accurate picture may have been obtained if the number in each of the cue categories were held constant, and thus future studies should endeavor to follow such a design. The second limitation is related to this design issue. As there was a very large number of unrelated cues in the vigilance task, cues which were much less likely to draw on primed memories, the full magnitude of the semantic-to-autobiographical priming effect may not have been realized here. Thus, the design used here may have enhanced the effects of unprimed memories, masking the full effects of primed memories, where a design using equal number of cue types may have revealed much larger influences of priming.

In conclusion, this study has demonstrated semantic-to-autobiographical priming on the vigilance task with three novel stimulus forms. As these findings add to the stimuli already shown to prime autobiographical memories, and together they represent most of the stimuli that one experiences in everyday cognition, we have argued that semantic-to-autobiographical priming is ubiquitous, as originally hypothesized by Conway (2005). We further note that semantic-to-autobiographical priming may be the only form of priming that displays such a wide degree of flexibility. We note that this may be because concept comprehension is the common factor linking different stimuli in semantic-to-autobiographical priming, and this observation may signal a role for autobiographical memory in concept processing. Nevertheless, such a hypothesis awaits support from future research. An unambiguous role for semantic-to-autobiographical priming is its involvement in involuntary remembering. As the many different stimulus forms involved in semantic-to-autobiographical priming have been shown to influence involuntary memories, we believe that this is strong evidence for the notion that semantic-to-autobiographical priming plays a significant role in involuntary remembering in everyday life. As this conclusion may help to explain some involuntary remembering, it may be especially important to investigate the possible role for semantic-to-autobiographical priming in concept comprehension. If autobiographical memories are shown to be involved in concept comprehension, then involuntary remembering may be an extension of this process, or some other as yet unknown, but related, cognitive process. Thus, answering this question may help solve the question of involuntary memory function (e.g., Mace & Atkinson, 2009; Rasmussen & Berntsen, 2009), helping to determine if they are functional, or merely byproducts of other processes (e.g., priming).

Open Practices Statement

The experiments were not preregistered, and due to restrictions in the ethics protocol for this study, the data are not available online. However, requests for the data can be made to the first author.

Appendices

Appendix A

Words and sounds (in parentheses) used in the sound and auditory-word priming groups

Dog (barking dog)

Cell Phone (phone ringing)

Cat (cat meowing)

Car (car engine starting)

Airplane (airplane taking off)

Ballpark (crack of bat and cheering)

Train (train horn and train moving on tracks)

Television (news intro music)

Band (band instruments playing)

Bowling (bowling ball hitting pins)

Fireworks (squeal and pop of fireworks)

Cooking (oil cracking in frying pan)

Storm (thunder and rain)

Singing (people singing)

Laughing (a child laughing)

Ocean (waves crashing and seagull sounds)

Children (children talking and giggling)

Guitar (electric guitar playing)

Alarm clock (beeping alarm clock)

Basketball (ball bouncing and fans cheering)

Lawnmower (pull of chain and sound mower engine)

Baby (baby fussing and crying)

Horse (horse neighing)

Wedding (organ playing the wedding march)

Snoring (a person snoring)

Bus (air brake release and coins entering fare box)

Skateboard (skateboard wheels moving on pavement)

Swimming (swimming and splashing sounds)

Restaurant (dishes clanking and groups of people talking)

Bumble Bee (a bee buzzing)

Appendix B

Words and objects used in the tactile and word priming groups

Ball

Glasses

Cell Phone

Remote

Bottle

Cup

Pen

Book

Video Tape

Key

Mask

Medicine Bottle

Backpack

Ring

Earbuds

Notebook

Coin

Credit card

Comb

Ruler

Glove

Watch

Hat

Sneaker

Button

Toothbrush

Sunglasses

Appendix C

Words and videos (actions/scenes in parentheses) used in the video and word priming groups

School (teacher talking to her class)

Parade (band marching in a parade)

Art (putting paint on a canvas)

Doctor (a patient getting a physical)

Pet (dogs playing)

Clothing (person trying on outfits)

City (scanning city buildings overhead)

Cat (cat walking)

Reading (person reading a book)

Hiking (person on hiking trail)

Movie (people at the movie theatre)

Sports (people playing different sports)

Car (car driving down a road)

Garden (person working in a garden)

Mountain (scanning the tops of mountains)

Beach (waves coming onto the sand)

Bike (person riding a bike)

Swimming (person swimming)

Running (people in a race)

Lake (moving scenes of a lake)

Zoo (animals at the zoo)

Bowling (person bowling)

Storm (rain and wind)

Singing (children singing)

Winter (scene of snow falling)

Woods (person walking in the woods)

Cooking (people cooking)

Sledding (people sledding)

Boat (boats moving in the water)

Shopping (person shopping for groceries)

Graduation (students receiving diplomas)

Haircut (child getting a haircut)

Traffic (cars in traffic jam)

New Year (counting clock and ball drop in Times Square NYC)

Fishing (a child reeling in a fish)

Sailing (people on a sailboat)

Ocean (scenes of waves crashing)

Baseball (scenes of a baseball field)

Basketball (playing basketball)

Vacuuming (child vacuuming)