
==== Front
bioRxiv
BIORXIV
bioRxiv
2692-8205
Cold Spring Harbor Laboratory

39071285
10.1101/2024.07.15.603640
preprint
4
Article
A third kind of episodic memory: Context familiarity is a distinct process from item familiarity and recollection
http://orcid.org/0000-0002-3957-0188
Addante Richard J. 123
Clise Evan 1
Waechter Randall 4
Bengson Jesse 5
http://orcid.org/0000-0001-8747-672X
Drane Daniel L. 6
Perez-Caban Jahdiel 1
1 Florida Institute of Technology, Department of Psychology, 150 W. University Dr., Melbourne, FL 32905, USA
2 Florida Institute of Technology, Department of Biomechanical Engineering, Melbourne, FL 32905, USA
3 Neurocog Analytics, LLC, Palm Bay, FL, 32905
4 Windward Islands Research and Education Foundation (WINDREF), Saint George University Medical School, Grenada, West Indies
5 University of California, Davis, Davis, CA, USA
6 Emory University, Atlanta, GA, USA
Contributions: RJA designed the study, supervised the data collection, analyzed the data, created the figures, interpreted the findings, wrote the manuscript, and responded to Reviews. EC and JPC contributed to data analysis. RW, JB, and DLD contributed to interpretation of findings and manuscript preparation.

11 9 2024
2024.07.15.603640https://creativecommons.org/licenses/by-nc-nd/4.0/ This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which allows reusers to copy and distribute the material in any medium or format in unadapted form only, for noncommercial purposes only, and only so long as attribution is given to the creator.
nihpp-2024.07.15.603640.pdf
Episodic memory is accounted for with two processes: ‘familiarity’ when generally recognizing an item and ‘recollection’ when retrieving the full contextual details bound with the item. Paradoxically, people sometimes report contextual information as familiar but without recollecting details, which is not easily accounted for by existing theories. We tested a combination of item recognition confidence and source memory, focusing upon ‘item-only hits with source unknown’ (‘item familiarity’), ‘low-confidence hits with correct source memory’ (‘context familiarity’), and ‘high-confidence hits with correct source memory’ (‘recollection’). Results across multiple within-subjects (trial-wise) and between subjects (individual variability) levels indicated these were behaviorally and physiologically distinct. Behaviorally, a crossover interaction was evident in response times, with context familiarity being slower than each condition during item recognition, but faster during source memory. Electrophysiologically, a Condition x Time x Location triple dissociation was evident in event-related potentials (ERPs), which was then independently replicated. Context familiarity exhibited an independent negative central effect from 800–1200 ms, differentiated from positive ERPs for item-familiarity (400 to 600 ms) and recollection (600 to 900 ms). These three conditions thus reflect mutually exclusive, fundamentally different processes of episodic memory. Context familiarity is a third distinct process of episodic memory.

Graphical Abstract

Recognition
Context
Familiarity
Recollection
Familiarity
==== Body
pmcIntroduction

Sometimes people remember information about the source of a prior event correctly (such as background, location, or other kinds of context) without fully recollecting the details of the event (Addante, Ranganath, & Yonelinas, 2012; Hicks et al., 2002; O’Neill & Diana, 2017). For instance: “this place seems familiar, but I don’t know why or what from” or “I know the general place where I read something [e.g. the bottom left hand part of a page of a book], but I can’t remember what it was [e.g., neither the page nor the specific information]” (Whittlesea & Williams, 2000; Whittlesea & Williams, 2001). It is not clear how this phenomenon occurs in memory. Cognitive models have typically accounted for episodic memory through two processes (for Review, see (Eichenbaum et al., 2007)): ‘familiarity’ recognizes an item from the past but with varying levels of memory strength and lacking contextual details, whereas recollection retrieves the item bound within the contextual details it was experienced with in the past, such that researchers typically used context retrieval as a proxy for recollection. Familiar recognition of a context in the absence of recollection is thus a puzzle for memory theories because it is difficult to resolve the paradox with just the two processes of recollection and familiarity- but these memories happen nonetheless (Addante, Ranganath, & Yonelinas, 2012; Ball et al., 2014; Chen et al., 2018; Cleary et al., 2012; Cleary, K. McNeely-White, et al., 2021; Hicks et al., 2002; Koriat et al., 2003; Kurilla & Westerman, 2010; Lampinen et al., 2005; Ryals et al., 2011; Whittlesea & Williams, 2000).

A prior study investigated this conundrum using event-related potentials (ERPs) to focus on the unique memory condition in which people reported low confidence recognition of an item but had accurate source memory for the item’s context (Addante, Ranganath, & Yonelinas, 2012). It was found that familiar context can be retrieved without relying upon recollection, thereby demonstrating that these constructs are dissociable and not dependent (Figure 1).The ERPs associated with this condition were electrophysiologically distinct from those associated with the memory process of recollection. The condition was therefore named ‘context familiarity’, based upon the framework of the Binding-in-Context (BIC) model (Diana et al., 2007; Montaldi & Mayes, 2010; Yonelinas et al., 2019), which proposes that items and contexts are processed differently in the parahippocampal gyrus of the medial temporal lobe, and that the hippocampus binds the item within the context it was experienced in, e.g. recollection.

The ERP effects for this ‘context familiarity’ manifested generally as a widespread fronto-central-parietal negative-going effect. This occurred at approximately 600 milliseconds (ms) after stimulus presentation and extended later in the epoch from 800 to 1200 ms when compared to both correct rejections as well as when compared to incorrect source judgments, too (Addante, Ranganath, & Yonelinas, 2012). This relatively late, broad central negativity (BCN) was dissociated from the late positive component (LPC) effect that is well-associated with recollection and did not show signs of an FN400 effect often related to familiarity. However, that study did not directly address whether the effects of context familiarity were dissociable from those for item familiarity. Therefore, it has remained unknown whether there is one type of overarching familiarity process that includes representations of both item and contexts (e.g. the BIC model), or, if there might be two separable forms of familiarity: one for items and one for context.

As a brief primer, in human long term memory, theoretical models of conscious episodic recognition have been largely governed by the dual cognitive processes of familiarity and recollection (Atkinson & Juola, 1974; Diana et al., 2007; Eichenbaum et al., 2007; Mandler, 1980; Parks & Yonelinas, 2009; Ranganath, 2010; Ranganath & Ritchey, 2012; Reagh & Ranganath, 2018; Rugg & Yonelinas, 2003; Squire et al., 2007; Wixted, 2007; Wixted & Mickes, 2010; Wixted & Squire, 2011a, 2011b; Yonelinas, 1994; A. P. Yonelinas, 2001; Yonelinas, 2002; Yonelinas et al., 2010; Yonelinas & Parks, 2007; Yonelinas et al., 2019). Recollection is operationalized as the declarative retrieval of episodic information (i.e., an event) specifically with both the item and context of the event bound together into a cohesive retrieval of the episode (for review see (Diana et al., 2007)). Recollection is usually associated with the confident retrieval of contextual information surrounding the item of the event (for reviews see (Eichenbaum et al., 2007; Rugg & Yonelinas, 2003; A. P. Yonelinas, 2001)). The item in the event, however, may be retrieved without recollection, instead relying upon a process of general familiarity of the episodic information, typically conceptualized as declarative, conscious retrieval of an item from a prior episode but without the associated confidence, recollection, or contextual information in which it occurred.

These two memory phenomena are dissociable cognitive processes (Yonelinas, 2002), with differential neural substrates in the medial temporal lobes (Daselaar et al., 2006; Diana et al., 2012, 2013; Ranganath et al., 2004), are neuropsychologically dissociable among patient impairments(Addante, Ranganath, Olichney, et al., 2012; Duzel et al., 2001; Mecklinger et al., 1998), and demonstrate spatio-temporally distinct patterns of electrophysiology at the scalp (event-related potentials, ERPs) (Addante, Ranganath, & Yonelinas, 2012; Curran, 2000; Friedman, 2013; Rugg, Mark, et al., 1998); for Review see (Kwon et al., 2023; Rugg & Curran, 2007). Familiarity has been associated with positive ERP differences in memory trials during a negative-going ERP peak at the mid-frontal scalp sites that occurs at approximately 400 to 600 milliseconds (ms) post stimulus, often called the mid-frontal old-new effect, or FN400 (for frontal-N400 effect). On the other hand, recollection has been associated with positive differences between memory conditions occurring at a peak in the ERP at the parietal region of the scalp from approximately 600 to 900 ms, often referred to as a late parietal component (LPC) or parietal old-new effect (Addante, Ranganath, & Yonelinas, 2012; Leynes et al., 2005) for reviews see (Friedman, 2013; Kwon et al., 2023; Rugg & Curran, 2007).

Among studies of episodic recollection, ‘source memory’ has been one of the most reliable methods of assessing recollection for episodic events in both behavioral and cognitive neuroscience paradigms. That is, it has traditionally been assumed that if someone remembers the source of information, then they must have retrieved information about the context of that memory and hence ‘must’ be recollecting that episode. As such, source memory has served as one of the most useful proxies for understanding memory and amnesia deficits (Addante, Ranganath, Olichney, et al., 2012), despite researchers widely acknowledging that source memory is not process-pure as an exclusive measure of recollection (Bowers & Schacter, 1990; Hicks et al., 2002; Hoffman, 1997; Johnson et al., 1993; Leynes et al., 2016; Mitchell & Johnson, 2009; Parks, 2007; Yonelinas, 1999; Yonelinas & Parks, 2007). Prior research has also identified the role that other processes can contribute to source memory (Ecker et al., 2007), including item-familiarity (Diana et al., 2008; Leynes et al., 2016; Mecklinger, 2006; Parks et al., 2011) and unitization of two items into a new association (Bader et al., 2010; Bastin et al., 2013; Diana et al., 2011; Diana et al., 2008; Parks & Yonelinas, 2015; Tu et al., 2017).

The findings of ERP dissociations for context familiarity and recollection as occurring at different times and different sites on the scalp (Addante, Ranganath, & Yonelinas, 2012) provided direct physiological evidence in support of the BIC model’s theory and related variants (Diana et al., 2007; Eichenbaum et al., 2007; Kwon et al., 2023; Montaldi & Mayes, 2010; Reagh & Ranganath, 2018; Yonelinas et al., 2019); however, there remained limitations. In particular, while those findings had demonstrated that context familiarity was reliably distinct from recollection (Addante, Ranganath, & Yonelinas, 2012), it was unknown if context familiarity is also directly dissociable from item familiarity, or whether it is operating merely as a component element of the same memory process of generic familiarity that governs both items and contextual representations (Diana et al., 2007). The notion that context can be differentially familiar than item information has been suggested in prior studies (Addante, 2012; Whittlesea & Williams, 2000; Whittlesea & Williams, 2001), but was viewed only as context familiarity being merely a variant of the same underlying process of familiarity, and lacked evidence of an independent process until the aforementioned ERP study (Addante, Ranganath, & Yonelinas, 2012). Here, we put forth evidence for a fundamentally different view: that familiarity of context differs qualitatively from item familiarity.

The focus of the present investigation was to assess if the ERP effects for context familiarity can be reliably dissociated from those for item familiarity, and to add novel insights from behavioral measures of response times. When prior investigations observed dissociable ERP activity in time and topography for conditions of recollection (LPC) and familiarity (FN400), it was considered compelling evidence of two independent memory processes (Curran, 2000; Rugg & Curran, 2007; Rugg, Mark, et al., 1998), that reflect dual-process models of episodic recognition, since a single process would not have (and has not) been able to account for such patterns of neurophysiology without substantial post-hoc modifications (Eichenbaum et al., 2007; Kwon et al., 2023). In turn, if the effects of context familiarity are found to be dissociable in ERPs and/or behavior, it could be taken as evidence that context familiarity is a distinct additional process of episodic memory that extends beyond the traditional dual processes of recollection and item-familiarity.

Towards that end, we re-analyzed a published ERP dataset (N = 54) (Addante et al., 2023; Muller et al., 2021) that utilized the same paradigm as prior studies (Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012) and tested a combination of item recognition confidence and source memory judgments. We directly assessed behavioral measures of response times and physiological measures of ERPs for the three conditions of ‘item-only hits with source unknown’ (‘item familiarity’ (Addante et al., 2023; Addante et al., 2011)), ‘low-confidence recognition hits that had correct source memory’ (‘context familiarity’(Addante, Ranganath, & Yonelinas, 2012)), and ‘high-confidence hits that had correct source memory’ (‘recollection’ (Addante, Ranganath, & Yonelinas, 2012)) (Figure 1). To preface our findings, we found that context familiarity, item familiarity, and recollection were each dissociable in neural activity patterns occurring at different times and places on the scalp, as well as in behavioral measures of response times for both item and source memory judgments (Experiment 1), and that the electrophysiological findings replicated across several independent studies (Experiment 2).

2.0 Methods

2.1 Participants.

The current study consisted of a re-analysis of previously published ERP data (Addante et al., 2023; Muller et al., 2021), so the sample size was determined by the number of participants’ data that was available in the archival dataset; the study was not pre-registered. The participants consisted of 61 right-handed students free from neurological disorder and memory problems. Data was not used for four participants due to noncompliance issues (i.e., pressed only one button throughout the task or ignored experimenter’s instructions), data lost due to experimenter error (N=1) or EEG data deemed unusable (N=2) due to excess motion artifacts/noise that resulted in an exclusion of the majority of EEG trials. This presented a working data set of N = 54 for the current study, which was more than double the size of the original study reporting physiological correlates of context familiarity (N = 25) (Addante, Ranganath, & Yonelinas, 2012), and triple the sample size standards of many of the foundational ERP studies of recognition memory (Duarte et al., 2004; Duzel et al., 2001; Duzel et al., 1997; Paller & Kutas, 1992; Rugg, Mark, et al., 1998; Rugg, Walla, et al., 1998; Wilding et al., 1995; Wilding & Rugg, 1996; Woodruff et al., 2006).

As described by the data’s original publication (Muller et al., 2021): Participants were recruited through a combination of methods including advertisements placed around CSUSB or through the schoolwide research pool SONA. Participants recruited through advertisements were paid $10 an hour for sessions that lasted approximately 2 hr. The majority of participants were women (N = 48); 57% were Hispanic, 23% Caucasian, 11% Asian, and 10% identified as more than one ethnicity. The average age was 23.5 years old (SD = 4.82). None of the participants reported any visual, medical, or physical issues that would interfere with the experiment. Most participants spoke English as their first language (N = 47) and those who had indicated speaking a different first-language had been speaking English for an average of 16.73 years (SD = 4.74, SE = 1.22, Median = 17.0, Minimum = 7, Maximum = 25). Written informed consent was obtained for participation in the experiment and protocol approved via the Institutional Review Board of California State University – San Bernardino, and data analyses observed the privacy rights of human subjects, consistent with the Declaration of Helsinki and as approved by the Institutional Review Board of Florida Institute of Technology.

2.2 Procedure.

The paradigm used was the same item- and source-memory confidence paradigm that has been routinely used in our prior reports to characterize various facets of memory (Addante, 2015; Addante et al., 2023; Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012; Addante et al., 2011; Muller et al., 2021; Roberts et al., 2018), with slight modifications as noted below (Muller et al., 2021). This paradigm consisted of an encoding phase containing four sequential study sessions during which participants studied 54 words per session, followed by a retrieval phase that contained six test sessions in which the participant’s memory was tested for 54 words in each session. Item recognition was tested on a five-point confidence scale with ‘1’ indicating being sure the item was new, and ‘5’ indicating being sure the item was ‘old’ (Figure 1).

During the encoding phase, participants were given instructions to make a simple decision about the word presented. The participants were either asked to judge if the item was man-made or if the item was alive. The instructions were presented in one of two counterbalanced orders: ABBA or BAAB. The participants viewed four lists of 54 words during the encoding phase. After the encoding phase was complete, the EEG cap was applied to the scalp of participants. During the memory test, participants viewed a total of 324 words, 216 of which were presented in the encoding phase and 108 of which were unstudied items (new items). All words presented during study and test were presented in white font on black background screen. Word stimuli were selected from the same source as stimuli used in the original study in 2012 and preserved the same characteristics. Word stimuli were selected from the Medical Research Council Psycholinguistics Database (http://www.psych.rl.ac.uk/MRC Psych Db.html). Word stimuli were all nouns, had an average rating of concreteness of 589 (min=400, max=670), imageability of 580 (min=424, max=667), Kucera–Francis Frequency of 30 (min=3, max=198), and an average number of 4.9 letters in each word (min=3, max=8).

During the memory retrieval phase, the participants were read instructions asking them to judge if the stimulus word presented was old (studied during the encoding phase) or new (not studied before in the encoding phase) (Figure 1). To begin a trial, a screen with a small white cross at the center was presented for one of three randomly chosen times: 1 second, 2.5 seconds, or 3 seconds. Then the participants were presented with a word in the middle of the screen, the numbers 1, 2, 3, 4, and 5 evenly spaced beneath the word, the word “New” on the left by the number 1, and the word “Old” on the right under the number 5. Participants pressed any number between 1 and 5 to indicate if they confidently believed the word was old (5), believe the word was old but was not confident (4), did not know if the word was old or new (3), believe the word was new but was not confident (2), or confidently believed the word was new (1). This prompt was subject-paced, and participants were told to choose the response that gave the most accurate reflection of their memory, and to respond as quickly and accurately as possible.

Immediately after each decision on item memory confidence, participants were asked to answer a source memory question about if the word came from the animacy decision task or the man-made decision task. The word and numbers remained on the screen but this time, word “Alive” was presented on the left by the number 1, and the word “Manmade” was presented on the right under the number 5. Participants were told to choose the response that gave us the most accurate reflection of their memory and could respond that they confidently believed the word was from the animacy task (1), believe the word was from the animacy task but was not confident (2), did not know the source of the word or had replied in the question directly before that the word was new (3), believe the word was from the manmade task but was not confident (4), or confidently believed the word was from the manmade task (5). Trials were terminated by subject response. Each session consisted of a list of 54 words. Six lists of 54 words were presented during the retrieval phase. The retrieval phase of the paradigm also included a simple metacognitive question to participants for estimating their performance, which occurred once every ten trials and for which the data has been previously reported elsewhere for metacognition effects (Muller et al., 2021). The current investigation focused instead upon the memory related responses, as the metacognitive response data has been reported previously. Each session consisted of a list of 54 words; overall, six lists of 54 words were presented during the retrieval phase.

2.3.1 Conditions of interest:

Following common convention in the field, we operationalized these processes based on the combination of responses given for item recognition confidence and source-memory judgments, consistent with the way prior research has defined these conditions. For recollection, we followed the convention of prior findings that it represented instances of retrieving high-confidence item recognition followed by correct source information: item responses of ‘5’ accompanied by source judgments that were correct (Figure 1; (Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012; Muller et al., 2021). For context familiarity, we followed the same approach used to originally identify this process (Addante, Ranganath, & Yonelinas, 2012): instances of low-confidence item recognition accompanied by correct source judgments (Figure 1). Item familiarity was defined based upon previous work (Addante et al., 2023; Addante et al., 2011) as instances in which participants retrieved successful item recognition (hits, combining 4 and 5 responses) but reported having no source memory information (source unknown, 3 responses) (Figure 1). These conditions of item- and context- familiarity are ones which have never been directly compared against each other in the literature while using behavioral or ERP measures, and so it remains to be determined if and how they may differ, and the extent to which the process of familiarity may be organized among these variables as a continuous or distinct process(es).

In order to maintain sufficient signal-to-noise ratio (SNR), all pairwise comparisons relied upon including only those subjects who met a pre-determined criterion of having a minimum number of 12 artifact-free ERP trials per condition being contrasted (Boudewyn et al., 2018; Gruber & Otten, 2010; Jensen & MacDonald, 2023; Otten et al., 2006; Ozubko et al., 2021). As was noted in the Introduction, the present work is not claiming any specifically process-pure ERP measurements of memory, but instead we adopt what others (Curran, 2000; Kwon et al., 2023; Mecklinger, 2006; Rugg & Curran, 2007) as well as our own previously published positions (Addante, 2015; Addante et al., 2023; Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012; Leynes et al., 2005; Muller et al., 2021; Ozubko et al., 2021) described as the ‘putative neural correlates’ of recognition memory processes.

2.3.2 Measuring source memory performance.

The current study’s interest focused specifically upon combinations of item + source memory responses described below. Like most studies of source memory, we wanted to know how accurate people were in discriminating between the two sources of information in the experiment, and so excluded the source unknown responses from the performance calculation of source discrimination. This procedure collapses high- and low- confidence source judgments into general ‘correct’ and ‘incorrect’ conditions. We thus calculated this source memory performance in the same way as had been done in prior studies (Addante, Ranganath, & Yonelinas, 2012): the total number of source correct trials divided by the total number of all correct & incorrect source memory judgments (excluding the source unknown responses) [ # correct source trials / (sum of # of correct + # of incorrect source trials)]. The source memory calculation we used is akin to the standard recognition measure of memory performance (%Hits-%FA) but applied to source memory discrimination, which would require excluding the ‘source unknown’ responses, since a false alarm in source memory would be a source misattribution (not the lack of source memory). When source memory discriminations occur between available options of the two sources for participants to choose from, they have a 50% chance or being right/wrong (source correct or source incorrect responses), thus chance levels are at 50% in that analysis (as opposed to 33% if assessing it relative to the three response options of correct, incorrect and unknown; see discussion below for further considerations of the limitations of such alternative calculation methods1).

In excluding the ‘source unknown’ responses from the calculation of performance on source memory discrimination, it excludes the trials of when people had an absence of source memory (Coane & Umanath, 2019; Liu et al., 2007). This approach was taken because we did not want to know how frequently people remembered correct source information at all (i.e.: out of all of the possible times that they could have), because this number would be artificially low, but rather how well they discriminated sources that they did remember. For example, consider a scenario in which people were tested on if episodic memory for where we learned various items of information in our lives. Then consider if the performance measures for memory measured the few times that we did remember correctly amid all the times when we couldn’t remember any source at all (likely most of the time, for naturalistic settings). Most people would [falsely] appear to be dense amnesiacs unable to retrieve source information above chance- despite likely having reasonable source discrimination capabilities for the times that they could retrieve source memory information. The same is true for clinical applications of the same principle: consider cases studying reality monitoring of auditory hallucinations in schizophrenia patients, whereby the pertinent question is about understanding how well the source determination is made when a source of the information is available to the patient- not the other instances in which it isn’t.

Our approach to measuring source memory performance is a common one that is sensibly used in other source memory studies (Diana et al., 2008). For example, our standard approach used to measure source memory performance is not just the same as we have published with previously (Addante, Ranganath, & Yonelinas, 2012) but is also the same approach used by additional other studies measuring source memory performance (Leynes & Mok, 2017), where they note that: “Conditional source measures control for overall differences in recognition by calculating the probability that the correct source is identified given that the item was correctly identified as old (p|source| = source hits/source hits + source misattributions)”. For these reasons noted above, we utilized the present method of calculation for source memory discrimination that excluded the ‘source unknown’ responses, as it provides the most appropriate treatment of participant’s memory while minimizing the extent of possible confounds. This approach also avoided the potential confounds of not knowing what response criterion a subject is adopting when making the ‘unknown’ response for source memory, as no two participants could be presumed to use the same such criterion. Importantly, it also ensured that the pursuit of replication could be achieved by using the same measures as were used in the preceding work we sought to asses for reproducibility (Ioannidis et al., 2012; Nosek et al., 2022).

2.4 Electrophysiological Acquisition.

Each subject was tested individually inside a private chamber. Stimulus presentation and behavioral response monitoring were controlled using Presentation software on a Windows PC. EEG was recorded using the actiCHamp EEG Recording System with a 32-channel electrode cap conforming to the standard International 10–20 System of electrode locations and was acquired at a rate of 1024 Hz. The EEG cap was sized while the participant’s face was wiped free of skin oil and/or makeup in preparation for attaching ocular electrodes. Five ocular electrodes were applied to the face to record electrooculogram (EOG): two above and below the left eye in line with the pupil to record electrical activity from vertical eye movements, two on each temple to record electrical activity from horizontal eye movements, and one electrode in the middle of the forehead in line horizontally with the electrode above the left eye as the ground electrode. EOG was monitored in the horizontal and vertical directions, and this data was used to eliminate trials contaminated by blinks, eye-movements, or other related artifacts. The EEG cap was placed on the participant’s head and prepared for electrical recording. Gel was applied to each cap site and impedances were lowered below 15 KOhms via gentle abrasion to allow the electrodes to obtain a clear electrical signal. Subjects were instructed to minimize jaw and muscle tension, eye movements, and blinking.

2.5 Electrophysiological Analysis.

Physiological measurements of brain activity were recorded using EEG equipment from Brain Vision LLC. All EEG data was processed using the EEGLAB and ERPLAB toolboxes in MATLAB (Delorme & Makeig, 2004; Lopez-Calderon & Luck, 2014). The EEG data was first re-referenced to the average of the mastoid electrodes, passed through a high-pass filter at 0.1 hertz as a linear de-trend of drift components, and then down sampled to 256 hertz (Hz). The EEG data was epoched from 200 milliseconds prior to the onset of the item recognition stimulus to 1200 milliseconds after the stimulus was presented(Muller et al., 2021). Independent components analysis (ICA) was performed using Infomax techniques in EEGLab (Bell & Sejnowski, 1995) for artifact correction and the resulting data was individually inspected for artifacts, rejecting trials for eye blinks and other aberrant electrode activity. During ERP averaging, trials exceeding ERP amplitudes of +/− 250 mV were excluded, as described in the previous report of this data (Addante et al., 2023; Muller et al., 2021). A 30 Hz low pass filter was applied to each subject’s ERPs as a non-causal, infinite impulse response (IIR, Butterworth) filter, implemented through ERPLAB toolbox (Lopez-Calderon & Luck, 2014).

2.6.1 Statistics.

Our hypotheses focused upon identifying three different processes of episodic memory: recollection, item familiarity, and context familiarity. We hypothesized that recollection would exhibit both an FN400 for item-familiarity and an LPC for recollection but would not exhibit a late broad central negativity (BCN) effect, whereas context familiarity would exhibit only an BCN and neither an FN400 or LPC, and that item-familiarity would be associated with only an FN400 effect but not an LPC nor BCN effect. As the current investigation was based upon clear a priori-defined hypotheses derived from prior findings (Addante, Ranganath, & Yonelinas, 2012), analyses utilized planned paired t-tests to assess differences between the targeted conditions (two-tailed unless indicated otherwise) with an alpha level of .05. In cases where exploratory analyses were conducted to explore unplanned comparisons, ANOVA was used to qualify potential differences that may exist and were corrected with Geisser-Greenhouse corrections when necessary. Analyses conducted to assess the extent to which prior results could be found to replicate in independent data sets and/or were principally based upon a priori predictions that effects would differ in one way (i.e.: context familiarity exhibiting more negative-going ERP amplitudes than correct rejections, item-familiarity (Addante, Ranganath, & Yonelinas, 2012)); such analyses thus applied targeted directional hypotheses utilizing one-tailed t-tests where noted in order to maintain a properly principled statistical approach, though most were also still significant with non-directional two-tailed tests as well (i.e.: multiply the p value by 2).

In reporting results of behavioral response times, we followed convention in the field to report the mean values of reaction times; however, because median values can provide a more stable measure of central tendency that is less prone to the influence of extreme values, we also assessed those values as well. When outlier data points were removed from the mean values (N = 2, using criteria of exceeding 3 standard deviations from the mean) the results from the analysis of mean response times remained the same. In this case, the two outlier subjects had longer mean RT values, merely indicating that some subjects sometimes responded a few seconds slower on uncertain memory conditions, which is common knowledge of reasonable variability in cognitive capabilities during conditions of uncertainty. Also, these inclusive data points are the aggregate of multiple responses for a person instead of a singular outlying value that might otherwise be considered as potential artifacts. Removing outliers based upon their post-hoc values would thus be inappropriately biasing the data based upon post hoc visual inspection of it. Therefore, we report mean reaction times with outlier data points included, since they represent the real average response times of subjects.

Because a non-significant p-value in conventional frequentist t-test methods is unable to inform whether there is actually evidence for the null hypothesis or if there is merely not sufficient evidence for any conclusion at all, analyses that revealed a null finding were subsequently quantified using Bayes factor analysis. Bayes factor analysis is a tool that allows researchers to quantify the relative strength of evidence for the null hypothesis (invariance) (Aczel et al., 2020; Bayes, 1991; Bengson & Luck, 2015; Etz et al., 2018; Etz & Vandekerckhove, 2018; Jarosz & Wiley, 2014; Jeffreys, 1998; Morey & Rouder, 2011; Rouder et al., 2009; Schönbrodt & Wagenmakers, 2018; Wagenmakers et al., 2018). A resulting Bayes factor value can generally be considered as the ratio of how likely a hypothesis is as compared to the likelihood to another hypothesis (for specific details see (Jarosz & Wiley, 2014; Rouder et al., 2009)). For interpreting Bayes factor result values, it is conventionally viewed that a Bayes factor (BF01, evidence for the null hypothesis) of 0–1 typically represents there being no evidence available for the null hypothesis, 1–3 typically represents anecdotal evidence for the null hypothesis, a BF of 3–10 represents relatively moderate evidence for the null hypothesis, and a BF of 10–30 represents relatively strong evidence for the null hypothesis (Jarosz & Wiley, 2014; Jeffreys, 1998; Schönbrodt & Wagenmakers, 2018). Thus, Bayes Factor results are reported as such for the pertinent analyses as representing the relative strength of evidence for the null finding as noted above. The Bayes factor analyses were computed for paired or one-sample t-test designs. Results were calculated using the standard scale of r = 0.707 and the resulting outputs are provided as values in favor of the null hypothesis using the recommended Jeffrey-Zellner-Siow Prior (JZS, Cauchy distribution on effect size) (Rouder et al., 2009).

2.6.2 Reproducibility.

The reproducibility of the effects of context familiarity was assessed in several ways through the present investigation. First, the present investigation (Experiment 1) was an assessment of the extent to which previously-reported effects (Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012; Addante et al., 2011) of context familiarity originally differentiated from the neural correlates of recollection in a sample of N = 25 could be found to be reproducible amid a larger sample (N = 54) that had been previously published in the domain of memory metacognition (Addante et al., 2023; Muller et al., 2021). Those original effects replicated in entirety in the present data. The sample of data of the present study thus demonstrated sensitivity of effect sizes and statistical power to detect the ERP effects of context familiarity originally discovered by the original study(Addante, Ranganath, & Yonelinas, 2012) by virtue of it having reproduced those findings of differences between context familiarity and recollection in a sample (N = 54) (Muller et al., 2021) that was more than double the original sample size (N = 25) (Addante, Ranganath, & Yonelinas, 2012).

Second, when the present findings (Experiment 1) then discovered that the neural correlates of context familiarity were dissociable from those for item familiarity, we sought to assess the extent to which these findings would have been observed in the original study (Addante, Ranganath, & Yonelinas, 2012) if those authors had only thought to look (they hadn’t). These effects of the present investigation were also found to be replicated in entirety (Experiment 2), both within-subjects and between-subjects, and across the same control analyses. Third, we also sought the extent of reproducibility of the findings across a range of several control analyses conducted in both Experiments 1 and 2, which systematically varied different conditions to assess potential confounds and possible alternative accounts. The control analyses found that the results remained reproducible even when reducing signal to noise ratio of trials and when reducing statistical power limited to only low levels of familiarity strength (excluding trials of strong familiarity responses).

Fourth, as the present study was a secondary analysis of existing independent datasets that had already been published with sufficient power and sample size to detect differences in previously published papers (Addante et al., 2023; Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012; Muller et al., 2021), it was not possible nor necessary (Heinsberg & Weeks, 2022; Hoenig & Heisey, 2001; Lakens, 2022; Levine & Ensom, 2001; Zhang et al., 2019) to perform an a priori power analyses for the present investigation since the sample size was already immutably established before the secondary investigation of analyses began2. Nevertheless, post-hoc analyses of power and sensitivity were performed because a previous version of this manuscript had been editorially forced to provide a post-hoc power & sensitivity analysis3 at Nature Communications: Psychology (which was then properly critiqued by another Reviewer for having done the ill-suited post-hoc analyses). The post-hoc results confirmed the sufficient power and sensitivity to detect all primary findings in the present report, and are available in the original open-access version of the manuscript for those interested (Method section, Supplemental Table 1) (Addante et al., 2024). The present version of the manuscript specifically excludes those results because of widespread agreement in the literature4 that such “post-hoc analyses are not informative” (Heinsberg & Weeks, 2022), are “fundamentally flawed” (Hoenig & Heisey, 2001), are “conceptually flawed and analytically misleading…and do not provide sensible results” (Zhang et al., 2019), and “should not be used” (Heinsberg & Weeks, 2022) since they are “simply not informative for data interpretation” (Heinsberg & Weeks, 2022). As such, we echo the conclusions recommended by others that: “we encourage authors and peer-reviewers to avoid using or requesting post hoc power calculations” (Heinsberg & Weeks, 2022). The present results were fully replicated in independent data, and in twice the sample size of previously published results that the present investigation also replicated; thus, the sufficiency for statistical power in the study is self-evident.

2.7.1 Experiment 2: Participants and Procedure

The participants included in Experiment 2 were aggregated across three previously published studies (Addante, 2015; Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012; Addante et al., 2011) that used nearly identical experimental protocols of item recognition confidence followed by source memory judgments as were used in the present Experiment 1 (Addante et al., 2023; Muller, 2019), for a total sample of N = 56. These studies had only directly contrasted conditions of recollection and context familiarity, but had never directly assessed if the ERPs for item- and context-familiarity differed(Addante, 2012). Each study’s cohort of participants have been previously reported for having intact levels of familiarity-based recognition on both behavioral and electrophysiological measures that did not significantly differ as a function of younger/middle aged adults (Addante, 2012, 2015; Addante, Ranganath, Olichney, et al., 2012). They have therefore been previously treated as equivalent cohorts of familiarity-based recognition, which was the subject of the present investigation (Addante, 2015; Addante, Ranganath, Olichney, et al., 2012), and thus we also treated them accordingly here as well.

The participants were reported in prior studies (Addante, Ranganath, & Yonelinas, 2012) as consisting of twenty-five healthy right-handed undergraduate students (Mean age = 20.4 years [SD = 2.9, SE = .60, Min = 18, Max = 30], mean education level = 14.4 years [SD = 1.8, SE = .4, Min = 12, Max = 19] demographics of 52% Caucasian, 20% Hispanic, 20% Asian, 8% more than once race; seventeen female) and twenty-two (Addante, Ranganath, & Yonelinas, 2012; Addante et al., 2011) right-handed undergraduate students (Mean age = 20.9 years [SD = 3.1, SE = .70, Min = 18, Max = 29], Mean education 14.5 years [SD = 1.4, SE = .30, Min = 13, Max = 17], demographics of 55% Caucasian, 9% Hispanic, 5% African-American, 32% Asian; ten females), all recruited from the University of California–Davis Psychology Department subject pool. They received credit for participation and were free from neurological, visual, motor, or other medical disorders. An additional nine middle-aged participants (six female; Mean = 45.4 years, SD = 12.5, SE = 4.2, Min = 27, Max = 57) were included whom were tested on the same memory protocol (six male) (Addante, 2015; Addante, Ranganath, Olichney, et al., 2012) and had a comparable profile of education (M = 16.8 years, SD = 5.7, SE = 1.9, Min = 12, Max = 31) and majority demographics (78% Caucasian, 22% more than once race). The experiments had been conducted as approved by the University of California–Davis Institutional Review Board protocol for research on human subjects.

2.7.2 Experiment 2: Measures.

These analyses were conducted using the same sites and latencies that had been reported as representative of effects in the present and previous studies (mid-central Cz; bilateral frontal sites (Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012) F3, F4). The null hypothesis for these analyses is that ERPs for the two conditions of context- and item-familiarity would exhibit no differences, if in fact the two memory responses were being supported by the same kind of underlying memory processing, and/or if the main results of Experiment were representing a form of type I error. The alternate hypothesis was that context familiarity would exhibit more negative-going ERPs than item-familiarity, and that if there is any such difference in ERP activity for the two conditions then it would be indicative that the two memory response types could not be measuring the same kind of memory process, and thus reflecting different kinds of memory processing people were using to making these different memory judgments (Duzel et al., 2001; Eichenbaum et al., 2007; Kwon et al., 2023; Rugg & Curran, 2007; Rugg, Mark, et al., 1998; Yonelinas, 2002).

3.0 Results: Experiment 1

3.1 Behavioral performance:

In the current study, when people had low-confidence recognition (item 4 judgments) they went on to provide accurate source discrimination judgments significantly above chance (considered as .50, see Method) (M = .572, SD = .18, SE = .03; t(47) = 2.85, one-tailed p = .003; Cohen’s d = .40, 95% CI [.521, .623]) (Table 1), which was (and replicated) the same level of source memory performance as was previously reported (Addante, Ranganath, & Yonelinas, 2012) (M = .576, SD = .15, SE = .03, 95% CI [.517, .635]). This null difference was quantified by finding no significant differences between the present study and the preceding one (Addante, Ranganath, & Yonelinas, 2012) (t(71) = .081, p = .936, 95% CI [−.080, .087], Bayes Factor = 3.95). Having established the reproducibility of the prior behavioral performance of memory responses(Addante, Ranganath, & Yonelinas, 2012), we next explored the response speeds and electrophysiological correlates of this paradoxical combination of memory response patterns.

3.2 Response Times: Is context familiarity behaviorally dissociable from recollection and item-familiarity?

We analyzed only subjects who had provided pairwise responses in each of the six conditions of interest (N=38). In our paradigm, participants first made a recognition judgment immediately followed by a source memory judgment for each item (Figure 1); we separately assessed the response times for the item and source judgments (Table 2).

First, in item recognition judgments, reaction times were assessed across the factor of the three memory conditions (recollection, context familiarity, and item familiarity), using the same low confidence source memory responses in recollection and context familiarity conditions as was done in prior studies (Addante, Ranganath, & Yonelinas, 2012), since it held constant the variable of source memory confidence. Response times for the three memory conditions were subjected to a one-way repeated measures ANOVA with three levels, which revealed a significant main effect of condition (F(2,37) = 25.16, p < .001, n2p =.41). Post hoc pairwise comparisons using the Bonferroni correction revealed that responses significantly differed for item familiarity and context familiarity, such that context familiarity was responded to reliably slower than item familiarity (MD = 348.57 msec, p = .048, 95% CI [2.26, 694.88], Cohen’s d = .48, as well as recollection (MD = 900.15 msec, p < .001, 95% CI [547, 1253], Cohen’s d = .76) (Figure 2). Item familiarity responses were also slower than recollection (MD = 551.58 msec, p < .001, 95% CI [297.97, 805.19], Cohen’s d = 1.24).

Within source memory judgments, response times for the three memory conditions revealed a significant main effect of condition (F(1.51,55.81) = 8.28, p = .002, n2p = .18). Further investigation using post hoc tests with Bonferroni correction revealed that context familiarity responses were faster than responses for both recollection and item familiarity (MD = 521.06, p < .001, 95% CI [ 273.01, 769.10], Cohen’s d = .42; MD = 537.21, p = .012, 95% CI [100.20, 974.23], Cohen’s d = .43). (Figure 2). Response times during source memory for recollection and item familiarity conditions did not differ (MD = 16.16, p = 1.00, 95% CI [−399.64, 431.96], Bayes Factor = 5.70).

Because item- and context-familiarity conditions also differed in item confidence (item familiarity included high and low confidence responses, whereas context familiarity included only low confidence recognition responses [Method, Figure 1), it was possible that differences between item- and context-familiarity conditions could have been attributable to the high confidence recognition responses included in the item familiarity condition. To test this directional hypothesis, we conducted a targeted analysis of these conditions while using only low confidence recognition with source unknown (item familiarityLow, no source memory responses), and low confidence recognition with source correct (context familiarityLow, all correct source memory responses). While this excluded high-strength item-familiarity responses (high confidence recognition), such that a null finding would not be able to falsify the hypothesis, it nevertheless represented a valuable challenge of the alternative hypothesis using only weaker-strength item familiarity matched for item-familiarity strength (Addante et al., 2023; Diana et al., 2010; Wixted & Mickes, 2010; Yonelinas, 2002; Yonelinas et al., 2010). Since these control analyses’ conditions were specifically defined by removing response types and thereby inherently weakening its power to detect differences, they nevertheless still produced significant results, and they also went on to be mirrored by similar results in the physiological measurements

The reliable difference originally observed between item- and context-familiarity responses was still preserved (t(36) = 2.09, p = .022, one-tailed, Cohen’s d = .34 95% CI [−inf, −49.49]; cFamLow: M = 2703, SD = 1033, SE = 149; iFamLow: M = 2896, SD = 852, SE = 140), and maintained a comparable pattern for source memory responses as were seen in the main analysis above (t(36) = 1.94, p = .030, one-tailed, Cohen’s d = .32, 95% CI [−inf, 57.76]; iFamLow: M = 2418, SD = 1697, SE = 279). These findings ensured that differences between familiarity responses (context, item) were not confounded by inclusion of higher confidence levels of memory strength in the item familiarity condition.

3.3 Electrophysiological Results

The EEG analyses began with an assessment of general memory ERPs (hits vs correct rejections) to identify if, consistent with the behavioral findings, there were three different physiological patterns of activation supporting episodic memory retrieval. This adopted the same approach as originally taken to characterize the conventional dual processes of recollection and familiarity (Allan et al., 1998; Rugg, Mark, et al., 1998) but was applied here to three different locations and latencies selected based upon hypotheses from prior findings (Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012). This was then followed by a series of more specific analyses of the particular memory condition contrasts.

For the first analysis, we used a basic contrast of measuring ERPs for hits vs. correct rejections in every subject (N=54) using a priori predictions from the same sites as the previous reports (Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012): mid-central site Cz from 400–600 ms for the FN400, left parietal site P3 from 600–900 ms for the LPC, and fronto-central site Fc1 from 800–1200 ms for the BCN. These ERPs were subjected to a 2×3×3 repeated measures ANOVA with factors being memory, time, and site. Results revealed main effects of site (F(1.20, 53) = 46.71, p < .001, n2p = .47) and time (F(1.6, 53) =13.271, p <.001, n2p =.20) and, importantly, a significant 3-way interaction of memory x site x time (F(2.29, 53) = 20.018, p < .001, n2p =.27) [in addition to interactions of memory x site (F(1.57, 53) = 13.60, p < .001, n2p = .20), memory x time (F(1.61, 53) = 22.89, p < .001, n2p =.30) and site x time (F(1.75, 53) = 14.58, p < .001, n2p = .22) (Figure 3).

Consistent with our a priori directional hypotheses, hits were more positive than correct rejections at Cz from 400–600 ms (i.e. an FN400 effect; t(53) = 3.86, p < .001, one-tailed, Cohen’s d = .53, 95% CI [.47, inf]; Hit: M = −3.66, SD = 3.51, SE = .48; CR: M = −4.50, SD = 3.72, SE = .51) and at P3 from 600–900 ms (an LPC effect; t(53) = 3.10, p = .002, one=tailed, Cohen’s d = .42, 95% CI [.38, inf], Hit: M = 3.46, SD = 2.54, SE = .346; CR: M = −.430, SD = 2.28, SE = .310), but were more negative at fronto-central site Fc1 from 800–1200 ms (t(53) = 2.84, p = 003, one-tailed, Cohen’s d = .39, 95% CI -[inf, −.348]; Hit: M = −3.86, SD = 3.31, SE = .451; CR: M = −3.01, SD = 3.28, SE = .45) with comparable results also evident from 800–1200 ms at adjacent site Cz, t(53) = 2.50, p = .007, one-tailed, Cohen’s d = .34, 95% CI [−inf, −.242]; Hit: M = −2.27, SD = 3.08, SE = .42; CR: M =−1.54, SD = 3.11, SE = .423). These findings provide strong support for three dissociable neural signals underlying basic episodic recognition, operating in different times and places on the scalp (Mecklinger, 2006; Mecklinger & Jäger, 2009; Rugg, 2009; Rugg & Coles, 1995; Rugg & Curran, 2007; Rugg, Mark, et al., 1998; Shallice, 1988). A series of subsequent analyses were then carried out to systematically characterize each condition (context familiarity, recollection, and item-familiarity) as compared to correct rejections (Allan et al., 1998; Kwon et al., 2023; Rugg, Mark, et al., 1998), described below.

Experiment 1: Verifying if recollection and context familiarity exhibit different neural correlates.

Since the current project hypothesized that context familiarity is distinct from recollection, we first sought to establish that the dataset validly reproduced prior findings (Addante, Ranganath, & Yonelinas, 2012) before endeavoring to explore differences within familiarity of items and context. A sample of N = 27 subjects provided observations in both conditions of interest for this analysis and satisfied inclusion criteria for numbers of valid ERP trials (see Method). A 2×3×2 repeated measures ANOVA for factors of memory condition (recollection, context familiarity), time interval (400–600ms, 600–900ms, and 800 1200 ms), and site (P3, Cz) was performed based upon a priori predictions derived from the preceding analyses and prior reports dissociating context familiarity from recollection-based processing (Addante, Ranganath, & Yonelinas, 2012). Results revealed a significant main effect of memory condition (F(2,26) = 9.418, p = .005, n2p = .27), a main effect of time interval (F(1.138, 26) = 6.262, p = .015, n2p = .19), and a main effect of site (F(1,26) = 15.89, p < .001, n2p = .38), along with a significant time by site interaction (F(1.35, 26) = 18.81, p < .001, n2p =.42) and a memory x time x site interaction (F(1.5, 26 = 7.69, p = .002, n2p = .23) (Figure 4).

ERPs for context familiarity compared to the conventional ERP control condition of correct rejections revealed reliable evidence of the same negative-going effect from 800–1200 ms that had been directionally hypothesized from previous reports(Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012) (Figure 4), while also reproducing the prior findings of there being no evidence for ERP correlates of recollection or familiarity (LPC, FN400) (Addante, Ranganath, & Yonelinas, 2012). ERPs for context familiarity were more negative-going than correct rejections (CR) from 800–1200 ms at mid central site Cz (t(26) = −2.53, p = .009, one-tailed; Cohen’s d = .49, 95% CI [−inf, −.369]; cFam: M = −2.91, SD = 3.575, SE = .69; CR: M = −1.772, SD = 3.72, SE = .72). Alternatively, there was no LPC effect evident for context familiarity from 600–900 ms (t(26) = 0.161, p = .437, one-tailed; Cohen’s d = .03, 95% CI [−.732, inf, Bayes factor = 4.32]; cFam.: M = −.857, SD = 3.00, SE = .58; CR: M = −.932, SD = 2.15, SE = .413), nor any evidence for an FN400 effect for context familiarity from 400–600ms (t(26) = −0.056, p = .522, one-tailed, Cohen’s d = .01, 95% CI [−.753, inf], Bayes factor = 5.12; cFam: M = −5.241, SD = 3.83, SE = .737; CR: M = −5.217, SD = 4.03, SE = .78).

ERPs for the recollection condition were also found to be consistent with the prior findings (Addante, Ranganath, & Yonelinas, 2012) when compared to correct rejections (Figure 4): eliciting both an FN400 effect at Cz from 400–600ms, (t(26) = 2.82, p = .004, one-tailed, Cohen’s d = .54, 95% CI [.487, inf]; rec: M = −3.939, SD = 3.54, SE = .68; CR: M = −5.22, SD = 4.03, SE = .78) and an LPC at left parietal site of P3 from 600–900ms ( t(26) = 2.98, p = .003, one-tailed, Cohen’s d = .57, 95% CI [.665, inf]; rec: M = .627, SD = 2.94, SD = .57; CR: M = −.932, SD = 2.15, SE = .41) but no evident differences during the later time of 800 to 1200 ms (t(26) = .002, p = .499, one-tailed; Cohen’s d= .00, 95% CI = [−.923, inf], Bayes Factor of 4.91). These results indicated that the neurophysiology for conditions of recollection and context familiarity were indeed distinct in time and place on the scalp. Overall, these results thus reproduced the prior findings that neural processing for the retrieval of familiar contexts was distinct from recollection (Addante, Ranganath, & Yonelinas, 2012), and established the foundation for exploring if the context familiarity may also be directly dissociable from item familiarity.

ERPs for the item familiarity condition also reliably exhibited their predicted effects of an early old-new effect from 400–600 ms, maximal at site Cp1 (N = 19, t(18) = 2.02, p = .029, one-tailed, Cohen’s d = .46, 95% CI [.160, ∞]; iFam: M = −1.32, SD = 3.47, SE = .797; CR: M = −2.47, SD = 2.57, SE = .59), providing concurrent validity with similar other findings of posterior distributions attributed to absolute familiarity (Bridger et al., 2014; Mecklinger & Bader, 2020; Wolk et al., 2007). The condition did not exhibit any evidence for an LPC from 600–900 ms at P3 (t(18) = 1.22, p = .119, one-tailed, Cohen’s d = .28, 95% CI [−.238, ∞], Bayes Factor = 1.27; iFam: M = −5.45, SD = 3.49, SE = .802; CR: M = −1.11, SD = 2.37, SE = .543) nor any later negative effect from 800–1200 ms at Cz (t(18) = .910, p = .187, one-tailed, Cohen’s d = .21, 95% CI [∞, .576], Bayes Factor = 1.83; iFam: M = −1.93, SD = 3.60, SE = .83; CR: M = −1.29, SD = 2.92, SE = .67). Together, the consistent findings of distinct ERP effects of an FN400, LPC, and BCN among both the established general conditions (Hits vs. CR) and the more specific conditions of response combinations (iFam, cFam, Rec., each vs Cr.) provided convergent validity of similar findings observed from different measures, while also demonstrating concurrent validity of the newer specific ERP measures of memory processing with the more established general ERP measures of memory processes.

3.5 Are the neural correlates of context familiarity different from item familiarity?

To test whether the context familiarity condition reflected the same or different kind of familiarity processing as traditional item-familiarity, we investigated if the two conditions varied during the conventional time window traditionally reported for familiarity-based processing of the FN400 (300–500 ms (Rugg & Curran, 2007; Rugg, Mark, et al., 1998)), performing a within-subjects (N=11) targeted analysis on mid-central site Cz where prior analysis identified peak activation for these conditions in both the present and prior study (Addante, Ranganath, & Yonelinas, 2012) that predicted a one-directional difference of context familiarity being more negative than item familiarity. ERPs for context familiarity were revealed to be significantly more negative than those for item familiarity (t(10) = −2.57, p = .015, one-tailed, Cohen’s d = .76, 95% CI [−inf, .−622]; cFam: M = −4.62, SD = 3.32, SE = .70; iFam: M = −2.42, SD = 3.22, SE = 1.33; Figure 5A), and topographies indicated a broad negative-going distribution of this effect across central scalp sites and throughout the epoch. To assess if these effects varied throughout the epoch a 2×3 ANOVA was performed using factors of memory condition (item familiarity, context familiarity) and time interval (300–500, 600–900, 800–1200 ms). Results revealed a marginal main effect of memory (F(1, 10) = 4.251, p = .066, n2p = .298, Cohen’s d = .65), a significant main effect of time (F(1.18, 10) = 5.13, p = .039, n2p = .339), and no memory x time interaction (F(1.34, 10) = 1.48, p = .255).

To test if these results could be attributable to the potential confound of high confidence recognition responses included in the item familiarity condition, we sought a similar control test as was described for the behavioral data of response times: by removing the high-confidence responses from the item-familiarity condition so that it was matched to the low-strength responses of the context familiarity condition5. ERPs for this contrast were available from only a limited sub-sample of the primary dataset due to there being only N = 3 participants available meeting our strict a priori inclusion criteria of the ‘item4+source unknown’ item familiarity condition (Muller et al., 2021) (see Methods section 2.3.1) (though note that the present paradigm has repeatedly been shown in prior studies to be effective and sensitive to detecting reliable differences in similar small sample sizes of specialized clinical patients on ERP measures of familiarity, confidence, and implicit memory repetition effects (Addante, 2015; Addante et al., 2023; Addante, Ranganath, Olichney, et al., 2012), in addition to conditions of metacognition (Muller et al., 2021)).

We used a directional t-test based upon the hypothesis that context familiarity exhibits more negative-going ERPs than item familiarity at the same mid-central size (Cz) and time (300 to 500 ms) that was found in the primary analysis (Figure 5C). Despite the limited sample, ERPs for context familiarityLow remained more negative than those for item familiarityLow: while this difference was not initially significant from 300–500 ms (t(2) = 1.67, p = .119, one-tailed, Cohen’s d = .962, 95% CI [−∞, −1.37], Bayes Factor of .610; iFam: M = −1.67, SD = 3.99, SE = 2.30; cFam: M = −3.49 , SD = 2.19, SE = 1.27) there was a significant effect maintained in the nearly-identical time interval of 400–600 ms that has also been commonly used for measuring the FN400 in prior studies using this specific dataset (Addante et al., 2023; Muller et al., 2021): (t(2) = 5.57, pbonf = .015, one-tailed, corrected, Cohen’s d = 3.21, 95% CI [−∞, −1.5]; iFam: M = −1.36, SD = 2.38, SE = 1.38, values: −3.29, −2.03, 1.31; cFam: M = −4.51, SD = 1.54, SE = .89, values: −5.40, −5.404, −2.73) (Figure 5C), indicating that the main effect was preserved in the more stringent control analysis.

Since we used a relatively conservative criteria for inclusion of participants in the within-subjects analysis (n = 12 artifact-free ERP trials pairwise, see Method), it provided a rigorous threshold to foster better signal-to-noise ratio of ERP effects but also resulted in a fairly small sample (N = 11). Therefore, we also performed a parallel analysis with a more liberal inclusion criteria for subjects, which could increase the sample size and challenge the findings via testing them with reduced signal-to-noise ratio (Boudewyn et al., 2018; Luck, 2014). The same results persisted when lowering the trial inclusion criteria to n = 5, which provided a larger sample of N = 25 subjects: ERPs for context familiarity still exhibited a negative difference from ERPs of item familiarity that was maximal at adjacent site Fc1 and extending broadly across bilateral mid-central regions (Figure 5B). We used a similar 2×3 ANOVA as employed in the primary analysis, using factors of memory condition (item familiarity, context familiarity) and adjacent time interval (400–600, 600–900, 800–1200 ms), and like the primary results using the more conservative inclusion criteria with a smaller sample, there was a main effect of memory (F(1, 24) = 4.37, p = .047, n2p = .15, Cohen’s d = .42) with no significant evidence for effects of time nor a memory x time interaction (both F’s < 1), indicating that ERPs for context familiarity were again significantly different from item familiarity beginning early in the epoch and persisting throughout despite lower signal-to-noise ratio (Boudewyn et al., 2018; Luck, 2014). Overall, the convergence of findings indicated that ERPs for item familiarity and context familiarity differed reliably during the traditional time window of familiarity processing, and thus that these two memory conditions represented reliably different kinds of familiarity processing on a trial-wise level within subjects (Mecklinger, 2006; Mecklinger & Jäger, 2009; Rugg, 2009; Rugg & Coles, 1995; Rugg & Curran, 2007; Rugg, Mark, et al., 1998; Shallice, 1988).

As the preceding analysis demonstrated moment to moment differences of modular cognitive processes within the same individuals over time, we next sought to determine if these patterns would persist as individual characteristics across individuals(Muller et al., 2021). The within-subjects approach used above required paired observations in both conditions of infrequent memory response patterns; however, the two conditions of item-familiarity and context-familiarity tend not to co-occur often within subjects. That is, those who tend to retrieve source memory well (i.e. better performers) also rarely report lacking source information (lesser memory), and those who tend to fail retrieving source memory do not also frequently report retrieving it. We thus utilized a between-subjects approach that treated those with sufficient trials in one condition but not the other as separate groups, in a mutually independent way (total sample of N = 26; item familiarity group: N = 9, context familiarity group = 17). This resulted in the exclusion of those from the within-subject analysis, but nevertheless converged to provide the same findings with an entirely different subset of participants, despite the inherent challenges to the hypothesis from a smaller sample.

Using a between-groups t-test, the same pattern of ERPs persisted between subjects from 300–500 ms such that ERPs for context familiarity were significantly more negative-going than from item familiarity (t(24) = 2.82, p = .005, one-tailed, Cohen’s d = 1.16, 95% CI [1.93, inf], Figure 6A), (cFam: N = 17, M = −4.62, SD = 4.63, SE = 1.23; iFam: N = 9, M = .250, SD = 3.34, SE = 1.12), and was not statistically significant during the later time intervals (600–900 ms: t(24) = 1.21, p = .119, one-tailed; 95% CI [−.698, inf], Bayes Factor = 1.57; 800–1200 ms: t(24) = .664, p = .257, 95% CI [−1.72, inf], Bayes Factor = 2.27). Overall, this mirrored the results from the preceding within-subjects analysis, independently demonstrating that these are distinct physiological patterns for familiar memory responses both at the single trial level (within-subjects) and at the level of individual variability (between-subjects), and that the context familiarity condition cannot be attributed to merely a form of the same kind of familiarity process supporting familiarity of item recognition.

Control analyses of range-normalized values revealed that the group differences between the memory conditions were not attributable to scaling artifacts or cohort effects that might be operating independent of the memory conditions. Since it is possible that the results of the between-subjects analyses could be driven in part by group-level differences in overall magnitudes of the ERP amplitudes between the ‘cohorts’ (e.g. scaling artifacts; (Haig et al., 1997; McCarthy & Wood, 1985; Urbach & Kutas, 2002; Wilding, 2006), we conducted follow-up analyses to control for this possibility. We range-normalized the ERP amplitudes of the individual conditions within each group for item-familiarity and context-familiarity, respectively (Figure 6A, bottom). This approach serves to overcome potential scaling artifacts of differential amplitudes, as it normalizes the ERPs in the dimension of amplitude magnitude and allows for investigation of where they are spatially located (Rugg & Coles, 1995; Rugg, Mark, et al., 1998; Wilding, 2006) as has been done in prior work (Addante et al., 2023; Ozubko et al., 2021; Rugg, Mark, et al., 1998). Results revealed that range-normalized values for item- and context-familiarity ERPs remained significantly different from 300–500 ms at mid-central site Cz (t(24) = 2.36, p = .027, Cohen’s d = .97, 95% CI [−1.24, −.084]; iFam: N = 9, M = .75, SD = .64, SE = .21; cFam: N = 17, M = .096, SD = .69, SE = .17) (Figure 6A), and thus that the differences between the memory conditions were not attributable to scaling artifacts or cohort effects of groups that might be operating independent of the memory conditions. Collectively, these findings indicated that the observed differences in familiarity response types reflected differential neural processing at the individual subject level of variability, in addition to the trial-wise nature of the memory processing that was identified in the preceding within-subjects analyses. These results thus converge to reveal that the two conditions represent fundamentally different kinds of familiarity operating in service of episodic recognition.

4.0 Results: Experiment 2

If there are differential neural correlates of context familiarity and item familiarity operating as distinct processes of episodic memory, then the differences should be evident in other, independent experiments. Therefore, we assessed questions of reproducibility and sought validity of the findings through replication (Ioannidis, 2005; Ioannidis et al., 2012; Nosek et al., 2022). We investigated if the differences between ERPs for context familiarity and item familiarity (Figures 3, 4, 5, & 6) could be reproduced among independent datasets that were aggregated from three previously-published studies using a nearly-identical paradigm and which had previously reported both traditional ERP correlates of item familiarity as well as ERP effects of context familiarity being separately dissociated from those of recollection, but had not directly contrasted conditions of item-familiarity to context-familiarity (Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012; Addante et al., 2011).

4.3.2 Experiment 2: Within-subjects Familiarity.

First, we assessed if the within-subjects results observed between item familiarity and context familiarity in Exp. 1 (Figure 5) could be seen at the mid-central site of Cz, using a 2×3 ANOVA with factors of memory (context familiarity, item familiarity) and time (300 – 500 ms, 600 – 900 ms, 800 – 1200 ms; N = 20). This revealed that there was a significant main effect of memory (F(1,19) = 4.55, p = .046, n2p = .193, Cohen’s d = .48), a main effect of time (F(1.26, 19) = 19.72, p < .001, n2p = .509), but no significant memory x time interaction (F(1.68, 19) = .437, p = .616, n2p = .022) (Figure 7). These findings indicated that the two conditions differed reliably at the mid central site throughout the epochs from 300 to 1200 ms, and represented a faithful reproduction of the differences originally observed between ERPs for context- and item-familiarity in the present study’s main findings (Figure 5).

Because the previous datasets (Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012) had reported the ERP effects for context familiarity also manifesting at bilateral fronto-central sites, we expanded the aforementioned analyses to also include the frontal sites F3 and F4, using a 2×3×3 ANOVA. This analysis continued to provide a faithful replication of the original findings of the present work, as there was a main effect of memory (F(1,19) = 6.69, p = .018, n2p = .261, Cohen’s d = .49), an effect of time (F(1.29, 19) = 9.69, p = .003, n2p = .338), and a site x time interaction (F(2.24, 19) = 15.20, p <.001, n2p = .444) as the previously-reported frontal sites became more pronounced later in the epoch (Figure 7). Importantly, these results provide an independent replication of the critical findings reported in the present study, comprised of entirely different participants from a different laboratory (the UC Davis Dynamic Memory Lab (2008–2010, majority Caucasian) vs. California State University – San Bernardino (2019, majority Hispanic), using different EEG hardware (BioSemi vs. ActiCHamp). Thus, the results of Experiment 1 in the main text of the present study are bolstered as valid, reliable findings observed across diverse participants from four different independent experiments (Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012; Addante et al., 2011; Muller et al., 2021).

4.3.3 Experiment 2: Controlling for confounds of confidence levels in item familiarity.

One possible alternative interpretation of the ERP differences we observed between the conditions of item familiarity and context familiarity (Figure 5) is that they could have been confounded by inclusion of high confidence recognition trials in the item familiarity condition. That is, the conditions of item- and context familiarity also vary on the factor of item recognition confidence, since the item familiarity condition includes both high and low confidence responses of strong and weak item familiarity levels, whereas the context familiarity condition contains only the low confidence recognition judgments that had correct source memory. One approach to address this issue is to exclude the high confidence responses from the item familiarity condition, even though, as noted in Experiment 1, the context familiarity condition could still differ from strong forms of item familiarity that were excluded from the analysis (Wixted, 2007; Wixted & Mickes, 2010; Wixted & Squire, 2011a). This control analysis was performed in the behavioral results reported in Experiment 1 and was found to successfully preserve the findings that the conditions for item- and context-familiarity were reliably distinct in response times.

For the ERP version of this control analysis in Experiment 1, there was only a small sample (N =3) meeting our inclusion criteria for ERPs of pairwise comparisons of the item familiarity condition with context familiarity (Figure 5C) that still provided a significant effect, but would benefit from an independent replication in a larger sample. In Experiment 2 there were more participants available (N = 12). We thus assessed ERPs for ‘item4+source unknown’ (iFamLow.) vs. ‘item4+source correct’ (cFam.), using a directional t-test based upon the a priori hypothesis derived from the prior findings in Experiment 1 that context familiarity exhibits more negative-going ERPs than item familiarity at the same mid-central size (Cz) and time (300 to 500 ms) as was found in Experiment 1. This revealed that there were still significant effects (t(11) = 1.97, p = .037, one-tailed, Cohen’s d = .60, 95% CI [−∞, −.101]; iFam: M = −.849, SD = 2.74, SE = .79; cFam: M = −1.99, SD = 2.06, SE = .593), and thus the main finding was replicated again and with a larger sample than original finding in Experiment 1.. Thus, the inclusion of the high confidence recognition responses in the condition of item familiarity is not contaminating nor confounding the results we observed of reliable differences between item- and context-familiarity (Figure 5A, Figure 7A).

4.3.4 Experiment 2: Between Subjects Familiarity.

The between-subjects ERP effects from Exp. 1 (Figure 6A) were also reproduced when aggregating across participants from prior studies (Addante, Ranganath, & Yonelinas, 2012; Addante et al., 2011) (Figure 6B). As was done in the main results, participants were included in conditions of either item familiarity or context familiarity if they met the trial inclusion criterion of sufficient trial responses (n = 12) in exclusively one condition but not the other, and these were entirely different participants than those included in the within-subjects analysis described above. This resulted in a subgroup of N = 13 for the item familiarity condition, and N = 11 for context familiarity. As familiarity effects for these prior studies were reported for 400 to 600 ms (Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012), a targeted analysis was conducted during this latency to assess if context familiarity ERPs exhibited the same negative-going difference from item-familiarity ERPS at the mid-frontal site characteristic of previous familiarity effects reported in these participants (Addante, Ranganath, & Yonelinas, 2012; Addante et al., 2011) (Fz). A between-group t-test revealed the ERPs to reliably differ (t(22) = 1.923, p = .034, one-tailed, Cohen’s d = .79, 95% CI [−∞, −.426]), with ERPs for context familiarity again being more negative-going than those for item-familiarity (N = 11, M= −2.69, SD = 4.98, SE = 1.50; N = 13, M = 1.25, SD = 5.00, SE = 1.39, respectively).

4.3.5 Experiment 2: Within-subjects Memory Condition Effects.

Next, the results of each individual condition’s memory effect were compared to correct rejections, as was done in Figure 4 of the main findings of Exp. 1. ERP correlates of context familiarity again exhibited a reliable effect from 800 – 1200 ms (N = 33) that was distributed broadly across the central-posterior scalp and maximal at site Cp1 (t(32) = −3.22, p = .003, Cohen’s d = .56, 95% CI [−3.17, 7.12]; cFam: M = 1.84, SD = 4.35, SE = 0.76; CR: M = 3.78, SD = 3.94, SE = 0.69), but for which there was no significant FN400 effects evident from 400–600 ms (t(32) = .163, p = .113, 95% CI [−.195, 1.75], Bayes Factor = 1.25; cFam: M = −.006, SD = 4.24, SE = .74; CR: M = −.782, SD = 3.84, SE = .67), nor any evident LPC at P3 from 600–900 ms (t(32) = 1.07, p = .293, 95% CI [−1.66, .515], Bayes Factor = 3.17); cFam: M = 2.19, SD = 4.42, SE = 0.77; CR: M = 2.76, SD = 3.66, SE = 0.64) (Figure 8). This finding reproduced and confirmed the Exp. 1 finding (Figure 4) that context familiarity was associated with a broad central negative effect, that it held no evidence of association with the FN400 and provided an evidence for absence of the LPC effect.

On the other hand, item familiarity was associated with significantly more positive ERPs than correct rejections at mid-central Cz from 400 to 600 ms (N = 38) (Figure 8), reproducing the FN400 effect reported in Figure 4 of the main results section (t(37) = 3.77, p < .001, Cohen’s d = .61, 95% CI [ −2.38, −.715]; iFam: M = .362, SD = 3.78, SE = .61; CR: M = −1.18, SD = 3.79, SE = .62). During the later latencies of 600 to 900 ms and then 800 to 1200 ms associated with recollection and context familiarity, respectively, there were no reliable differences evident between the two conditions (t(37) = 0.95, p = .348, 95% CI [−1.44, .519], Bayes Factor = 3.76; iFam: M = 3.27, SD = 4.47, SE = .725; CR: M = 2.81, SD = 4.34, SE = .703); t(37) = .160, p = .877, 95% CI [−1.08, .923], Bayes Factor = 5.66; iFam: M = 3.49, SD = 4.39, SE = .71; CR: M = 3.41, SD = 4.19, SE = .68) (Figure 8), reproducing the same pattern of results reported in the present study’s main results (Figure 4). These findings indicated that the neurophysiological correlates of the item familiarity condition were linked with a FN400 effect but no comparable evidence for any LPC or BCN effects (and with Bayes Factor evidence of absence for each other effect, respectively), consistent with the well-established profile of item-familiarity ERPs observed within and across the two present Experiments and among the broader literature of other studies (Addante, Ranganath, & Yonelinas, 2012; Friedman, 2013; Mecklinger, 2006; Ozubko et al., 2021; Rugg & Curran, 2007; Rugg, Mark, et al., 1998).

Furthermore, similar to the findings of the present study, the recollection response condition exhibited both a reliable FN400 and LPC effect when compared to correct rejections (N = 54), (t(53) = 6.14, p < .001, Cohen’s d = .84, 95% CI [1.43, 2.81]; Rec.: M = .207, SD = 4.92, SE = .668; CR: M = −1.915, SD = 4.53, SE = .61; t(53) = 4.73, p < .001, Cohen’s d = .64, 95% CI [.942, 2.33]; Rec. M = 3.92, SD = 3.96, SE = .54), but provided no evidence for any BCN (t(53) = .585, p = .561, 95% CI [−1.55, .848], Bayes Factor of 5.72) ( Figure 8, 9). Collectively, these analyses of previously published datasets using the same memory paradigm as the present study provide a replication of the results observed in the current investigation. These patterns of effects reported here across both behavior and neurophysiology, both within and between subjects, and their replications among several other independent studies cannot be accounted for by a priori predictions from existing models of only a single- or dual processes of episodic memory.

4.3.6 Experiment 2: Within-subjects Old-New Memory Effects.

Using the same aggregated data also reproduced the other findings reported in the present study of a triple dissociation of ERP effects for item-familiarity, recollection, and context familiarity among the general memory contrast of hits minus corrections (Figure 3), whereby a 3-way interaction was originally found among an FN400, LPC, and BCN. For this analysis, we focused upon the latencies where effects had been observed in the prior studies for the FN400, LPC, and BCN (400 to 600 ms, 600 to 900 ms, 1000 to 1400 ms, respectively6), and focusing on sites Cz, P3, and Fc1 reported for effects in the present study (N=56). Results from a 2×3×3 ANOVA revealed that, similar to the main findings from Experiment 1 of the present study, there was no initial main effect of memory (F(1,55) = 2.19, p = .145, n2p = .038), but there were significant main effects of site (F(1.68, 55) = 5.82, p = .007, n2p = .096), and significant interactions of memory x time (F(1.52, 55) = 32.77, p < .001, n2p = .373), site x time (F(1.57, 55) = 23.59, p < .001, n2p = .300), plus a significant memory x site x time interaction (F(3.20, 55) = 4.77, p = .003, n2p = .080) (Figure 9).

Consistent with the a priori hypotheses, results from directional t-tests based upon our a priori hypotheses revealed that there were significant effects of the FN400 (t(55) = 6.16, p < .001, one-tailed, Cohen’s d = .82, 95% CI [1.27, inf]; Hit: M = −.208, SD = 4.39, SE = .588; CR: M = −1.95, SD = 4.48, SE = .599), LPC (t(55) = 3.06, p = .002, one-tailed, Cohen’s d = .41, 95% CI [.402, inf]; Hit: M = 3.13, SD = 3.63, SE = .49; CR: M = 2.24, SD = 3.99, SE = .534), and BCN (t(55) = 1.85, p = .035, one-tailed, Cohen’s d = .25, 95% CI [−∞, −.083]; Hit: M = .916, SD = 3.68, SE = .49; CR: M = 1.80, SD = 4.60, SE = .62). The BCN effect identified at the a priori selected location of Fc1 was observed to have extended broadly and been maximal at site Pz: t(55) = 4.33, p < .001, one-tailed, Cohens d = .58, 95% CI [−inf, −1.14], Together, these results revealed the same finding of a significant memory x site x time interaction in an independent data set, effectively reproducing the original findings reported in the present study (Figure 3, Figure 9). Specifically, these results confirm that the three different memory effects are present within the most general measures of memory historically used in the field, in addition to the specific conditions described herein.

4.3.7. Are the neural correlates of context familiarity reflecting non-conscious memory processing?

While the responses given for context familiarity were defined as being explicit declarative responses from people for item recognition hits and also had overtly-declarative correct source memory judgments, it nevertheless remained theoretically possible that they reflected a form of non-conscious processing such as that related to guessing from 100–300ms (Voss & Paller, 2010) or fluency that has been known to occur characteristically in earlier latencies of 200–400ms (Leynes & Zish, 2012; Leynes & Addante, 2016; Ozubko et al., 2021) for Review see (Mecklinger & Bader, 2020). Thus, to rule this possibility out, we sought to verify if the memory processing associated with the context familiarity condition’s effects in these participants was different from the hallmark signs of non-conscious, implicit memory signals that have been extensively characterized in the prior literature (i.e.: recognition misses (Addante et al., 2023; Mecklinger & Bader, 2020; Ozubko et al., 2021; Rugg, Mark, et al., 1998), old items endorsed as 1 or 2 responses during the memory retrieval test). The null hypothesis of this analysis was that if the context familiarity condition was actually representing an implicit memory process, then it should show the same, or similar, activity as other implicit memory processing, such as the misses in the time windows when each effect is known to operate, respectively (200–400 ms for implicit guessing, 800–1200 ms for context familiarity), and thus predict no significant differences in ERPs.

Experiment 1: Misses:

We assessed ERPs for misses and context familiarity at the centro-parietal site of peak activity identified earlier for context familiarity (Cp1), using a within-subjects two-factor ANOVA with repeated measures on both factors (N = 27). This analysis revealed no significant main effect of condition (F(1,26) = 1.03, p = .321, n2p = .038), identified a significant main effect of latency (F(1.91, 26) = 19.95, p < .001, n2p = .434), and importantly revealed a significant interaction of memory condition and latency (F(2.16, 26) = 3.78, p = .026, n2p = .127), which was further investigated with follow-up analyses (Figure 10A). This revealed that ERPs for context familiarity were significantly more negative-going than misses at centro-parietal sites (Cp1) from 800 – 1200ms (t(26) = 2.31, pholm = .029, corrected, Cohen’s d = .44, 95% CI −2.40, −.140]; Context Familiarity M = −2.69, SD= 3.46, SE = .66; Miss M = . −1.43, SD = 2.34, SE = .45) though not in the earlier latency of 200–400ms (t(26) = .443, p = .660; Bayes Factor = 4.37 indicating moderate evidence for the null). This demonstrated that the context familiarity condition was exhibiting substantively different physiology from recognition misses at centro-parietal sites during 800 to 1200 ms, refuting the null hypothesis that they were both reflecting the same implicit process outside of conscious awareness (e.g. guessing driven by implicit factors).

Since topography maps of Experiment 1 also indicated an early frontal activity pattern for this contrast, we conducted an exploratory analysis of the evident frontal region at its maximal left frontal site of F3. Inspection of these findings revealed that ERPs for context familiarity were more positive-going than misses in the early latency of 200–400ms (F3: t(26) = 2.55, p = .034holm, corrected, Cohens d’ = .49, 95% CI [−2.07, −.221]; cFam: M = 1.57, SD = 2.89, SE = .55; Miss: M = .427, SD = 3.01, SE = .57), continuing marginally 400–600 ms (t(26) = 1.90, pholm = .068, corrected, Cohen’s d = .37, 95% CI [−1.85, .07]; cFam: M = −4.36, SD = 3.62, SE = .697; Miss: M = −5.26, SD = 4.11, SE = .792) (Figure 10A). Overall, these analyses revealed reliable, significant differences between conditions of context familiarity and implicit memory misses in Experiment 1 but were also assessed for the replicability in Experiment 2.

Experiment 2: Misses:

The same pattern of differences between context familiarity and misses was also observed as being independently reproduced by targeted analyses in Experiment 2’s data sets (N = 27) based upon directional hypotheses derived from the preceding findings for the same contrasts in Exp. 1, during the 800 to 1200 ms time interval of context familiarity effects, occurring at the same site as the primary results (Cp1: (t(26) = 1.79, p = .043, one-tailed, Cohen’s d = .34, 95% CI [.046, ∞], cFam: M = 1.68, SD = 3.68, SE = .707; Miss: M = 2.68, SD = 3.35, SE = .644; maximal activation site of P4 (t(26) = 2.35, p = .013, one-tailed, Cohen’s d = .45, 95% CI [.396, ∞], cFam: M = .689, SD = 3.33, SE = .64; Miss : M = 2.14, SD = 3.34, SE = .681) (Figure 10B). Similar results to the primary findings in the earlier latencies of 200–400ms and 400–600 ms were also observed in Experiment 2’s datasets, revealing more positive ERP activity for context familiarity, albeit in more central-parietal sites of these participants (maximal at Cp5) (t(26) = 2.62, p = .007, one-tailed, Cohen’s d = .50, 95% CI [∞, −.400], cFam: M = 1.93, SD = 2.71, SE = .52; Miss: M = .786, SD = 2.84, SE = .55; t(26) = 1.81, p = .041, one-tailed, Cohen’s d = .35, 95% CI [∞, −.062], cFam: M = .019, SD = 3.08, SE = .594; Miss: M = −1.08, SD = 3.48, SE = .67, respectively).

Together, these results from both Experiments 1 & 2 revealed that neural activity for explicitly declared context familiarity judgments was reliably differentiated from that of implicit misses at the human scalp, during early latencies (200–400 ms) and later latencies (800 – 1200 ms), and across several different independent studies. This line of converging evidence refuted the null hypothesis that both conditions of context familiarity and implicit misses had the same functional significance or were representing the same cognitive process (Allan et al., 1998; Bridger et al., 2012; Mecklinger & Jäger, 2009; Rugg, 2009; Rugg & Coles, 1995; Rugg, Mark, et al., 1998; Shallice, 1988). More broadly, recent work in the same experimental paradigm has also identified that other, more pure, kinds of implicit recognition processes can operate distinctly from misses(Addante et al., 2023), but that pattern of implicit processing (positive-going right parietal activity from 600–1000 ms) was entirely opposite of the negative-going central site patterns of ERPs characterized here for context familiarity, and thus not possible as a viable account of the present findings either. In sum, explicit context familiarity judgments were supposed to be equivalent to misses if representing similar implicit processing, but instead were significantly more positive than implicit misses in early epochs at frontal sites and then reversed to have significantly more negative voltage than at centro-parietal sites at later times in the epoch, thus providing clear evidence of divergent validity from the measures of implicit memory processing. This pattern for context familiarity cannot easily be attributable to accounts such as guessing or signals outside of episodic awareness and lending further weight against the possibility that context familiarity judgments were driven by implicit processes.

5.0 Discussion.

5.1 General Summary of Findings.

We began the investigation to understand how (by which process?) people make accurate source judgments for items they recognize with low confidence. Emphasizing reproducibility in science (Button et al., 2013; Collaboration, 2015; Ioannidis, 2005; Ioannidis et al., 2014; Ioannidis et al., 2012; Moonesinghe et al., 2007; Nosek et al., 2022; Yong, 2012) the current study first replicated the original behavioral performance and ERP effects of context familiarity as previously reported (Addante, Ranganath, & Yonelinas, 2012). Extending those findings, the main finding here was that the effects of item familiarity, context familiarity, and recollection were reliably dissociable in multiple measures of both behavior and neural activity; these ERP effects were then reproduced in independent datasets. These patterns of effects cannot be accounted for by a priori predictions made by models of only a single- or dual processes of episodic memory without making post-hoc modifications. This provides strong evidence that these three observations are mutually exclusive, distinct episodic memory processes. That is, item- and context-familiarity appear to be operating as fundamentally different kinds of memory processes, based upon established norms for identifying cognitive processes via neuroimaging and behavioral data (Duzel et al., 2001; Duzel et al., 1997; Henson, 2005; Leynes et al., 1998; Mecklinger & Jäger, 2009; Rugg, 2009; Rugg & Coles, 1995; Rugg, Mark, et al., 1998; Shallice, 1988; Simons, 2009; Yonelinas, 2009). Discussion of each finding is expanded upon in the sections below.

5.2 Behavioral findings.

Item familiarity and context familiarity were found to be behaviorally distinct in measures of response time, both in item judgments and in source judgments. Control analyses that held memory strength constant among conditions still preserved these differences and demonstrated that they were not confounded by recognition confidence levels. Item familiarity responses were faster during item judgments, whereas context familiarity responses were faster during source memory. It is difficult to account for these patterns of dissociations if attributing the responses to the same underlying cognitive processes of a single familiarity signal (Ranganath, 2010; Wixted & Mickes, 2010; Yonelinas, 2002; Yonelinas et al., 2010). Context familiarity responses were also responded to significantly different than were responses of recollection, with recollection being faster than both item- and context- familiarity conditions during recognition and slower during source memory (Figure 2). There are a few ways to view these patterns of results : (1) They are consistent with existing findings of the relative speeds of recollection and familiarity (recollection being slower and familiarity being faster (Johnson et al., 1993; Yonelinas, 2002); (2) they are contrary to traditionally-held patterns of recollection being slower than familiarity. Each view will be discussed in more detail below, but importantly, regardless of interpretation, the memory conditions were reliably distinct in behavioral measures, indicating that mnemonic processing for context familiarity is distinct from item familiarity and recollection.

Traditionally, recollection has been conceptualized as a slower, deliberative process, whereas familiarity was modeled as a faster, automatic process (Yonelinas, 2002). As reviewed in the Introduction, source memory judgments are typically considered a general (though not exclusive) measure of recollection, which in the present study was reliably slower during source memory than the conditions of both context- and item-familiarity responses. Those results from source memory RTs are thus consistent with traditional characterizations of recollection as a slower, deliberative process (Yonelinas, 2002). Accordingly, item recognition tasks are usually considered to be a relatively better measure of familiarity-based responding, and faster response times observed here for item-familiarity than for context-familiarity in the present findings can thus be seen as it being the item (not the context) that is familiar during this early stage of the memory task. In this view, that recollection responses were also faster in item recognition would be seen merely that their inclusion of the highest confidence responses was driving a concurrent higher item familiarity and thus faster recognition time, consistent with traditional patterns. As the context familiarity responses then go on to become faster than item familiarity responses during the source judgments, it preserves the traditional notion that familiarity is also a faster automatic process- this time it is familiarity for context that is differentially faster (not items). That is, in a traditional view of RTs results are seen as showing that the two different types of familiarity (item, context) are faster than each other in each of the respective domains in which they are tested for representing differentially (item recognition, source memory), and ultimately consistent with the common findings that familiarity operates faster, while recollection is a slower process.

Alternatively, the pattern of faster recollection during item recognition responses (and slower than context familiarity in the source judgments) could also be seen as departing from common findings that recollection is generally a slower process than item familiarity. While some studies have found that recollection can sometimes be faster than familiarity in some instances (Gillund & Shiffrin, 1984; Trott et al., 1999), that discrepancy was suggested to likely be due to instructions to participants that they respond after both memory processes are complete- thereby delaying the responses times (Yonelinas, 2002). The current paradigm also requires subjects respond to a combination of both item and source memory confidence ratings on each trial, which could similarly impose complex demands on their memory judgements. Moreover, recent studies noted below have reported recollection to have faster responses than familiarity during recognition, suggesting that the traditionally assumed speeds of recollection and familiarity should not necessarily always be taken as simply 1-directional (i.e.: Familiarity > Recollection) and that is likely more complex in directionality.

For instance, a review of pupillometry studies of recognition memory concluded that, contrary to common observation of behavioral studies, recollection is not always the characteristic ‘slow and effortful’ process typically assumed, but can also be quite ‘fast and easy’ in its manifestation as a memory process (Goldinger & Papesh, 2013). When more advanced modeling methods were used for measuring the speeds of memory processes across six different experiments, recollection was again found to operate as both fast and slow (Brainerd et al., 2019), similar to findings in the present data. Such patterns are also consistent with proposals that recollection is a two-stage process that includes both an initial rapid assessment of memory, plus a later second stage that is slower and more deliberative (Moscovitch, 2008), and with behavioral studies suggesting a third recognition process of familiarity (Coane et al., 2011). Together, the response time profiles observed in the present work (Figure 2) are consistent with other studies (Linde-Domingo et al., 2019), while also adding insight from specific memory response conditions that have not previously measured response times, to reveal that regardless of chosen interpretation views, the conditions remain nevertheless as neurocognitively distinct.

5.3 Physiological findings.

In physiological measures, first, a triple dissociation of activity patterns was evident in general ERP effects observed during the basic memory measure of hits vs. correct rejections, qualified by a significant condition x time x location interaction that revealed a mid-central FN400 effect from 400–600 ms, a left parietal LPC Effect from 600–900 ms, and a broad central negativity effect (BCN) from 800–1200 ms, and replicated in a second independent data set. Second, when investigating individual ERP effects of the specific memory conditions of item-familiarity and context-familiarity, each was found to be reliably different within-subjects and across subjects and was dissociated across the different latencies and locations on the scalp as noted above, and each was also dissociated from recollection. The same results were observed for each independent condition when contrasted against the standard ERP baseline condition of correct rejections; all of these findings were also reproduced in archival data from independent studies. This suggests that differential neurocognitive processes underlie each of the memory judgments, respectively (Duzel et al., 1997; Mecklinger & Jäger, 2009; Rugg, 2009; Rugg & Coles, 1995; Rugg, Mark, et al., 1998; Shallice, 1988).

The neural correlates for context familiarity did not exhibit the LPC nor the FN400 effects (the putative correlates of recollection and item-familiarity, respectively) in any analysis. The absence of such standard ERP effects was further bolstered by the Bayes Factor results, which provided moderate evidence that both the LPC and FN400 were not present for these memory judgments instead of being merely undetectable. This conclusion is further bolstered by the finding that context familiarity exhibited significant effects present elsewhere in time and scalp location, so the absence of LPC and FN400 correlates could not be merely a case of insufficient statistical power to detect such differences.

5.3.1 Why might context familiarity effects occur later than item familiarity?

In our main ERP contrasts of each memory condition (vs. correct rejections, Figure 4, Supplementary Figure 1), context familiarity manifested effects later than item familiarity (800–1200 ms vs. 400–600 ms, respectively), and behavioral response times for context familiarity judgments were similarly later than those for item familiarity (2710 ms, 2361 ms, Figure 2, Table 2) [ though we note that during the source memory judgment context familiarity responses were faster than item familiarity (2075 ms, 2612 ms, respectively)]. To a certain extent, the present work is not designed to directly answer why the times differ, but importantly is reporting the original finding that they are, in fact, different at all. One speculative reason is that context information may be broader than item information: encompassing additional elements/features that require neural pattern completion (Anderson et al., 2016; Horner et al., 2015; Leutgeb & Leutgeb, 2007; Rolls, 2013; Yassa & Stark, 2011)- and that this could take the extra few hundred milliseconds that we observe here. We note here that the later ERP time interval (800–1200 ms) was not the only time that context familiarity was found to exert an effect: it also differed reliably from item familiarity ERPs during the early epoch of 300–500 ms (Figures 5, 6, 7), too. We underscore that this early finding was important to the present work, because it demonstrated that the two memory conditions are occurring through differential patterns of neural activity early in the epoch (300–500 ms)- not just later.

There should also be caution in over-interpreting the response time patterns of ERP effects earlier than the behavioral effects. The timings of ERP effects are occurring substantially before the later behavioral memory judgments, which might seem odd to researchers unfamiliar with ERP research on memory. However, this pattern of faster ERP times than behavioral times for the same memory processes has been generally true for most studies throughout the literature. This is because ERP effects have generally never been thought to represent a direct 1:1 mapping of memory process (as discussed in (Addante, Ranganath, Olichney, et al., 2012)) but instead are widely regarded as “the putative neural correlates” of the memory processes (Rugg & Curran, 2007). That is, the ERPs are covert measures of memory-related activity(Luck, 2014), but the overt memory behavior is generally thought to occur later in time (when people indicate their memory response) after the initial neural processing has set the stage for integrating the ensuing cognition and then deploying the actual motor response. So, there should be caution exerted in assuming too much about the timing of ERPs- what matters most, in our view here, is that they differ at all, and thus must be representing at least partially non-overlapping cognitive processing (Duzel et al., 1997; Rugg, 2009; Rugg & Coles, 1995; Rugg & Curran, 2007; Rugg, Mark, et al., 1998; Shallice, 1988).

5.4.0 Alternatives, Limitations, and Future Directions.

How are study participants making accurate source memory retrieval for items recognized with low confidence? There are several different accounts which could potentially be used to interpret them for future explorations. We will address four possible alternatives and limitations here for what other accounts might explain the pattern of findings for context familiarity: a) generic item familiarity, b) delayed recollection, c) the late parietal negativity (LPN), and d) implicit memory or guessing. observed in other studies for controlled search processes. Each is reviewed below.

5.4.1 A single signal of generic familiarity?

One possibility to account for the present findings is that results for context familiarity could perhaps ‘simply’ reflect a variant of generic familiarity-based processing for which item and context are merely differences in levels but not kinds, such as the distinction between absolute (pre-experimental) and relative familiarity (in-the-experiment) of the stimuli (Bridger et al., 2014; Mecklinger & Bader, 2020). This potential account would propose that the ERP measures of each condition of item- and context-familiarity would vary incrementally rather than fundamentally, similar to how ERP measures have been found to vary incrementally to incremental gradations of item familiarity strength (Addante, Ranganath, & Yonelinas, 2012; Woodruff et al., 2006; Yu & Rugg, 2010). Yet, that pattern was not found in the present study, which instead revealed the two conditions to vary in fundamental, not incremental, ways. The present effects of context familiarity were significantly dissociable from the effects of item familiarity- in measures of both behavior and physiology, so cannot be the same (Mecklinger & Jäger, 2009; Rugg, 2009; Rugg & Coles, 1995; Rugg & Curran, 2007; Rugg, Mark, et al., 1998; Shallice, 1988). Thus, a shared, single generic process of familiarity cannot account for this pattern of memory responses that have been observed here and in several independent data sets.

The ERPs for conditions of item- and context-familiarity were found to significantly differ in N400 potentials from 300–500 ms, and the N400 component has also been known to be influenced by factors such as context (Bader et al., 2023; Bridger et al., 2012; Kutas & Federmeier, 2011; Leynes & Mok, 2020; Leynes & Upadhyay, 2022; Leynes et al., 2017; Voss & Federmeier, 2011). This association with context is consistent with the present interpretations of the psychological constructs we describe as ‘context familiarity’ and was differentiated from the opposing patterns observed for item-familiarity. The posterior distribution of the FN400 effect for the item-familiarity condition is consistent with other findings ascribing it to a sub-type of absolute familiarity (as opposed to relative familiarity, which is reported to have a more anterior scalp distribution (Bader & Mecklinger, 2017; Bridger et al., 2012; Mecklinger & Bader, 2020; Stróżak et al., 2020). The FN400 effect observed in the present study had the same mid-central topography as was reported originally for this paradigm(Addante, Ranganath, & Yonelinas, 2012), which is between the frontal and posterior distributions observed in other studies that manipulated factors of relative and absolute familiarity (Bader & Mecklinger, 2017; Bridger et al., 2014; Leynes & Mok, 2020; Leynes & Upadhyay, 2022; Leynes et al., 2017; Mecklinger & Bader, 2020; Stróżak et al., 2020), respectively. This mid-central distribution may likely reflect a combination of both relative and absolute familiarity with the stimuli in this particular paradigm since the study was not designed to differentiate those factors. Finally, accurate source memory has also been shown to be achieved through reliance on cognitive processes such as item familiarity (Leynes et al., 2016; Parks et al., 2011) and unitization (Bader et al., 2010; Bastin et al., 2013; Diana et al., 2008; Parks & Yonelinas, 2015; Quamme et al., 2007; Rhodes & Donaldson, 2007; Tu et al., 2017), and the Source Monitoring Framework (Johnson et al., 1993; Mitchell & Johnson, 2009) has proposed that familiarity can support source judgments in some contexts. However, weighing against such possibilities of the source judgments being contributed to by those forms of item-familiarity is that prior ERP findings for item familiarity and unitization contributing to source judgments are found as positive-going ERP difference effects (Leynes et al., 2016; Mecklinger & Bader, 2020; Rhodes & Donaldson, 2007), instead of the significant negative-going ERP effects observed in the current data.

5.4.2 A delayed recollection process occurring after the ERP measures?

Another possible account for the effects of the context familiarity condition is that perhaps recollection is a bit delayed and happening later in the epoch7. There are several broad reasons why it is likely that the ERP effects of the context familiarity condition could not plausibly be reflecting a “Delayed recollection” process, which we will follow with several additional specific reasons. For instance, first there is scant evidence to support a contribution of recollection to the condition we measured as context familiarity: there has been no ERP evidence of the putative neural correlates of recollection for this condition across several different studies, and this absence of evidence has been quantified by Bayes Factor analyses providing consistent evidence of absence. Alternatively, the evidence that has accrued across various studies and measures continues to unambiguously indicate that the condition of context familiarity differs from recollection trials in substantive ways that include behavioral response times for both recognition and for source memory judgements, and several physiological measures across several experiments.

The context familiarity condition also tends to be low confidence source judgments, which argues against the premise that a later process of slow recollection might be driving their trials following a delayed memory search. This is because if recollection occurred later then it could be assumed to have more high confidence correct source judgments (and yet the response data doesn’t have that profile). That is, if recollection is “delayed”, then the common features of recollection should also be evident in at least a reasonable proportion of those source memory response conditions for when that response is ‘eventually’ happening, and those features of recollection commonly include high confidence responses for source memory (Curran, 2004; Parks & Yonelinas, 2009; Ratcliff & Starns, 2009; Woodruff et al., 2006; Yonelinas, 2001; Yonelinas, 1999; Yonelinas, 2002; Yonelinas et al., 2010; Yonelinas & Parks, 2007). Since that, too, runs contrary to the pattern of responses evident in the present data, it weighs against the theory of ‘delayed recollection’. This alternative account of delayed recollection would also de facto hold that a delayed recollection arrives later in time, even though the source judgments for this particular condition are provided faster than the source judgments in the actual recollection condition. As such, it is difficult for that to be reconciled in light of the pattern of behavioral responses. The ‘delayed recollection’ account also would have trouble explaining why the context familiarity ERPs differed reliably from the item familiarity ERPs during the early latency of 300–500 ms if it was only representing a delayed process happening later after 1200 ms. For these reasons, it is improbable (and unsupported by evidence) to speculate that the effects of context familiarity could plausibly be reflecting a delayed recollection; however, more specific reasons to reject this theory also exist and are detailed below.

The term of ‘delayed recollection’ appears only sparsely in the memory literature. One of the few extant studies of “delayed recollection” using ERPs (Liesefeld et al., 2016) found that ‘delayed recollection’ occurred under conditions of “changed viewing conditions”- which were not a part of, nor directly applicable, to the present study. Notably, that study did not find any negative-going ERP effects to be associated with their condition of ‘delayed recollection’, which was instead associated with positive LPC measures (the opposite of the present findings). This provides evidence against a ‘delayed recollection’ account and is difficult to support as a viable account of our present findings. Another related ERP study (Rivas-Fernández et al., 2020) claiming ‘delayed recollection’ did report a later negative-going difference for hits vs correct rejections that they attributed as late parietal negativity (LPN, see section 5.4.3 below), but in their paradigm their reported LPN effect actually became more reduced in their ‘delayed memory condition’, and was instead larger in their condition of immediate memory retrieval. This, too, is the opposite of what would be predicted from a ‘delayed recollection’ account of the present findings, and thus it, too, ends up providing evidence against the ‘delayed recollection’ account. Could the BCN effects we attribute to context familiarity instead be reflecting processes of non-criterial recollection (Yonelinas & Jacoby, 1996)? If so, one could predict that then there would still be an LPC associated with recollecting the non-criterial information- a pattern of results that we did not see in the present series of studies, and for which there was Bayes Factor evidence of absence.

The alternative theory of ‘delayed recollection’ is further complicated by other problematic aspects: it would predict no differences in ERPs between the condition we call context familiarity and that of correct rejections because if there wasn’t memory/processing yet (i.e. it occurred via a ‘delayed recollection’) then there would be no reason for the condition to differ from a non-memory condition baseline of correct rejections. But that is not what was found in the data: there is in fact a significant difference between the condition of context familiarity and correct rejections, and we need to then know why- or, at least, be able to offer a plausible account to explain why that difference exists. Our account provides at least one viable and parsimonious explanation (it is the dissociable memory process of context familiarity), whereas the delayed recollection account does not provide an explanation.

The theory of delayed recollection is additionally difficult to sustain when one thinks about it more deeply to its natural consequent end. It is, in effect, saying that recollection might be present later in time if only we were to look (a potentially unfalsifiable hypothesis). One problem with that is that it is improperly assuming a direct 1:1 mapping of cognitive processes to ERP effects, de facto assuming that the timing of one effect (ERPs) must coincide with the timing of another effect (the cognitive process). However, it is widely understood that ERP effects of memory are considered to be covert measures that occur separate to the overt subjective experience of conscious memory reported by the subject (Addante et al., 2023; Kandel, 2013; Luck, 2014; Rugg, 2009; Rugg & Coles, 1995; Rugg & Curran, 2007). Thus, it requires acknowledging that both item-familiarity and recollection are both normatively occurring in a delayed manner evident in their reported response times (in the present data: 2.4 seconds, 1.8 seconds, respectively) that is preceded in time by their respective ERP correlates (e.g. ~400 msec, ~800 msec, respectively). That is, all recollection is delayed, at least in the sense that the traditional recollection condition’s responses are normatively delayed from the earlier ERPs effect correlates that are well-established as the LPC. Why then would it now be something different and special for this one condition of accurately retrieving the source- especially-especially when considering that the context-familiarity source memory responses are notably faster (e.g. happening less delayed) than both recollection and item-familiarity? The post-hoc modifications needed to support the delayed recollection account’s uniqueness to the context familiarity condition while leaving the canonical ERP effects of item-familiarity and recollection unaccounted for are thus deemed untenable.

5.4.3 The late parietal negativity (LPN.)

Another common suggestion about the broad central negative (BCN) ERP effects in the present study is that they could represent the Late Parietal Negativity (LPN)8, which is a heterogeneous effect sometimes seen in studies of source memory and other tasks (for Reviews see (Johansson & Mecklinger, 2003; Mecklinger et al., 2016)). The LPN has been linked with a controlled search process for episodic memory and is coarsely thought to reflect the mental processes of re-constructing the prior episode. As a brief primer, the LPN was proposed to reflect a reconstructive or evaluative process in memory search when memory features are not fully recoverable (Johansson & Mecklinger, 2003; Mecklinger et al., 2007; Mecklinger et al., 2016). Prior work studying the LPN has identified early and late variants of the LPN, with the earlier variant occurring approximately 600 to 1200 ms post stimulus onset and being variously interpreted as a) a mnemonic search for context (Herron, 2007), b) re-activating context-specifying information from an encoded episode (James et al., 2016; Mecklinger et al., 2007), or c) during later epochs that we did not measure (1200–1900 ms), associated with the evaluation of fluency (Herron, 2007; Kurilla & Gonsalves, 2012; Wolk et al., 2007) or d) reflective of the absence of information (1200 to 2400 ms) (Leynes, 2012).

The results here may appear at first glance to be similar to the late parietal negativity (LPN), in that they are negative-going ERP differences that occur relatively late in the epoch, and extend to some parietal regions (though observed here & previously to emanate from foci in more central regions (Addante, Ranganath, & Yonelinas, 2012)). As such, the present findings could potentially be seen to fit one/some of those 4 different accounts of the LPN’s functional significance. However, ‘similar’ does not mean ‘same’, and upon closer inspection there are several critical distinctions between the present findings that differentiate them on orthogonal vectors of sensitivity to different memory conditions, reported scalp locations, and timing. Thus, there are a series of reasons why the LPN interpretation remains problematic and difficult to reconcile with the available data. Because the general similarities of the two patterns of effects can seem ambiguous at first, we provide a thorough address to disambiguate the two below.

First in determining if the present study’s broad central negativity (BCN) might instead be the same effect as the late parietal negativity (LPN) is a direct appeal to what the authors of the primary Review of the LPN have said about the matter (Mecklinger et al., 2016), where it was specifically noted that the LPN was not the same effect: “… in the studies by Addante et al. (2012a,b)… these components are unlikely the same ERP component as the LPN.” (page 631), and “Addante et al. (2012a) examined ERP correlates of source memory for items that were recognized low confidence… the rather broad scalp distribution and the early onset of the negativity do, however, not necessarily support an LPN interpretation of this effect”. (page 634(Mecklinger et al., 2016)). The original report of the negative ERP effects of context familiarity (Addante, Ranganath, & Yonelinas, 2012) also similarly provided why the two were not the same effects, concluding that: “…these factors make it difficult to attribute the observed effects as an LPN” (page 448). So, neither of the authors of either of the two effects thought they were measuring the same things, and this bears some weight towards a measured interpretation of the findings.

Another possible account of the present findings as an LPN effect is that they may reflect a “controlled search process” that is one of the core (albeit vague) functional significances ascribed to the LPN (Mecklinger et al., 2016). There are several lines of reasoning weighing against this possibility. In theory, every trial of the study should be inducing participants to engage a ‘controlled memory search’ simply by virtue of the memory question being asked of them. That is, the participants are expected to be searching their memory on every trial for if they remember the item and the source of its encoding condition, then reporting it. In that sense, the controlled memory search would be a variable that is held constant across all conditions, thus canceling out in the contrasts, and leaving open the question of why this particular memory condition would show the negative ERP effect when compared to other conditions that also involve the searching of memory do not show the effect. Furthermore, if the present effect was representing a controlled search process, then it would have also presumably been seen in the recollection condition when compared to correct rejections, if participants searched memory for controlled recollection of the information. However, the same negative ERP effects weren’t seen in the recollection condition (which had far more trials included for statistical power, too), nor were the same effects seen for the item-familiarity condition in which participants were searching for the unknown source information as well, as described further below.

Expanding upon this line of reasoning, consider the item familiarity condition (i.e. successful recognition memory hits, but lacking source memory): participants report that they do not have source memory and spend significantly longer response times on the source memory judgments. If participants had made a correct item judgment on the basis of item familiarity, and could not recollect the source of the event, then presumably they would certainly be engaging in a controlled search process of their memory to see if they could find that source information. This provides a solid premise to infer that the condition would also include a “controlled search of memory” for the source information that still comes up empty for them, and makes a clear hypothesis that we should see a major negative-going ERP effect on these trials if it reflected the same thing as the LPN (Mecklinger et al., 2016). Yet, that is not what the present data shows, nor in several of the independent replications of the finding, and instead we see that the item familiarity condition differs reliably and consistently from the context familiarity condition, and that they each reliably differ from a baseline condition of correct rejections. Hence, the present differences must be attributable to another variable than just a controlled search of memory, and this creates a challenge for the LPN interpretation.

An additional point of differentiation of the present results from the LPN is the different patterns of topographic and temporal activities. Differences in topographical and temporal profiles of neural correlates of behavioral responses has been widely taken to reflect dissociable cognitive processes (Addante et al., 2023; Allan et al., 1998; Kwon et al., 2023; Mecklinger & Jäger, 2009; Rugg, 2009; Rugg & Coles, 1995; Rugg & Curran, 2007; Rugg, Mark, et al., 1998; Woodruff et al., 2006). The LPN is generally found by other studies to be maximal in posterior parietal areas, whereas the current effect was observed in superior central regions of the scalp and distributed broadly to both frontal and central-parietal areas, and had been previously differentiated from posterior sites (Addante, Ranganath, & Yonelinas, 2012). In regard to timing, the LPN has been previously characterized as beginning during periods after the epoch ended in the present study (1200 ms) (Mecklinger et al., 2016). Among those irreconcilable factors was sensitivity to memory conditions: the LPN has been reported to either be invariant to source accuracy (i.e.: it was present for both correct and incorrect conditions (Herron, 2007)) or even to be larger for incorrect source judgments (Wilding, 1999). However, the current conditions’ effects have been found to be specific for correct source memory when compared directly to incorrect source memory (Addante, Ranganath, & Yonelinas, 2012). Furthermore, other lines of evidence weighing against the LPN account of our findings comes from testing a direct prediction made that the LPN should be present “not necessarily tied to successful memory retrieval but also observed during misses…” (pg. 635 (Mecklinger et al., 2016)). We did not find that prediction supported in the present data, as the present findings directly dissociated ERP effects of misses from the ERP correlates of context familiarity, which was also replicated, and thus adds to the considerable line of evidence weighing against an interpretation of the present findings for context familiarity as an LPN.

A related problematic issue for an LPN account of the present findings is that the LPN is well-researched but poorly defined for its functional significance. The LPN is operationalized broadly as either a “reconstructing of the past” or a “controlled search process”, which, colloquially, only means “trying to remember”, which should occur on every memory retrieval trial to one extent or another and is difficult to reconcile with the present pattern of results. Due to the poor operational definitions available to the LPN, it is possible that the LPN findings from at least some other studies could instead be reflecting context familiarity (James et al., 2016). That is, there is also the other possibility that some instances of reported results interpreted as an LPN might have actually been misinterpreted and reflecting the effects of context familiarity, instead of an LPN.

For instance, another laboratory’s findings of a Cz-centric ERP negativity at approximately 1000 milliseconds occurring for source memory retrieval were interpreted as an LPN in that study (James et al., 2016) but their results’ characteristics could also be interpreted as being consistent with the current findings proposed to reflect context familiarity instead. The main, non-specific feature shared among the present findings and the LPN is that each is a negative-going ERP effect- but that in itself is not a very helpful descriptor, just as many other positive-ERP effects are not assumed to reflect the same thing just by virtue of their being a positive difference. In sum, the present findings are clear that the broad central negativity observed in the present studies is evident to the specific condition of context familiarity and not reasonably ascribed to other negative-going ERP effects such as the LPN, nor cognitive operations associated with the LPN such as a controlled search process or reconstruction of the past, concurring with the prior observations of their differentiation (Addante, Ranganath, & Yonelinas, 2012; Mecklinger et al., 2016), respectively.

5.4.4. Are effects representing context familiarity driven by fluency or guessing?

It is also possible that the current results might reflect a form of guessing. In this view, effects may be reflecting conceptual implicit memory contributions to source memory, as has been seen in the form of guessing observed in prior studies of item recognition (Voss & Paller, 2010). However, ERP findings of guesses indicated that guess processing was occurring much earlier in time (200–400 ms and 600–700 ms) than what we found in the current study (800–1200 ms) (Voss & Paller, 2010), so the current findings do not bear those hallmarks of guessing. The fact that two explicit declarative responses (item, source) were observed for judgments in the condition is also inconsistent with guessing, too, as is the above-chance performance of the responses. While it’s possible people could have been successfully guessing in these item + source judgments, this possibility becomes increasingly less likely when considering that participants were given the option of saying they ‘don’t know’ the source in our task’s paradigm (i.e.: ‘source unknown’, Figure 1), and which are thus excluded from the condition.

Furthermore, if results were influenced by guesses, then it is notable that the purportedly accurate ‘guesses’ would actually be driven by an implicit memory process (i.e.: ‘recognition without awareness’). We conducted several direct analyses to assess if context familiarity ERP activity resembled the known hallmark signs of other implicit processing (i.e., misses) that would have been presumably supporting the guess behavior, but instead found that it was significantly different in both time and scalp location from implicit misses. That finding is further bolstered by another, different implicit activity recently observed in the same dataset for a purer form of implicit memory (IMAP) (Addante et al., 2023) that was found instead to manifest in right parietal regions in positive-going ERP effects from approximately 400 to 1000 ms (not the fronto-central sites observed here from 800 to 1200 ms in negative-going ERP effects). Prior work discussing the requirements for a condition being ‘recognition without awareness has established the criteria that information about context should be absent from the memory (Craik et al., 2015) – which also rules out the condition of interest for context familiarity in the present study. Taken together, the evidence weighs against interpretations of context familiarity being primarily the result of processes associated with guessing from underlying implicit memory signals.

5.4.5. Are the negative ERP effects reflecting implicit fluency?

It is probable that, much like many explicit memory processes, context familiarity is also supported by forms of implicit processes (i.e., repetition priming, perceptual priming, contextual fluency) that then give rise to the explicit memory judgments, as has been found in other studies of item familiarity and cued recall (Ozubko et al., 2021; Voss et al., 2010, 2012). Familiarity of context may naturally represent various forms of fluency in processing information which has linked negative-going ERP effects for old items to repetition and perceptual fluency (Bruett & Leynes, 2015; Leynes & Zish, 2012; Leynes & Mok, 2020; Leynes & Addante, 2016; Leynes et al., 2016). As such, it is possible that these explicit response trials for context familiarity could be supported by an underlying fluency heuristic that leads one to accurate recognition and source judgments.

According to this explanation, the negative ERP effect likely reflects the fluency heuristic. This effect might purportedly be obscured during high-confidence trials by the predominance of positive LPC effects associated with recollection, which could override the presumed baseline negative ERP activity postulated to occur in recollection conditions as well. While such an explanation is hypothetically conceivable in this familiar context, where various possibilities exist, it remains speculative and lacks parsimony with the data. This is because it fails to explain why conditions of item familiarity did not display similar negative ERPs in the apparent absence of recollection. Additionally, the effects of context familiarity manifest later in time (800–1200 ms) than the very early latencies (200–400 ms) where fluency is often found exerting its effects (Leynes & Addante, 2016; Mecklinger & Bader, 2020). Through the series of systematic analyses here we demonstrate that there are still core distinctions between the kinds of familiarity represented by items and context, both of which surely are multiply determined by earlier sensory and contributory processes. We believe that based upon the combination of converging data and the divergent findings for studies measuring fluency at much earlier latencies, locations, and conditions, the present findings cannot be explained by mere fluency.

5.5 Future Directions.

There are several next steps which can be envisioned for future research based upon the present findings. For instance, intracranial measures of subsections of the medial temporal lobe regions such as the parahippocampal cortex and perirhinal cortex (Fernández et al., 1999; Goyal et al., 2020; Staresina, Cooper, et al., 2013; Staresina et al., 2012; Suthana et al., 2011; Suthana & Fried, 2012; Suthana, Donix, et al., 2015; Watrous & Buchanan, 2020; Watrous, Deuker, et al., 2015; Watrous et al., 2018) as hypothesis-driven tests of model predictions from the present work (see Section 5.5 below, Fig. 8). Such kinds of studies can include both invasive and non-invasive stimulation approaches to both enhance (Addante, 2021; Ezzyat et al., 2017; Ezzyat et al., 2018; Hanslmayr et al., 2019; Inman et al., 2018; Jacobs et al., 2016; Khambhati et al., 2019; Roberts et al., 2018; Suthana et al., 2012; Suthana, Parikshak, et al., 2015; Titiz et al., 2017) and disrupt (Jacobs et al., 2016) memory operations of different episodic processes, and assessing their relevance in real-world environments/context (Aghajan et al., 2017; Maoz et al., 2023; Stangl et al., 2023; Topalovic et al., 2023). Other potential future approaches include exploring for neuropsychological dissociations among various types of MTL patients (Yonelinas et al., 2002) or those with specified focal lesions of MTL subregions (Addante, 2015; Addante, Ranganath, Olichney, et al., 2012; Bowles et al., 2007), or electrophysiological searching for differential oscillatory bandwidths that may possibly be selective for each process (Addante et al., 2011; Duzel et al., 2005; Fell & Axmacher, 2011; Fell et al., 2001; Fell et al., 2008; Fell et al., 2011; Fellner et al., 2019; Goyal et al., 2020; Griffiths et al., 2016; Hanslmayr et al., 2019; Hanslmayr et al., 2009; Hanslmayr & Staudigl, 2014; Jacobs et al., 2006; Klimesch et al., 2000; Klimesch et al., 2001; Klimesch et al., 2006; Mazaheri & Jensen, 2010; Mazaheri et al., 2018; van Diepen & Mazaheri, 2018; van Dijk et al., 2010; Watrous & Buchanan, 2020; Watrous, Deuker, et al., 2015; Watrous, Fell, et al., 2015; Watrous et al., 2018).

In pursuing future studies, note that the specialized condition used in the present study to measure context familiarity (low-confidence recognition hits accompanied with accurate source memory) does not necessarily mean that it is the only potential way to measure context familiarity. Rather, it is likely just one of many possible approaches to capturing an elusive cognitive process. The findings here suggest the importance of including multiple measures of memory into experimental protocols, as it permitted the leverage needed to identify the combined conditions needed to measure context familiarity in the present experiments. We would expect future studies to find context familiarity to accordingly be measurable in other ways pending properly sensitive experimental designs that do not conflate it with recollection or item-familiarity or guessing. For example, we surmise that context familiarity could be found in future studies using different paradigms that systematically vary repetition of context while holding items constant and vice versa, similar to repetition paradigms used in prior fMRI studies (Diana et al., 2010, 2012; Hannula et al., 2013; O’Neill & Diana, 2017; Ranganath et al., 2004) that found results similar to the present data. Our measure here was a combination of multiple responses (memory confidence scales) across multiple measures (item and source memory judgments), and thus reflects the sophisticated paradigmatic sensitivity that will likely be needed for follow-up explorations. Towards this end, studies should use an unequal confidence scale for source memory judgments to permit a ‘source unknown’ response in order to avoid the guesses that are inherently integrated in balanced confidence scales (Coane & Umanath, 2019; Liu et al., 2007; Malpass & Devine, 1981; Voss & Paller, 2010); that is, be sure to move from a commonly-used 6-point confidence scale(Ranganath et al., 2004) to a 5-point confidence scale (Addante et al., 2023; Addante, Ranganath, & Yonelinas, 2012; Addante et al., 2011; Muller et al., 2021). That is, including a ‘source unknown’ option into source memory tests should be essential for future researchers to avoid contaminating source memory responses with instances of guessing or the source unknown.

Another direction for future studies could be to assess the possibility of delayed recollection supporting the condition identified here as context familiarity. One way to do that could be by studying response locked analyses of ERP or oscillatory analyses (Griffiths et al., 2016; Mazaheri, 2022; Mazaheri & Jensen, 2010; Mazaheri et al., 2018; van Diepen & Mazaheri, 2018; van Dijk et al., 2010), as opposed to the stimulus-locked events studied here for the purposes of consistency to the literature of known EPR effects of episodic memory (Curran, 2000; Friedman, 2013; Kwon et al., 2023; Mecklinger, 2006; Mecklinger & Jäger, 2009; Ozubko et al., 2021; Rugg & Curran, 2007). But it is important to also note that such studies of response-locked activity in this paradigm risk being confounded or conflated in their baseline period due to the preceding activity of the item recognition prompt/decision in in the present paradigm- so future studies would need to be very carefully designed to ensure those and other factors are properly disambiguated.

5.6 Implications for models of memory.

The medial temporal lobe region of the parahippocampus has been modeled by several frameworks to be a neural substrate supporting contextual representations of episodic memory (Aminoff et al., 2013; Diana et al., 2007; Eichenbaum et al., 2007; Meeter et al., 2005; Montaldi & Mayes, 2010) (Figure 11).This is based upon neuroimaging studies identifying activations of that region for contextual processing (Diana et al., 2012, 2013; Ranganath et al., 2004), while ascribing item familiarity to the adjacent perirhinal cortex (Diana et al., 2010, 2012; Hannula et al., 2013; Haskins et al., 2008). This has been paralleled by findings from studies using intracranial recordings showing temporal dissociations of familiarity and recollection processing amid the parahippocampal gyrus and hippocampus, respectively (Fernández et al., 1999; Linde-Domingo et al., 2019; Staresina et al., 2012; Staresina, Fell, et al., 2013; Staresina & Wimber, 2019), which has been similarly observed using functional neuroimaging methods (Staresina, Cooper, et al., 2013).

Neuropsychologically, converging reports of the perirhinal cortex role in item familiarity come from a rare clinical patient, NB, with an epileptic resection lesion specific to the perirhinal cortex that left the parahippocampal cortex intact (Köhler & Martin, 2020). This patient was deficient in familiarity for items (words) but showed intact abilities of familiarity for pictures and faces (Aggleton & Brown, 2006; Bowles et al., 2007; Bowles et al., 2010; Martin et al., 2011), which we consider here as broadly representing kinds of context in episodic events and this could be seen as performance consistent with the patient’s intact parahippocampus service of context familiarity (though it is possible that faces and/or picture stimuli might have potentially instead be treated by that patient as an item in certain studies with other contexts, or alternatively the face/picture memory performance might have potentially been supported by the other hemisphere’s intact perirhinal cortex). Similar reports of familiarity for contexts in the absence of object familiarity or recollection have emerged from neuropsychological (Levine et al., 1999) and normative studies using visual scenes of mazes, whereby it is the spatial layout of the scene that is experienced as familiarity and not the items or objects (Cleary et al., 2012; Cleary, K. McNeely-White, et al., 2021; Cleary, J. Neisser, et al., 2021; A. M. Cleary et al., 2021).

Together with the present studies, these findings converge to suggest an updated framework for the organization of long-term episodic memory from the canonical declarative memory model used ubiquitously throughout textbooks. We propose an updated model to conceptualize the organization of memory processes (Figure 8), which is unique in being specifically physiologically-informed, as prior models of memory have instead been either informed by cognitive psychology or neuroanatomical/neuroimaging studies (Cohen & Eichenbaum, 1993; Diana et al., 2007; Eichenbaum et al., 2007; Milner et al., 1998; Squire, 2009; Squire et al., 2007; Squire & Zola, 1997; Wixted & Squire, 2010, 2011b; Yonelinas et al., 2019). The new trivariate model provides an updated representation of the canonical declarative memory framework that is inclusive of findings from physiological data indicating that recollection and familiarity are distinct processes of episodic memory (Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012; Rugg & Curran, 2007; Rugg, Mark, et al., 1998), and that context- and item-familiarity are also dissociable processes in episodic memory (Figures 4, 5, 6, 7, (Addante, Ranganath, & Yonelinas, 2012; Hicks et al., 2002)). This physiologically-informed model also depicts the emerging views that differential forms of implicit fluency contribute to explicitly-aware familiarity (Bruett & Leynes, 2015; Leynes & Zish, 2012; Leynes et al., 2023; Leynes & Mok, 2020; Leynes & Upadhyay, 2022; Leynes et al., 2017; Mecklinger & Bader, 2020; Stróżak et al., 2020) and that other forms of implicit memory depend upon the hippocampus (Addante, 2015; Addante et al., 2023; Duss et al., 2014; Hannula & Greene, 2012; Hannula et al., 2017; Henke, 2010; Reber, 2013; Reber et al., 2015; Reber et al., 2012).

The episodic process of recollection is modeled as a threshold process based upon physiological findings in the current and former data, since the LPC effects were found to be limited to high-confidence responses and did not extend to low-confidence responses even if source memory is retrieved (Addante, Ranganath, & Yonelinas, 2012); for other related cognitive frameworks see (Parks, 2007; Parks et al., 2011; Parks & Yonelinas, 2009). Context familiarity and item familiarity are each depicted as distinct continuous processes of signal detection for relatively old and new information distributions (Addante, Ranganath, & Yonelinas, 2012; Mruczek & Sheinberg, 2007; Woodruff et al., 2006; Yu & Rugg, 2010), based upon the current and prior data, and as opposed to the traditional model of them both merged within a single continuous process of general familiarity. The model also updates the understanding of the parahippocampal and perirhinal cortices as processing differential familiarity signals (Figure 11B), as opposed to previous models depicting them to operate amid a single familiarity signal along both regions of the larger parahippocampal gyrus (Diana et al., 2007; Eichenbaum et al., 2007; Meeter et al., 2005; Ranganath, 2010; Ranganath & Ritchey, 2012; Yonelinas et al., 2019).

A cognitive framework can be built to understand how the condition of context familiarity may arise or manifest. This account both echoes and extends models provided by prior work investigating the role of context in episodic memory (Addante, Ranganath, & Yonelinas, 2012; Hannula et al., 2013; Montaldi & Mayes, 2010; Reagh & Ranganath, 2018), while integrating fluency heuristics (Leynes & Zish, 2012; Leynes & Mok, 2020; Mecklinger & Bader, 2020). It has been proposed that people use a fluency heuristic to make recognition judgments (Leynes & Zish, 2012; Mecklinger & Bader, 2020), and the current findings extend that by suggesting that a fluency heuristic can also be used to support accurate source memory judgments, too. This would be based upon an underlying fluency for context, as opposed to fluency for items or conceptual/perceptual dimensions. By our account, as a retrieval probe is first shown on a screen by itself, the item doesn’t at first seem familiar (i.e., no FN400 for item familiarity from ~300–600ms that we observed). The context, however, may seem familiar, even if they do not recollect the details of either the item or how the context was bound with other details from the prior episode. Akin to how item familiarity can be derived from underlying fluency of the stimuli, so too would context familiarity be facilitated by a separate underlying fluency for the context. If the context does not initially seem familiar from ~300–600 ms, then in the ensuing time (600–900 ms, LPC) the participant would continuing to mentally look back in time for any context in which to place the item being shown on the screen (e.g. ‘where might I have seen that before?’), seeking the pattern completion process to autocomplete a recollection of the relevant details from the prior event. During that time (600–900 ms), that autocompletion process of binding any existing neural signals of context together with the presented item fails (i.e.: our observations of an absence of evidence for the LPC correlate of recollection), consistent with the slower response times observed for these item judgments.

As time continues (800–1200 ms) the person continues to assess their memory for any connections of details, something about the stimuli remains diagnostic for prior exposure (the broad central negativity effect emerges to reliably differentiate from correct rejections and item familiarity), which could be either the explicit familiarity of a contextual detail such as the encoding source or the implicit sense of its contextual fluency for associated information about the item. The successful retrieval of some familiar contextual information would be enough to logically deduce that the item must have been old (Whittlesea & Williams, 2000; Whittlesea & Williams, 2001). That is, one can then draw upon that awareness that the context seems familiar to derive the logical inference that the item is likely old, akin to how other studies have found item familiarity can lead to false alarms in associative recognition by similar logical deductions (Addante et al., 2015; Jones & Bartlett, 2009; Jones & Jacoby, 2005; Jones et al., 2001; Lampinen et al., 2005). Ultimately, this basis of a retrieved prior context is sufficient to support an accurate hit in item recognition, albeit with low confidence because the item itself is not remembered directly, while the correct source is identified in the ensuing judgment based upon that retrieved familiarity of the context of its prior exposure (Whittlesea & Williams, 2001). The item is thus accurately recognized due to the context seeming familiar: context familiarity.

Importantly, our trivariate account provides an improved precision in characterizing human episodic memory, and as such could help to better understand disparate findings from memory studies and neuropsychological patient impairments that occasionally arise in the literature. That is, if prior studies had been conflating distinct familiarity processes for items and context, and conflating context familiarity with instances of recollection, when occurring in differential proportions and under differential experimental conditions, it could easily produce disparate clinical findings that have been reported about source memory, context, recollection, and neuropsychological impairments thereof (Addante, Ranganath, Olichney, et al., 2012; Bowles et al., 2007; Bowles et al., 2010; Eichenbaum et al., 2007; Manns et al., 2003; Martin et al., 2011; Squire et al., 2007; Wais et al., 2006; Wixted & Squire, 2004, 2010, 2011a; Yonelinas et al., 2010; Yonelinas et al., 2002; Yonelinas et al., 2004; Yonelinas et al., 2019). If we are now able to delineate between these three distinct but overlapping processes of memory, it could help illuminate a better understanding of the nature of memory impairments and lead to better precision in types of amnesia diagnoses.

The idea that there can be different kinds of familiarity for items and for contexts has been previously mentioned by several others (Addante, 2012; Addante, Ranganath, & Yonelinas, 2012; Montaldi & Mayes, 2010; Whittlesea & Williams, 2000). Whereas prior work established the premise that there can be separable familiarity of items and familiarity of contexts (Montaldi & Mayes, 2010; Whittlesea & Williams, 2000), they nevertheless proposed that each was stemming from the same singular memory process of underlying familiarity. For instance, the Discrepancy Attribution Hypothesis offered that people can produce either piece of mnemonic information separately (Whittlesea & Williams, 2000), stating:

“contextual information can also cause the person to experience a feeling of familiarity: encountering a stimulus in a context that is unusual for that stimulus (e.g., a clerk on the bus) can cause this.” However, and importantly, they go on to propose that “the use of contextual information to perform judgments about the past is not qualitatively different from the use of identity [emphasis added]” (page 561) (Whittlesea & Williams, 2000).

Here, we put forth a fundamentally different view: that production of familiar context does differ qualitatively from the production of familiar item information and provide converging lines of support that the familiarity of items and familiarity of contexts are evidently operating as separable independent neurocognitive processes.

5.7 Conclusions.

The field of memory research is replete with complexities in memory behaviors in normative and clinical patient populations that are difficult for existing theories to fully account for. We show here that the common measure of source memory can be supported by a disparate process that is independent of both recollection and item-familiarity. The memory condition of context familiarity was found to be distinct from item familiarity, in addition to previously identified dissociations from recollection. This conclusion was evident from triple dissociations observed in both behavioral and physiological measures. Moreover, the differences in item-familiarity and context familiarity reflected differential neural processing at both the individual subject level as well as trial-wise differences in memory processing that was identified within-subjects and were replicated across several independent studies. These findings cannot be accounted for by relying upon using only two processes of episodic memory for interpreting the results (Yonelinas, 1999; A. P. Yonelinas, 2001). Item familiarity and context familiarity are therefore seen as fundamentally different kinds of familiarity processes independently operating in service of episodic recognition (Allan et al., 1998; Mecklinger & Jäger, 2009; Rugg, 2009; Rugg & Coles, 1995; Rugg, Mark, et al., 1998; Shallice, 1988): one process for items and another for context, and each distinct from recollection. This extends traditional dual process models of recognition to reflect that context familiarity constitutes a third process of episodic memory.

Supplementary Material

Supplement 1

Supplement 2

Supplement 3

Acknowledgments:

Funding support of this work came from the National Institutes of Health (NIH), which included grants 2 L30 NS112849–03, 2 L30 NS112849–02, and 1 L30 NS112849–01 awarded to RJA from the National Institute of Neurological Disorders and Stroke (NINDS), from Florida Institute of Technology (FIT) grants to RJA from the Institutional Research Incentive (IRI) program and, via direct funding from the FIT Provosts Office of Dr. Marco Carvalho. DLD received support from grants R01 NS124585 from the NIH/NINDS and A1321808 from Medtronic, Inc. JPC and RJA received support from NASA/FL Space Consortium Grant. Authors thank Alana Muller, Lindsey Sirianni, Anderson Wilder, and Cristo Rey for research assistance. Publication of this article was funded in part by the Open Access Subvention Fund and the John H. Evans Library at Florida Institute of Technology.

Data Availability and Open Practices Statement:

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request, and the primary data set is available via the previously published study using the same data(Addante et al., 2023), freely available under the terms of the GNU General Public License at https://github.com/IMAP-Lab/IMAP-Lab-IMAP-Implicit-Memory-Automated-Pipeline.

Figure 1. Memory test schematic and combinations of conditions analyzed.

Panel A (left): For each of 216 items studied and 108 new item lures, participants responded to a memory test that first asked them to rate their recognition confidence on a scale from 1 to 5 for if they thought the word on the screen was an old item from the study phase or if they thought it was a new item. Immediately following the item recognition judgment, participants were asked to indicate their source memory, e.g. if they remembered which study task the item was encountered in during the study phase. If they could not remember, they were instructed to then press the ‘3’ key to indicate that the source was unknown. Panel B (right): schematic identifying the response profile for the three main conditions of interest analyzed in the current study. Green: item recognition hits collapsed across confidence levels, and which received responses of having no source memory, were operationalized as ‘item familiarity’ (iFam.(Addante et al., 2023; Addante et al., 2011)). Blue: low confidence recognition hits (responses of ‘4’) that received correct source memory judgments were operationalized as ‘context familiarity’ (cFam. (Addante, Ranganath, & Yonelinas, 2012)). Red: high confidence recognition hits (responses of ‘5’) that received correct source memory judgments were operationalized as ‘recollection’ (Rec. (Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012; Yonelinas, 2002; Yonelinas et al., 2010; Yonelinas et al., 2019)). Right: Representative ERP results of context familiarity and recollection are depicted from the original findings (Addante, Ranganath, & Yonelinas, 2012) (reproduced with permission of the owner), shown as topographic maps of each condition plotted in comparison to correct rejections during the latency of 600 to 800 milliseconds in the memory epoch, with color scale depicting difference amplitudes (from −2 to 2 μV) of each condition as compared to correct rejections.

Figure 2. Reaction speeds for item recognition and source memory judgments.

Mean times in milliseconds are shown in black with standard error of the mean for responses of item recognition (left) and source memory judgments (right), for three different conditions of memory response patterns. Green: item familiarity (iFam.), operationalized as retrieval of recognition hits that did not have source memory (received responses of ‘source unknown’). Blue: context familiarity (cFam.), operationalized as retrieval of low-confidence item recognition that had correct source memory responses. Red: recollection (Rec.), operationalized as retrieval of high-confidence item recognition that had correct source memory responses. The boxplots demonstrate the median value, notches (95% confidence interval), and interquartile range (IQR) with lower and upper whiskers (Q1–1.5*IQR; Q3+1.5*IQR). The violin plots show density curves centered around the median value. Plots were generated using the scripts created by (Yeh & Koen, 2023).

Figure 3. Old-New Effects of Recognition Memory.

Top panel: Topographic maps depicting memory effects of the subtraction of hits minus correct rejections; color scale of topographic maps denotes amplitude differences from −2 to 2 mu-volts. Maps are plotted for each of the three time latencies of 400 to 600 milliseconds (ms), 600 to 900 ms, and 800 to 1200 ms that have been associated with the frontal negative old-new effect (FN400), late parietal component (LPC), and later broad central negativity (BCN), respectively. Bottom panel: Event-related potential (ERP) amplitudes (μV ) of the individual conditions of recognition hits (dashed line) and correct rejections (solid line) plotted through the 1200 ms epoch and beginning with a −200 ms baseline period. Shaded areas depict the standard error or the mean for each condition.

Figure 4. Three physiological effects of episodic memory retrieval.

Left panels: Topography maps are shown for each memory condition as compared to correct rejections in Experiment 1: (top) high-confident item recognition judgments that also had correct source memory retrieval; (middle) low-confident item recognition judgments of that also had correct source memory retrieval; and (bottom) recognition hits (collapsed across both high and low confidence) that did not have source memory retrieved (received responses of ‘source unknown’). Color scale of topographic maps denotes amplitude differences for each condition as compared to correct rejections, ranging from −2 to 2 mu-volts. Right panels: Event-related potential (ERP) amplitudes (μV) from representative sites of maximal activation for each of the respective effects during the memory retrieval epoch of −200 to 1200 milliseconds (ms); shaded areas for each plot represent the standard error of the mean.

Figure 5. Within-subjects ERP differences of context familiarity and item familiarity.

Results depict findings from Experiment 1. Panel A: within-subject effects for participants who shared pairwise response trials in both conditions of item familiarity and context familiarity (N = 11; note: N = subjects, n = trials, see Methods section 2.3.2). Effects were evident at 300–500 ms and distributed broadly across the scalp while being maximal over central midline site Cz. Topographic maps of difference waves for effects results from the subtraction of item familiarity from context familiarity conditions, respectively, with color scale depicting difference amplitudes from −2 to 2 mu-volts; white boxes denote regions of peak activity and targeted analyses; time intervals for each topography map is noted beneath it in milliseconds (ms). Shaded areas of event-related potentials (ERPs) depict the standard error or the mean for each condition. Panel B: Results from a sample of participants (N = 25) meeting a lower inclusion criterion for signal-to-noise ratio (n = 5 artifact-free ERP trials pairwise). Panel C: Control analysis of the same conditions but while holding constant the variable of item recognition confidence (depicting results from only the low confidence ratings of ‘4’ for each condition, and excluding the high confidence ‘5’ responses from item familiarity, N = 3, n > 11).

Figure 6. Between-subjects ERP differences of context familiarity and item familiarity.

Panel A, top: Between-subject effects in independent sub-groups of participants who had response trials in one condition but not the other (i.e.: mutually exclusive N = 26 total; item familiarity: N = 9, context familiarity: N = 17), from Experiment 1 (Addante et al., 2023; Muller et al., 2021). Event-related potentials (ERPs) are shown from representative midline central site Cz; Shaded areas of event-related potentials (ERPs) depict the standard error or the mean for each condition. Bottom: range-normalized topographic maps of individual conditions from 300 to 500 ms, with color scale depicting voltage on the numerical scales indicated. Panel B: Between-subjects effects from participants of Experiment 2 (Addante, Ranganath, & Yonelinas, 2012; Addante et al., 2011), (N = 24 total; item familiarity: N = 13, context familiarity: N = 11).

Figure 7. Independent Replication of ERP Differences for Context- and Item-Familiarity.

Within-subject effects for participants sharing pairwise response trials in both conditions of item familiarity and context familiarity (N = 20) in Experiment 2 data aggregated from previously published studies (Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012; Addante et al., 2011) (see Method). Panel A, top row: topographic difference maps of the resulting within-subjects difference between context familiarity minus item familiarity, in time intervals shown in milliseconds (ms). Color scale depicts difference amplitudes from −2 to 2 mu-volts, white boxes denote regions of targeted analyses revealing statistically significant effects. ERPs from representative sites Cz (mid-central), F3 (left frontal), and F4 (right frontal); shaded areas depict the standard error or the mean for each condition. Panel B: Same contrast as depicted as in Panel A but excluding high confidence recognition responses from the item familiarity conditions (compare to Figure 5C), to match conditions on item recognition confidence levels (N = 12).

Figure 8. Physiological effects of episodic memory retrieval conditions.

Left panels: Topography maps are shown for each memory condition as compared to correct rejections in Experiment 2, comprising data from three previously published studies utilizing the same paradigm as the present study (Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012; Addante et al., 2011). (top) High-confident item recognition judgments that also had correct source memory retrieval; (middle) low-confident item recognition judgments of that also had correct source memory retrieval; and (bottom) recognition hits (collapsed across both high and low confidence) that did not have source memory retrieved (received responses of ‘source unknown’). Color scale of topographic maps denotes amplitude differences for each condition as compared to correct rejections, ranging from −2 to 2 mu-volts. Right panels: Event-related potential (ERP) amplitudes (μV) from representative sites of maximal activation for each of the respective effects during the memory retrieval epoch of −200 to 1200 milliseconds (ms); shaded areas for each plot represent the standard error of the mean.

Figure 9. Old-New Effects for Recognition Memory in Experiment 2.

Data for Experiment 2 was aggregated from three previously published studies (Addante, Ranganath, Olichney, et al., 2012; Addante, Ranganath, & Yonelinas, 2012; Addante et al., 2011) (N = 56). Top row: topographic difference maps of the resulting difference between hits minus correct rejections, plotted for each of the three time latencies of 400 to 600 milliseconds (ms), 600 to 900 ms, and 1000 to 1400 ms that have been associated with the FN400, late parietal component (LPC), and late broad central negativity (BCN), respectively. Color scale depicting difference amplitudes from −2 to 2 mu-volts, white boxes denote regions of targeted analyses revealing statistically significant effects. Bottom panel: Event-related potential (ERP) amplitudes (μV) of individual conditions of recognition hits (dashed line) and correct rejections (solid line) plotted through the 1500 ms epoch and beginning with a −200 ms baseline period. Shaded areas depict the standard error or the mean for each condition.

Figure 10. Physiological effects of context familiarity versus recognition misses.

Panel A: Top: topographic maps of the difference map for activity of context familiarity minus activity for recognition misses (N = 27). Maps are plotted for each time interval of 200–400 milliseconds (ms), 400 to 600, 600 to 900 ms, and 800 to 1200 ms; color scale of topographic maps denotes amplitude differences from −2 to 2 mu-volts. Event-related potential (ERP) amplitudes (μV) for each condition plotted for representative sites of maximal activation (left frontal, F3, and centro-parietal Cp1), through an epoch of −200 to 1200 ms of a memory retrieval test. Shaded areas depict the standard error or the mean for each condition. Panel B: Replication data from Experiment 2. Data from three previously published studies (N = 27). ERPs for each condition plotted for representative sites from original analysis (Cp1) and of maximal activation (right parietal Cp5, 200–400 ms), through an epoch of −200 to 1200 ms of a memory retrieval test.

Figure 11. Physiologically informed model of long-term memory processes.

Top, A: The trivariate model presents an updated representation of the canonical declarative memory framework that is inclusive of physiological findings from physiological data indicating that recollection and familiarity are distinct processes of episodic memory (Addante, Ranganath, Olichney, et al., 2012; Rugg, Mark, et al., 1998), and that context- and item-familiarity are also dissociable processes in episodic memory (Figures 4, 5, 6, 7). Model depicts the emerging views that differential forms of implicit fluency contribute to driving explicit familiarity (Mecklinger & Bader, 2020). A sample schematic of recollection shown models it as a threshold process based upon physiological findings in the current and former data indicating that is limited to high-confidence responses and does not extend to low-confidence responses even if source memory is retrieved (Addante, Ranganath, & Yonelinas, 2012; Parks, 2007; Yonelinas & Parks, 2007). Context familiarity and item familiarity are each depicted as continuous processes of signal detection, with representative distributions of old and new information each. Bottom, B: Anatomical and conceptual representation of medial temporal lobe processing of mnemonic information, based upon the ‘binding-items-in-context (BIC)’ model (Diana et al., 2007; Eichenbaum et al., 2007), updated here to include differential familiarity processes for items and context in the perirhinal and parahippocampal cortices, respectively. Note that the model presents a simplified representation of the breadth of complex memory operations and neuroanatomical substrates.

Table 1. Proportions of overall response patterns for recognition hits and source judgment combinations for old items during the memory test.

	Source Incorrect		Source Correct		
					
Response Proportions	Sure	Unsure	Unknown	Unsure	Sure	Total	
	
Per total Old Items							
 Item 4: Unsure ‘Old’	< .01	.05	.02	.06	.02	.16	
 Item 5: ‘Sure Old’	.12	.07	.04	.11	.30	.64	
 Total	.14	.14	.22	.18	.33		
Per Confidence Level							
 Item 4: Unsure ‘Old’	.05	.32	.14	.38	.10	1.00	
 Item 5: ‘Sure Old’	.19	.11	.07	.17	.47	1.00	
Per Total Hits							
 Item 4: Unsure ‘Old’	.01	.07	.03	.08	.02	.20	
 Item 5: ‘Sure Old’	.15	.09	.05	.13	.37	.80	
 Total	.16	.15	.08	.21	.39		
Each participant (N = 54) had 216 trials of old items. Left hand column represents item recognition confidence responses, and successive columns denote source memory responses. Top subsection provides the overall proportion of responses of high- and low-confidence recognition hits for each level of source judgment, out of all old trials (214 trials for each of N = 54 participants). Middle subsection indicates the proportion of responses for each source judgment per high- and low-confidence recognition hits; e.g. when subjects did have a recognition hit, it indicated what proportion of times it ended up in each of the source memory categories of interest for the current investigation. Bottom subsection indicates proportion of responses for each source memory judgment out of total recognition hits (collapsing across high and low confidence levels).

Table 2. Descriptive Statistics of Response Time (ms) for Item and Source Memory Judgements.

	Item Memory	Source Memory	
			
Condition	M	SD	SE	M	SD	SE	
	
Recollection	1810	422	168	2596	1126	182	
Context Familiarity	2710	962	156	2075	968	157	
Item Familiarity	2361	684	111	2612	1577	255	
Time is reported in milliseconds, and data represents the sample of participants with pairwise observations included in the reported results (N = 38).

Significance statement/Summary:

Memory for past events is widely believed to operate through two different processes: one called ‘recollection’ when retrieving confident, specific details of a memory, and another called ‘familiarity’ when only having an unsure but conscious awareness that an item was experienced before. When people successfully retrieve details such as the source or context of a prior event, it has been assumed to reflect recollection. We demonstrate that familiarity of context is functionally distinct from familiarity of items and recollection and offer a new trivariate model. The three memory responses were differentiated across multiple behavioral and brain wave measures. What has traditionally been thought to be two kinds of memory processes are actually three, becoming evident when using sensitive enough multi-measures. Results are independently replicated across studies from different labs. These data reveal that context familiarity is a third process of human episodic memory.

Code Availability Statement: Data collection was performed using stimulus delivery software Presentation, Neurobehavioral Systems, Inc., and is available from corresponding author upon reasonable request. Analysis codes used for the current study are available from the corresponding author upon reasonable request.

1 Authors acknowledge a previous Reviewer for suggesting consideration of this alternative approach.

2 We thank a previous Reviewer for pointing this out while suggesting post-hoc analysis of power instead.

3 Authors acknowledge Editor Dr. Marike Shiffer at Nature Communications: Psychology for requiring this analysis, for which Reviews are provided with Author Responses on the pre-print server Biorxiv as Supplemental Material 1–3, https://www.biorxiv.org/content/10.1101/2024.07.15.603640v3.supplementary-material.

4 Hoenig & Heisel (2022): “Advocates of such post-experiment power calculations claim the calculations should be used to aid in the interpretation of the experimental results. This approach, which appears in various forms, is fundamentally flawed. We document that the problem is extensive and present arguments to demonstrate the flaw in the logic.” Heinsberg & Weeks (2022): “Post hoc power estimates are often requested by reviewers and/or performed by researchers after a study has been conducted. [But]…post hoc power should not be used. …post hoc power calculations are misleading and simply not informative for data interpretation. As such, we encourage authors and peer-reviewers to avoid using or requesting post hoc power calculations.”. Zhang et al (2019):“What is interesting is that some journals also ask for power analysis for the study data that were already analysed and reported in a manuscript before considering its publication. … As such post hoc power analysis is conceptually flawed, concerns have been raised over the years. Despite these warnings, some journals continue to ask for such information and use it as part of their decision process for manuscript publications.…Power analysis is an indispensable component of planning clinical research studies. However, when used to indicate power for outcomes already observed, it is not only conceptually flawed but also analytically misleading. … In general, post hoc power analyses do not provide sensible results.”

5 Authors thank an anonymous Reviewer for suggesting this analysis

6 Similar results were also found when assessing the time windows used in the present study (Figure 3) of 400 to 600 ms, 600 to 900 ms, and 800 to 1200 ms: with main effects of memory F(1, 55) = 5.144, p = .027, n2p =.086, site (F(1.68, 55) = 7.70, p = .002, n2p = .123), time (F1.45, 55) = 44.797, p < .001, n2p = .449), and significant interactions of memory x time (F(1.41, 55) = 26.34, p <.001, n2p =.324), site x time (F1.56, 55) = 25.52, p <.001, n2p = .317), and a 3-way interaction of memory x site x time (F2.78, 55) = 4.66, p = .004, n2p = .078).

7 Authors thank a Reviewer for suggesting this possibility.

8 Authors thank a Reviewer for suggesting this possibility
==== Refs
References

Aczel B. , Hoekstra R. , Gelman A. , Wagenmakers E. J. , Klugkist I. G. , Rouder J. N. , Vandekerckhove J. , Lee M. D. , Morey R. D. , Vanpaemel W. , Dienes Z. , & van Ravenzwaaij D. (2020). Discussion points for Bayesian inference. Nat Hum Behav, 4 (6 ), 561–563. 10.1038/s41562-019-0807-z 31988442
Addante R. J. (2012). The electrophysiological and neuropsychological organization of long term memory (Publication Number AAI3499396) APA PsycInfo®. https://portal.lib.fit.edu/login?url=https://www.proquest.com/dissertations-theses/electrophysiological-neuropsychological/docview/1324551493/se-2?accountid=27313 https://portal.lib.fit.edu/login?url=https://resolver.ebscohost.com/openurl?sid=&genre=dissertations&issn=&ISBN=978-1-267-23814-6&volume=&issue=&date=2013&spage=&pages=&title=&atitle=The+electrophysiological+and+neuropsychological+organization+of+long+term+memory&aulast=Addante%2C+Richard+James%20&id=DOI:
Addante R. J. (2015). A critical role of the human hippocampus in an electrophysiological measure of implicit memory. Neuroimage, 109 (0 ), 515–528. 10.1016/j.neuroimage.2014.12.069 25562828
Addante R. J. , Clise E. , Waechter R. , Bengson J. , Drane D. L. , & Perez-Caban J. (2024). A third kind of episodic memory: Context familiarity is a distinct process from item familiarity and recollection. bioRxiv, 2024.2007.2015.603640. 10.1101/2024.07.15.603640
Addante R. J. , de Chastelaine M. , & Rugg M. D. (2015). Pre-stimulus neural activity predicts successful encoding of inter-item associations. Neuroimage, 105 (0 ), 21–31. 10.1016/j.neuroimage.2014.10.046 25450109
Addante R. J. , Lopez-Calderon J. , Allen N. , Luck C. , Muller A. , Sirianni L. , Inman C. S. , & Drane D. L. (2023). An ERP measure of non-conscious memory reveals dissociable implicit processes in human recognition using an open-source automated analytic pipeline. Psychophysiology, n/a (n/a ), e14334. 10.1111/psyp.14334
Addante R. J. , Ranganath C. , Olichney J. , & Yonelinas A. P. (2012). Neurophysiological evidence for a recollection impairment in amnesia patients that leaves familiarity intact [Research Support, N.I.H., Extramural]. Neuropsychologia, 50 (13 ), 3004–3014. 10.1016/j.neuropsychologia.2012.07.038 22898646
Addante R. J. , Ranganath C. , & Yonelinas A. P. (2012). Examining ERP correlates of recognition memory: evidence of accurate source recognition without recollection. Neuroimage, 62 (1 ), 439–450. 10.1016/j.neuroimage.2012.04.031 22548808
Addante R. J. , Watrous A. J. , Yonelinas A. P. , Ekstrom A. D. , & Ranganath C. (2011). Prestimulus theta activity predicts correct source memory retrieval. Proc Natl Acad Sci U S A, 108 (26 ), 10702–10707. 10.1073/pnas.1014528108 21670287
Addante R. Y. , M ; Valencia R ; Greenwood C ; Marino R. (2021). Boosting brain waves improves memory. Frontiers in Young Minds.
Aggleton J. P. , & Brown M. W. (2006). Interleaving brain systems for episodic and recognition memory. Trends Cogn Sci, 10 (10 ), 455–463. 10.1016/j.tics.2006.08.003 16935547
Aghajan Z. , Schuette P. , Fields T. A. , Tran M. E. , Siddiqui S. M. , Hasulak N. R. , Tcheng T. K. , Eliashiv D. , Mankin E. A. , Stern J. , Fried I. , & Suthana N. (2017). Theta Oscillations in the Human Medial Temporal Lobe during Real-World Ambulatory Movement. Current Biology, 27 (24 ), 3743–3751.e3743. 10.1016/j.cub.2017.10.062 29199073
Allan K. , Wilding E. L. , & Rugg M. D. (1998). Electrophysiological evidence for dissociable processes contributing to recollection. Acta Psychol (Amst), 98 (2–3 ), 231–252. http://www.ncbi.nlm.nih.gov/pubmed/9621832 9621832
Aminoff E. M. , Kveraga K. , & Bar M. (2013). The role of the parahippocampal cortex in cognition. Trends Cogn Sci, 17 (8 ), 379–390. 10.1016/j.tics.2013.06.009 23850264
Anderson M. L. , James J. R. , & Kirwan C. B. (2016). An event-related potential investigation of pattern separation and pattern completion processes. Cogn Neurosci, 1–15. 10.1080/17588928.2016.1195804
Atkinson R. C. , & Juola J. F. (1974). Search and decision processes in recognition memory. In Contemporary developments in mathematical psychology: I. Learning, memory and thinking. (pp. xiii, 299–xiii, 299). W. H. Freeman.
Bader R. , & Mecklinger A. (2017). Separating Event-related Potential Effects for Conceptual Fluency and Episodic Familiarity. Journal of Cognitive Neuroscience, 29 (8 ), 1402–1414. 10.1162/jocn_a_01131 28387586
Bader R. , Mecklinger A. , Hoppstadter M. , & Meyer P. (2010). Recognition memory for one-trial-unitized word pairs: evidence from event-related potentials. Neuroimage, 50 (2 ), 772–781. 10.1016/j.neuroimage.2009.12.100 20045471
Bader R. , Tarantini L. , & Mecklinger A. (2023). Task context dissociates the FN400 and the N400. Psychophysiology, 60 (7 ), e14258. 10.1111/psyp.14258 36737886
Ball B. H. , DeWitt M. R. , Knight J. B. , & Hicks J. L. (2014). Encoding and retrieval processes involved in the access of source information in the absence of item memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 40 (5 ), 1271–1286. 10.1037/a0037204 24933700
Bastin C. , Diana R. A. , Simon J. , Collette F. , Yonelinas A. P. , & Salmon E. (2013). Associative memory in aging: the effect of unitization on source memory. Psychol Aging, 28 (1 ), 275–283. 10.1037/a0031566 23527745
Bayes T. (1991). An essay towards solving a problem in the doctrine of chances. 1763. MD Comput, 8 (3 ), 157–171. https://www.ncbi.nlm.nih.gov/pubmed/1857193 1857193
Bell A. J. , & Sejnowski T. J. (1995). An information-maximization approach to blind separation and blind deconvolution. Neural Comput, 7 (6 ), 1129–1159. http://www.ncbi.nlm.nih.gov/pubmed/7584893 7584893
Bengson J. J. , & Luck S. J. (2015). Effects of strategy on visual working memory capacity. Psychon Bull Rev. 10.3758/s13423-015-0891-7
Boudewyn M. A. , Luck S. J. , Farrens J. L. , & Kappenman E. S. (2018). How many trials does it take to get a significant ERP effect? It depends. Psychophysiology, 55 (6 ), e13049. 10.1111/psyp.13049 29266241
Bowers J. , & Schacter D. (1990). Implicit memory and test awareness. Journal of Experimental Psychology. Learning, Memory, and Cognition, 16 3 , 404–416.2140400
Bowles B. , Crupi C. , Mirsattari S. M. , Pigott S. E. , Parrent A. G. , Pruessner J. C. , Yonelinas A. P. , & Kohler S. (2007). Impaired familiarity with preserved recollection after anterior temporal-lobe resection that spares the hippocampus. Proc Natl Acad Sci U S A, 104 (41 ), 16382–16387. 10.1073/pnas.0705273104 17905870
Bowles B. , Crupi C. , Pigott S. , Parrent A. , Wiebe S. , Janzen L. , & Kohler S. (2010). Double dissociation of selective recollection and familiarity impairments following two different surgical treatments for temporal-lobe epilepsy. Neuropsychologia, 48 (9 ), 2640–2647. 10.1016/j.neuropsychologia.2010.05.010 20466009
Brainerd C. J. , Nakamura K. , & Lee W. A. (2019). Recollection is fast and slow. J Exp Psychol Learn Mem Cogn, 45 (2 ), 302–319. 10.1037/xlm0000588 29698044
Bridger E. K. , Bader R. , Kriukova O. , Unger K. , & Mecklinger A. (2012). The FN400 is functionally distinct from the N400. Neuroimage, 63 (3 ), 1334–1342. 10.1016/j.neuroimage.2012.07.047 22850570
Bridger E. K. , Bader R. , & Mecklinger A. (2014). More ways than one: ERPs reveal multiple familiarity signals in the word frequency mirror effect. Neuropsychologia, 57 , 179–190.24675676
Bruett H. , & Leynes P. A. (2015). Event-related potentials indicate that fluency can be interpreted as familiarity. Neuropsychologia, 78 , 41–50. 10.1016/j.neuropsychologia.2015.09.035 26432342
Button K. S. , Ioannidis J. P. , Mokrysz C. , Nosek B. A. , Flint J. , Robinson E. S. , & Munafo M. R. (2013). Power failure: why small sample size undermines the reliability of neuroscience. Nat Rev Neurosci, 14 (5 ), 365–376. 10.1038/nrn3475 23571845
Chen X. R. , Gomes C. F. A. , & Brainerd C. J. (2018). Explaining recollection without remembering. J Exp Psychol Learn Mem Cogn, 44 (12 ), 1921–1930. 10.1037/xlm0000559 29389189
Cleary Brown , A. S. , Sawyer B. D. , Nomi J. S. , Ajoku A. C. , & Ryals A. J. (2012). Familiarity from the configuration of objects in 3-dimensional space and its relation to déjà vu: A virtual reality investigation. Consciousness and Cognition, 21 (2 ), 969–975. 10.1016/j.concog.2011.12.010 22322010
Cleary McNeely-White , K. , Huebert A. , & Claxton A. (2021). Déjà vu and the feeling of prediction: an association with familiarity strength. Memory, 29 (7 ), 904–920. 10.1080/09658211.2018.1503686 30384796
Cleary Neisser , J. , McMahan T. , Parsons T. D. , Alwaki A. , Okada N. , Vosoughi A. , Kheder A. , Drane D. L. , & Pedersen N. P. (2021). Subjective distinguishability of seizure and non-seizure Déjà Vu: A case report, brief literature review, and research prospects. Epilepsy Behav, 125 , 108373. 10.1016/j.yebeh.2021.108373 34735965
Cleary A. M. , Neisser J. , McMahan T. , Parsons T. D. , Alwaki A. , Okada N. , Vosoughi A. , Kheder A. , Drane D. L. , & Pedersen N. P. (2021). Subjective distinguishability of seizure and non-seizure Déjà Vu: A case report, brief literature review, and research prospects. Epilepsy Behav, 125 , 108373. 10.1016/j.yebeh.2021.108373 34735965
Coane J. H. , Balota D. A. , Jacoby L. L. , & Dolan P. O. (2011). Not all sources of familiarity are created equal: The case of word frequency and repetition in episodic recognition [Article]. Memory and Cognition, 39 (5 ), 791–805. 10.3758/s13421-010-0069-5 21264634
Coane J. H. , & Umanath S. (2019). I don’t remember vs. I don’t know: Phenomenological states associated with retrieval failures. Journal of Memory and Language, 107 , 152–168. 10.1016/j.jml.2019.05.002
Cohen N. J. , & Eichenbaum H. (1993). Memory, amnesia, and the hippocampal system. MIT Press.
Collaboration O. S. (2015). Estimating the reproducibility of psychological science. Science, 349 (6251 ), aac4716. 10.1126/science.aac4716 26315443
Craik F. I. M. , Rose N. S. , & Gopie N. (2015). Recognition without awareness: Encoding and retrieval factors. Journal of Experimental Psychology. Learning, Memory, and Cognition, 41 5 , 1271–1281.26010824
Curran T. (2000). Brain potentials of recollection and familiarity. Mem Cognit, 28 (6 ), 923–938. http://www.ncbi.nlm.nih.gov/pubmed/11105518
Curran T. (2004). Effects of attention and confidence on the hypothesized ERP correlates of recollection and familiarity. Neuropsychologia, 42 (8 ), 1088–1106. 10.1016/j.neuropsychologia.2003.12.011 15093148
Daselaar S. M. , Fleck M. S. , & Cabeza R. (2006). Triple dissociation in the medial temporal lobes: recollection, familiarity, and novelty. J Neurophysiol, 96 (4 ), 1902–1911. 10.1152/jn.01029.2005 16738210
Delorme A. , & Makeig S. (2004). EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis [Research Support, Non-U.S. Gov’t Research Support, U.S. Gov’t, P.H.S.]. Journal of Neuroscience Methods, 134 (1 ), 9–21. 10.1016/j.jneumeth.2003.10.009 15102499
Diana R. A. , Van den Boom W. , Yonelinas A. P. , & Ranganath C. (2011). ERP correlates of source memory: unitized source information increases familiarity-based retrieval. Brain Res, 1367 , 278–286. 10.1016/j.brainres.2010.10.030 20965154
Diana R. A. , Yonelinas A. P. , & Ranganath C. (2007). Imaging recollection and familiarity in the medial temporal lobe: a three-component model. Trends Cogn Sci, 11 (9 ), 379–386. 10.1016/j.tics.2007.08.001 17707683
Diana R. A. , Yonelinas A. P. , & Ranganath C. (2008). The effects of unitization on familiarity-based source memory: testing a behavioral prediction derived from neuroimaging data. J Exp Psychol Learn Mem Cogn, 34 (4 ), 730–740. 10.1037/0278-7393.34.4.730 18605864
Diana R. A. , Yonelinas A. P. , & Ranganath C. (2010). Medial temporal lobe activity during source retrieval reflects information type, not memory strength. J Cogn Neurosci, 22 (8 ), 1808–1818. 10.1162/jocn.2009.21335 19702458
Diana R. A. , Yonelinas A. P. , & Ranganath C. (2012). Adaptation to cognitive context and item information in the medial temporal lobes. Neuropsychologia, 50 (13 ), 3062–3069. 10.1016/j.neuropsychologia.2012.07.035 22846335
Diana R. A. , Yonelinas A. P. , & Ranganath C. (2013). Parahippocampal cortex activation during context reinstatement predicts item recollection. J Exp Psychol Gen, 142 (4 ), 1287–1297. 10.1037/a0034029 23937182
Duarte A. , Ranganath C. , Winward L. , Hayward D. , & Knight R. T. (2004). Dissociable neural correlates for familiarity and recollection during the encoding and retrieval of pictures. Cognitive Brain Research, 18 (3 ), 255–272. 10.1016/j.cogbrainres.2003.10.010 14741312
Duss S. B. , Reber T. P. , Hanggi J. , Schwab S. , Wiest R. , Muri R. M. , Brugger P. , Gutbrod K. , & Henke K. (2014). Unconscious relational encoding depends on hippocampus. Brain, 137 , 3355–3370. 10.1093/Brain/Awu270 25273998
Duzel E. , Neufang M. , & Heinze H. J. (2005). The oscillatory dynamics of recognition memory and its relationship to event-related responses. Cereb Cortex, 15 (12 ), 1992–2002. 10.1093/cercor/bhi074 15772372
Duzel E. , Vargha-Khadem F. , Heinze H. J. , & Mishkin M. (2001). Brain activity evidence for recognition without recollection after early hippocampal damage. Proc Natl Acad Sci U S A, 98 (14 ), 8101–8106. 10.1073/pnas.131205798 11438748
Duzel E. , Yonelinas A. P. , Mangun G. R. , Heinze H. J. , & Tulving E. (1997). Event-related brain potential correlates of two states of conscious awareness in memory [Research Support, Non-U.S. Gov’t Research Support, U.S. Gov’t, P.H.S.]. Proc Natl Acad Sci U S A, 94 (11 ), 5973–5978. http://www.ncbi.nlm.nih.gov/pubmed/9159185 9159185
Ecker U. K. , Zimmer H. D. , Groh-Bordin C. , & Mecklinger A. (2007). Context effects on familiarity are familiarity effects of context - an electrophysiological study. Int J Psychophysiol, 64 (2 ), 146–156. 10.1016/j.ijpsycho.2007.01.005 17331603
Eichenbaum H. , Yonelinas A. P. , & Ranganath C. (2007). The medial temporal lobe and recognition memory. Annu Rev Neurosci, 30 , 123–152. 10.1146/annurev.neuro.30.051606.094328 17417939
Etz A. , Gronau Q. F. , Dablander F. , Edelsbrunner P. A. , & Baribault B. (2018). How to become a Bayesian in eight easy steps: An annotated reading list. Psychon Bull Rev, 25 (1 ), 219–234. 10.3758/s13423-017-1317-5 28660424
Etz A. , & Vandekerckhove J. (2018). Introduction to Bayesian Inference for Psychology. Psychon Bull Rev, 25 (1 ), 5–34. 10.3758/s13423-017-1262-3 28378250
Ezzyat Y. , Kragel J. E. , Burke J. F. , Levy D. F. , Lyalenko A. , Wanda P. , O’Sullivan L. , Hurley K. B. , Busygin S. , Pedisich I. , Sperling M. R. , Worrell G. A. , Kucewicz M. T. , Davis K. A. , Lucas T. H. , Inman C. S. , Lega B. C. , Jobst B. C. , Sheth S. A. ,…Kahana M. J. (2017). Direct Brain Stimulation Modulates Encoding States and Memory Performance in Humans. Curr Biol. 10.1016/j.cub.2017.03.028
Ezzyat Y. , Wanda P. A. , Levy D. F. , Kadel A. , Aka A. , Pedisich I. , Sperling M. R. , Sharan A. D. , Lega B. C. , Burks A. , Gross R. E. , Inman C. S. , Jobst B. C. , Gorenstein M. A. , Davis K. A. , Worrell G. A. , Kucewicz M. T. , Stein J. M. , Gorniak R. ,…Kahana M. J. (2018). Closed-loop stimulation of temporal cortex rescues functional networks and improves memory. Nat Commun, 9 (1 ), 365. 10.1038/s41467-017-02753-0 29410414
Fell J. , & Axmacher N. (2011). The role of phase synchronization in memory processes. Nat Rev Neurosci, 12 (2 ), 105–118. https://doi.org/nrn2979 [pii] 10.1038/nrn2979 21248789
Fell J. , Klaver P. , Lehnertz K. , Grunwald T. , Schaller C. , Elger C. E. , & Fernandez G. (2001). Human memory formation is accompanied by rhinal-hippocampal coupling and decoupling. Nat Neurosci, 4 (12 ), 1259–1264. 10.1038/nn759 11694886
Fell J. , Ludowig E. , Rosburg T. , Axmacher N. , & Elger C. E. (2008). Phase-locking within human mediotemporal lobe predicts memory formation. Neuroimage, 43 (2 ), 410–419. 10.1016/j.neuroimage.2008.07.021 18703147
Fell J. , Ludowig E. , Staresina B. P. , Wagner T. , Kranz T. , Elger C. E. , & Axmacher N. (2011). Medial temporal theta/alpha power enhancement precedes successful memory encoding: evidence based on intracranial EEG. J Neurosci, 31 (14 ), 5392–5397. 10.1523/JNEUROSCI.3668-10.2011 21471374
Fellner M. C. , Gollwitzer S. , Rampp S. , Kreiselmeyr G. , Bush D. , Diehl B. , Axmacher N. , Hamer H. , & Hanslmayr S. (2019). Spectral fingerprints or spectral tilt? Evidence for distinct oscillatory signatures of memory formation. PLoS Biol, 17 (7 ), e3000403. 10.1371/journal.pbio.3000403 31356598
Fernández G. , Effern A. , Grunwald T. , Pezer N. , Lehnertz K. , Dümpelmann M. , Roost D. V. , & Elger C. E. (1999). Real-Time Tracking of Memory Formation in the Human Rhinal Cortex and Hippocampus. Science, 285 (5433 ), 1582–1585. 10.1126/science.285.5433.1582 10477525
Friedman D. (2013). The Cognitive Aging of Episodic Memory: A View Based On The Event-Related Brain Potential (ERP) [Review]. Front Behav Neurosci, 7 . 10.3389/fnbeh.2013.00111
Gillund G. , & Shiffrin R. M. (1984). A retrieval model for both recognition and recall. Psychological Review, 91 (1 ), 1–67. 10.1037/0033-295X.91.1.1 6571421
Goldinger S. D. , & Papesh M. H. (2013). Chapter Five - Recollection is Fast and Easy: Pupillometric Studies of Face Memory. In Ross B. H. (Ed.), Psychology of Learning and Motivation (Vol. 59 , pp. 191–222). Academic Press. 10.1016/B978-0-12-407187-2.00005-8
Goyal A. , Miller J. , Qasim S. E. , Watrous A. J. , Zhang H. , Stein J. M. , Inman C. S. , Gross R. E. , Willie J. T. , Lega B. , Lin J. J. , Sharan A. , Wu C. , Sperling M. R. , Sheth S. A. , McKhann G. M. , Smith E. H. , Schevon C. , & Jacobs J. (2020). Functionally distinct high and low theta oscillations in the human hippocampus. Nat Commun, 11 (1 ), 2469. 10.1038/s41467-020-15670-6 32424312
Griffiths B. , Mazaheri A. , Debener S. , & Hanslmayr S. (2016). Brain oscillations track the formation of episodic memories in the real world. Neuroimage, 143 , 256–266. 10.1016/j.neuroimage.2016.09.021 27622395
Gruber M. J. , & Otten L. J. (2010). Voluntary control over prestimulus activity related to encoding. J Neurosci, 30 (29 ), 9793–9800. 10.1523/JNEUROSCI.0915-10.2010 20660262
Haig A. R. , Gordon E. , & Hook S. (1997). To scale or not to scale: McCarthy and Wood revisited. Electroencephalogr Clin Neurophysiol, 103 (2 ), 323–325. 10.1016/s0013-4694(97)00009-6 9277635
Hannula D. E. , & Greene A. J. (2012). The hippocampus reevaluated in unconscious learning and memory: at a tipping point? Front Hum Neurosci, 6 , 80. 10.3389/fnhum.2012.00080 22518102
Hannula D. E. , Libby L. A. , Yonelinas A. P. , & Ranganath C. (2013). Medial temporal lobe contributions to cued retrieval of items and contexts. Neuropsychologia, 51 (12 ), 2322–2332. 10.1016/j.neuropsychologia.2013.02.011 23466350
Hannula D. E. , Ryan J. D. , & Warren D. E. (2017). Beyond Long-Term Declarative Memory: Evaluating Hippocampal Contributions to Unconscious Memory Expression, Perception, and Short-Term Retention. In Hannula D. E. & Duff M. C. (Eds.), The Hippocampus from Cells to Systems: Structure, Connectivity, and Functional Contributions to Memory and Flexible Cognition (pp. 281–336). Springer International Publishing. 10.1007/978-3-319-50406-3_10
Hanslmayr S. , Axmacher N. , & Inman C. S. (2019). Modulating Human Memory via Entrainment of Brain Oscillations. Trends in Neurosciences. 10.1016/j.tins.2019.04.004
Hanslmayr S. , Spitzer B. , & Bauml K. H. (2009). Brain oscillations dissociate between semantic and nonsemantic encoding of episodic memories. Cereb Cortex, 19 (7 ), 1631–1640. 10.1093/cercor/bhn197 19001457
Hanslmayr S. , & Staudigl T. (2014). How brain oscillations form memories--a processing based perspective on oscillatory subsequent memory effects. Neuroimage, 85 Pt 2 , 648–655. 10.1016/j.neuroimage.2013.05.121 23769913
Haskins A. L. , Yonelinas A. P. , Quamme J. R. , & Ranganath C. (2008). Perirhinal cortex supports encoding and familiarity-based recognition of novel associations. Neuron, 59 (4 ), 554–560. 10.1016/j.neuron.2008.07.035 18760692
Heinsberg L. W. , & Weeks D. E. (2022). Post hoc power is not informative. Genet Epidemiol, 46 (7 ), 390–394. 10.1002/gepi.22464 35642557
Henke K. (2010). A model for memory systems based on processing modes rather than consciousness. Nat Rev Neurosci, 11 (7 ), 523–532. 10.1038/nrn2850 20531422
Henson R. (2005). What can functional neuroimaging tell the experimental psychologist? Q J Exp Psychol A, 58 (2 ), 193–233. 10.1080/02724980443000502 15903115
Herron J. E. (2007). Decomposition of the ERP late posterior negativity: effects of retrieval and response fluency. Psychophysiology, 44 (2 ), 233–244. 10.1111/j.1469-8986.2006.00489.x 17343707
Hicks J. L. , Marsh R. L. , & Ritschel L. (2002). The role of recollection and partial information in source monitoring. Journal of Experimental Psychology: Learning, Memory, and Cognition, 28 (3 ), 503–508. 10.1037//0278-7393.28.3.503 12018502
Hoenig J. M. , & Heisey D. M. (2001). The Abuse of Power. The American Statistician, 55 (1 ), 19–24. 10.1198/000313001300339897
Hoffman H. G. (1997). Role of memory strength in reality monitoring decisions: evidence from source attribution biases. J Exp Psychol Learn Mem Cogn, 23 (2 ), 371–383. 10.1037//0278-7393.23.2.371 9080009
Horner A. J. , Bisby J. A. , Bush D. , Lin W.-J. , & Burgess N. (2015). Evidence for holistic episodic recollection via hippocampal pattern completion [Article]. Nat Commun, 6 . 10.1038/ncomms8462
Inman C. S. , Manns J. R. , Bijanki K. R. , Bass D. I. , Hamann S. , Drane D. L. , Fasano R. E. , Kovach C. K. , Gross R. E. , & Willie J. T. (2018). Direct electrical stimulation of the amygdala enhances declarative memory in humans. Proc Natl Acad Sci U S A, 115 (1 ), 98–103. 10.1073/pnas.1714058114 29255054
Ioannidis J. P. (2005). Why most published research findings are false. PLoS Med, 2 (8 ), e124. 10.1371/journal.pmed.0020124 16060722
Ioannidis J. P. , Munafo M. R. , Fusar-Poli P. , Nosek B. A. , & David S. P. (2014). Publication and other reporting biases in cognitive sciences: detection, prevalence, and prevention. Trends Cogn Sci, 18 (5 ), 235–241. 10.1016/j.tics.2014.02.010 24656991
Ioannidis J. P. , Nosek B. , & Iorns E. (2012). Reproducibility concerns. Nat Med, 18 (12 ), 1736–1737. 10.1038/nm.3020 23223056
Jacobs J. , Hwang G. , Curran T. , & Kahana M. J. (2006). EEG oscillations and recognition memory: theta correlates of memory retrieval and decision making. Neuroimage, 32 (2 ), 978–987. 10.1016/j.neuroimage.2006.02.018 16843012
Jacobs J. , Miller J. , Lee S. A. , Coffey T. , Watrous A. J. , Sperling M. R. , Sharan A. , Worrell G. , Berry B. , Lega B. , Jobst B. C. , Davis K. , Gross R. E. , Sheth S. A. , Ezzyat Y. , Das S. R. , Stein J. , Gorniak R. , Kahana M. J. , & Rizzuto D. S. (2016). Direct Electrical Stimulation of the Human Entorhinal Region and Hippocampus Impairs Memory. Neuron, 92 (5 ), 983–990. 10.1016/j.neuron.2016.10.062 27930911
James T. , Strunk J. , Arndt J. , & Duarte A. (2016). Age-related deficits in selective attention during encoding increase demands on episodic reconstruction during context retrieval: An ERP study. Neuropsychologia, 86 , 66–79. 10.1016/j.neuropsychologia.2016.04.009 27094851
Jarosz A. F. , & Wiley J. (2014). What Are the Odds? A Practical Guide to Computing and Reporting Bayes Factors. J. Probl. Solving, 7 .
Jeffreys H. (1998). The Theory of Probability. OUP Oxford. https://books.google.com/books?id=vh9Act9rtzQC
Jensen K. M. , & MacDonald J. A. (2023). Towards thoughtful planning of ERP studies: How participants, trials, and effect magnitude interact to influence statistical power across seven ERP components. Psychophysiology, 60 (7 ), e14245. 10.1111/psyp.14245 36577739
Johansson M. , & Mecklinger A. (2003). The late posterior negativity in ERP studies of episodic memory: action monitoring and retrieval of attribute conjunctions. Biol Psychol, 64 (1–2 ), 91–117. 10.1016/s0301-0511(03)00104-2 14602357
Johnson M. K. , Hashtroudi S. , & Lindsay D. S. (1993). Source monitoring. Psychol Bull, 114 (1 ), 3–28. http://www.ncbi.nlm.nih.gov/pubmed/8346328 8346328
Jones T. C. , & Bartlett J. C. (2009). When false recognition is out of control: the case of facial conjunctions. Mem Cognit, 37 (2 ), 143–157. 10.3758/MC.37.2.143
Jones T. C. , & Jacoby L. L. (2005). Conjunction errors in recognition memory: modality-free errors for older adults but not for young adults. Acta Psychol (Amst), 120 (1 ), 55–73. 10.1016/j.actpsy.2005.03.003 15876419
Jones T. C. , Jacoby L. L. , & Gellis L. A. (2001). Cross-Modal Feature and Conjunction Errors in Recognition Memory. Journal of Memory and Language, 44 (1 ), 131–152. 10.1006/jmla.2001.2713
Kandel E. R. (2013). Principles of neural science (5th ed.). McGraw-Hill.
Khambhati A. N. , Kahn A. E. , Costantini J. , Ezzyat Y. , Solomon E. A. , Gross R. E. , Jobst B. C. , Sheth S. A. , Zaghloul K. A. , Worrell G. , Seger S. , Lega B. C. , Weiss S. , Sperling M. R. , Gorniak R. , Das S. R. , Stein J. M. , Rizzuto D. S. , Kahana M. J. ,…Bassett D. S. (2019). Functional control of electrophysiological network architecture using direct neurostimulation in humans. Netw Neurosci, 3 (3 ), 848–877. 10.1162/netn_a_00089 31410383
Klimesch W. , Doppelmayr M. , Schwaiger J. , Winkler T. , & Gruber W. (2000). Theta oscillations and the ERP old/new effect: independent phenomena? Clin Neurophysiol, 111 (5 ), 781–793. https://doi.org/S1388-2457(00)00254-6 [pii]10802447
Klimesch W. , Doppelmayr M. , Yonelinas A. , Kroll N. E. , Lazzara M. , Rohm D. , & Gruber W. (2001). Theta synchronization during episodic retrieval: neural correlates of conscious awareness. Brain Res Cogn Brain Res, 12 (1 ), 33–38. https://doi.org/S0926-6410(01)00024-6 [pii]11489606
Klimesch W. , Hanslmayr S. , Sauseng P. , Gruber W. , Brozinsky C. J. , Kroll N. E. , Yonelinas A. P. , & Doppelmayr M. (2006). Oscillatory EEG correlates of episodic trace decay. Cereb Cortex, 16 (2 ), 280–290. 10.1093/cercor/bhi107 15888605
Köhler S. , & Martin C. B. (2020). Familiarity impairments after anterior temporal-lobe resection with hippocampal sparing: Lessons learned from case NB. Neuropsychologia, 138 , 107339. 10.1016/j.neuropsychologia.2020.107339 31930957
Koriat A. , Levy-Sadot R. , Edry E. , & de Marcas S. (2003). What do we know about what we cannot remember? Accessing the semantic attributes of words that cannot be recalled. Journal of Experimental Psychology. Learning, Memory, and Cognition, 29 6 , 1095–1105.14622049
Kurilla B. P. , & Gonsalves B. D. (2012). An ERP investigation into the strategic regulation of the fluency heuristic during recognition memory. Brain Research. 10.1016/j.brainres.2011.12.060
Kurilla B. P. , & Westerman D. L. (2010). Source memory for unidentified stimuli. J Exp Psychol Learn Mem Cogn, 36 (2 ), 398–410. 10.1037/a0018279 20192538
Kutas M. , & Federmeier K. D. (2011). Thirty years and counting: finding meaning in the N400 component of the event-related brain potential (ERP). Annu Rev Psychol, 62 , 621–647. 10.1146/annurev.psych.093008.131123 20809790
Kwon S. , Rugg M. D. , Wiegand R. , Curran T. , & Morcom A. M. (2023). A meta-analysis of event-related potential correlates of recognition memory. Psychonomic Bulletin & Review. 10.3758/s13423-023-02309-y
Lakens D. (2022). Sample Size Justification. Collabra: Psychology, 8 (1 ). 10.1525/collabra.33267
Lampinen J. M. , Meier C. R. , Arnal J. D. , & Leding J. K. (2005). Compelling untruths: content borrowing and vivid false memories. J Exp Psychol Learn Mem Cogn, 31 (5 ), 954–963. 10.1037/0278-7393.31.5.954 16248744
Leutgeb S. , & Leutgeb J. K. (2007). Pattern separation, pattern completion, and new neuronal codes within a continuous CA3 map. Learn Mem, 14 (11 ), 745–757. 10.1101/lm.703907 18007018
Levine B. , Freedman M. , Dawson D. , Black S. , & Stuss D. (1999). Ventral frontal contribution to self-regulation: Convergence of episodic memory and inhibition. Neurocase, 5 (3 ), 263–275.
Levine M. , & Ensom M. H. (2001). Post hoc power analysis: an idea whose time has passed? Pharmacotherapy, 21 (4 ), 405–409. 10.1592/phco.21.5.405.34503 11310512
Leynes A. , & Zish K. (2012). Event-related potential (ERP) evidence for fluency-based recognition memory. Neuropsychologia, 50 (14 ), 3240–3249. 10.1016/j.neuropsychologia.2012.10.004 23063967
Leynes P. , Kalelkar A. U. , Shaik H. T. , & Sawhney S. (2023). Event-Related Potential (ERP) evidence for fluency and disfluency effects on recognition memory. Brain Cogn, 167 , 105961. 10.1016/j.bandc.2023.105961 36893675
Leynes P. , & Mok B. (2020). Context Influences the FN400 Recognition Event-Related Potential. International Journal of Psychophysiology.
Leynes P. , & Upadhyay T. (2022). Context dissociations of the FN400 and N400 are evidence for recognition based on relative or absolute familiarity. Brain and Cognition, 162 , 105903. 10.1016/j.bandc.2022.105903 36030708
Leynes P. A. (2012). Event-related potential (ERP) evidence for source-monitoring based on the absence of information [Randomized Controlled Trial Research Support, Non-U.S. Gov’t]. International Journal of Psychophysiology, 84 (3 ), 284–295. 10.1016/j.ijpsycho.2012.03.007 22484301
Leynes P. A. , & Addante R. J. (2016). Neurophysiological evidence that perceptions of fluency produce mere exposure effects. Cognitive, Affective, & Behavioral Neuroscience, 16 (4 ), 754–767.
Leynes P. A. , Allen J. D. , & Marsh R. L. (1998). Topographic differences in CNV amplitude reflect different preparatory processes. International Journal of Psychophysiology, 31 (1 ), 33–44. http://www.ncbi.nlm.nih.gov/pubmed/9934619 9934619
Leynes P. A. , Askin B. , & Landau J. D. (2016). Visual perspective during remembering: ERP evidence of familiarity-based source monitoring. Cortex. 10.1016/j.cortex.2016.12.003
Leynes P. A. , Bruett H. , Krizan J. , & Veloso A. (2017). What psychological process is reflected in the FN400 event-related potential component? Brain Cogn, 113 , 142–154. 10.1016/j.bandc.2017.02.004 28235696
Leynes P. A. , Landau J. , Walker J. , & Addante R. J. (2005). Event-related potential evidence for multiple causes of the revelation effect. Conscious Cogn, 14 (2 ), 327–350. https://doi.org/S1053-8100(04)00090-X [pii] 10.1016/j.concog.2004.08.005 15950886
Leynes P. A. , & Mok B. A. (2017). Encoding focus alters diagnostic recollection and event-related potentials (ERPs). Brain and Cognition, 117 (Supplement C ), 1–11. 10.1016/j.bandc.2017.06.011 28683338
Liesefeld H. R. , Liesefeld A. M. , & Zimmer H. D. (2016). Recollection is delayed under changed viewing conditions: A graded effect on the latency of the late posterior component. Psychophysiology, 53 (12 ), 1811–1822. 10.1111/psyp.12760 27628129
Linde-Domingo J. , Treder M. S. , Kerrén C. , & Wimber M. (2019). Evidence that neural information flow is reversed between object perception and object reconstruction from memory. Nat Commun, 10 (1 ), 179. 10.1038/s41467-018-08080-2 30643124
Liu Y. , Su Y. , Xu G. , & Chan R. C. K. (2007). Two dissociable aspects of feeling-of-knowing: Knowing that you know and knowing that you do not know. Quarterly Journal of Experimental Psychology, 60 (5 ), 672–680. 10.1080/17470210601184039
Lopez-Calderon J. , & Luck S. J. (2014). ERPLAB: an open-source toolbox for the analysis of event-related potentials. Front Hum Neurosci, 8 , 213. 10.3389/fnhum.2014.00213 24782741
Luck S. J. (2014). An introduction to the event-related potential technique (Second edition. ed.). The MIT Press.
Malpass R. S. , & Devine P. G. (1981). Eyewitness identification: Lineup instructions and the absence of the offender. Journal of Applied Psychology, 66 (4 ), 482–489. 10.1037/0021-9010.66.4.482
Mandler G. (1980). Recognizing: The judgment of previous occurrence. Psychological Review, 87 (3 ), 252–271. 10.1037/0033-295X.87.3.252
Manns J. R. , Hopkins R. O. , Reed J. M. , Kitchener E. G. , & Squire L. R. (2003). Recognition memory and the human hippocampus. Neuron, 37 (1 ), 171–180. 10.1016/s0896-6273(02)01147-9 12526782
Maoz S. L. L. , Stangl M. , Topalovic U. , Batista D. , Hiller S. , Aghajan Z. M. , Knowlton B. , Stern J. , Langevin J.-P. , Fried I. , Eliashiv D. , & Suthana N. (2023). Dynamic neural representations of memory and space during human ambulatory navigation. Nat Commun, 14 (1 ), 6643. 10.1038/s41467-023-42231-4 37863929
Martin C. B. , Bowles B. , Mirsattari S. M. , & Kohler S. (2011). Selective familiarity deficits after left anterior temporal-lobe removal with hippocampal sparing are material specific. Neuropsychologia, 49 (7 ), 1870–1878. 10.1016/j.neuropsychologia.2011.03.012 21419788
Mazaheri A. (2022). 115The Relationship Between Evoked and Induced EEG/MEG Changes: Going Beyond Labels. In Gable P. A. , Miller M. W. , & Bernat E. M. (Eds.), The Oxford Handbook of EEG Frequency (pp. 0). Oxford University Press. 10.1093/oxfordhb/9780192898340.013.6
Mazaheri A. , & Jensen O. (2010). Rhythmic Pulsing: Linking Ongoing Brain Activity with Evoked Responses [Review]. Frontiers in Human Neuroscience, 4 . 10.3389/fnhum.2010.00177
Mazaheri A. , Slagter H. A. , Thut G. , & Foxe J. J. (2018). Orchestration of brain oscillations: principles and functions. Eur J Neurosci, 48 (7 ), 2385–2388. 10.1111/ejn.14189 30276895
McCarthy G. , & Wood C. C. (1985). Scalp distributions of event-related potentials: an ambiguity associated with analysis of variance models [Research Support, U.S. Gov’t, Non-P.H.S. Research Support, U.S. Gov’t, P.H.S.]. Electroencephalography and Clinical Neurophysiology, 62 (3 ), 203–208. http://www.ncbi.nlm.nih.gov/pubmed/2581760 2581760
Mecklinger , & Bader . (2020). From fluency to recognition decisions: A broader view of familiarity-based remembering. Neuropsychologia, 146 , 107527. 10.1016/j.neuropsychologia.2020.107527 32540265
Mecklinger A. (2006). Electrophysiological measures of familiarity memory. Clin EEG Neurosci, 37 (4 ), 292–299. http://www.ncbi.nlm.nih.gov/pubmed/17073167 17073167
Mecklinger A. , & Jäger T. (2009). Episodic memory storage and retrieval: Insights from electrophysiological measures. In Rösler F. , Ranganath C. , Röder B. , & Kluwe R. (Eds.), Neuroimaging of Human Memory: Linking cognitive processes to neural systems (pp. 0). Oxford University Press. 10.1093/acprof:oso/9780199217298.003.0020
Mecklinger A. , Johansson M. , Parra M. , & Hanslmayr S. (2007). Source-retrieval requirements influence late ERP and EEG memory effects. Brain Res, 1172 , 110–123. 10.1016/j.brainres.2007.07.070 17822684
Mecklinger A. , Rosburg T. , & Johansson M. (2016). Reconstructing the past: The Late Posterior Negativity (LPN) in episodic memory studies. Neurosci Biobehav Rev. 10.1016/j.neubiorev.2016.06.024
Mecklinger A. , von Cramon D. Y. , & Matthes-von Cramon G. (1998). Event-related potential evidence for a specific recognition memory deficit in adult survivors of cerebral hypoxia. Brain, 121 ( Pt 10 ), 1919–1935. http://www.ncbi.nlm.nih.gov/pubmed/9798747 9798747
Meeter M. , Myers C. E. , & Gluck M. A. (2005). Integrating incremental learning and episodic memory models of the hippocampal region. Psychol Rev, 112 (3 ), 560–585. 10.1037/0033-295x.112.3.560 16060751
Milner B. , Squire L. R. , & Kandel E. R. (1998). Cognitive neuroscience and the study of memory. Neuron, 20 (3 ), 445–468. http://www.ncbi.nlm.nih.gov/pubmed/9539121 9539121
Mitchell K. J. , & Johnson M. K. (2009). Source monitoring 15 years later: what have we learned from fMRI about the neural mechanisms of source memory? Psychol Bull, 135 (4 ), 638–677. 10.1037/a0015849 19586165
Montaldi D. , & Mayes A. R. (2010). The role of recollection and familiarity in the functional differentiation of the medial temporal lobes. Hippocampus, 20 (11 ), 1291–1314. 10.1002/hipo.20853 20928828
Moonesinghe R. , Khoury M. J. , & Janssens A. C. J. W. (2007). Most Published Research Findings Are False—But a Little Replication Goes a Long Way. PLoS Med, 4 (2 ), e28. 10.1371/journal.pmed.0040028 17326704
Morey R. D. , & Rouder J. N. (2011). Bayes factor approaches for testing interval null hypotheses. Psychol Methods, 16 (4 ), 406–419. 10.1037/a0024377 21787084
Moscovitch M. (2008). The hippocampus as a “stupid,” domain-specific module: Implications for theories of recent and remote memory, and of imagination [Research Support, Non-U.S. Gov’t]. Canadian Journal of Experimental Psychology, 62 (1 ), 62–79. 10.1037/1196-1961.62.1.62 18473631
Mruczek R. E. , & Sheinberg D. L. (2007). Context familiarity enhances target processing by inferior temporal cortex neurons. J Neurosci, 27 (32 ), 8533–8545. 10.1523/JNEUROSCI.2106-07.2007 17687031
Muller A. , Sirianni L. A. , & Addante R. J. (2021). Neural correlates of the Dunning–Kruger effect. European Journal of Neuroscience, 53 (2 ), 460–484. 10.1111/ejn.14935 32761954
Muller A. L. (2019). Neurological Correlates of the Dunning-Kruger Effect.
Nosek B. A. , Hardwicke T. E. , Moshontz H. , Allard A. , Corker K. S. , Dreber A. , Fidler F. , Hilgard J. , Kline Struhl M. , Nuijten M. B. , Rohrer J. M. , Romero F. , Scheel A. M. , Scherer L. D. , Schönbrodt F. D. , & Vazire S. (2022). Replicability, Robustness, and Reproducibility in Psychological Science. Annu Rev Psychol, 73 , 719–748. 10.1146/annurev-psych-020821-114157 34665669
O’Neill M. , & Diana R. A. (2017). The neurocognitive basis of borrowed context information. Cortex, 91 , 89–100. 10.1016/j.cortex.2017.01.014 28219653
Otten L. J. , Quayle A. H. , Akram S. , Ditewig T. A. , & Rugg M. D. (2006). Brain activity before an event predicts later recollection. Nat Neurosci, 9 (4 ), 489–491. 10.1038/nn1663 16501566
Ozubko J. D. , Sirianni L. A. , Ahmad F. N. , MacLeod C. M. , & Addante R. J. (2021). Recallable but not recognizable: The influence of semantic priming in recall paradigms. Cognitive, Affective, & Behavioral Neuroscience. 10.3758/s13415-020-00854-w
Paller K. A. , & Kutas M. (1992). Brain Potentials during Memory Retrieval Provide Neurophysiological Support for the Distinction between Conscious Recollection and Priming. Journal of Cognitive Neuroscience, 4 (4 ), 375–392. 10.1162/jocn.1992.4.4.375 23968130
Parks C. M. (2007). The Role of Noncriterial Recollection in Estimating Recollection and Familiarity. J Mem Lang, 57 (1 ), 81–100. 10.1016/j.jml.2007.03.003 18591986
Parks C. M. , Murray L. J. , Elfman K. , & Yonelinas A. P. (2011). Variations in recollection: the effects of complexity on source recognition. J Exp Psychol Learn Mem Cogn, 37 (4 ), 861–873. 10.1037/a0022798 21417513
Parks C. M. , & Yonelinas A. P. (2009). Evidence for a memory threshold in second-choice recognition memory responses. Proc Natl Acad Sci U S A, 106 (28 ), 11515–11519. 10.1073/pnas.0905505106 19564612
Parks C. M. , & Yonelinas A. P. (2015). The importance of unitization for familiarity-based learning. J Exp Psychol Learn Mem Cogn, 41 (3 ), 881–903. 10.1037/xlm0000068 25329077
Quamme J. R. , Yonelinas A. P. , & Norman K. A. (2007). Effect of unitization on associative recognition in amnesia. Hippocampus, 17 (3 ), 192–200. 10.1002/hipo.20257 17203466
Ranganath C. (2010). A unified framework for the functional organization of the medial temporal lobes and the phenomenology of episodic memory. Hippocampus, 20 (11 ), 1263–1290. 10.1002/hipo.20852 20928833
Ranganath C. , & Ritchey M. (2012). Two cortical systems for memory-guided behaviour. Nat Rev Neurosci, 13 (10 ), 713–726. 10.1038/nrn3338 22992647
Ranganath C. , Yonelinas A. P. , Cohen M. X. , Dy C. J. , Tom S. M. , & D’Esposito M. (2004). Dissociable correlates of recollection and familiarity within the medial temporal lobes [Clinical Trial Research Support, Non-U.S. Gov’t Research Support, U.S. Gov’t, P.H.S.]. Neuropsychologia, 42 (1 ), 2–13. http://www.ncbi.nlm.nih.gov/pubmed/14615072 14615072
Ratcliff R. , & Starns J. J. (2009). Modeling confidence and response time in recognition memory. Psychol Rev, 116 (1 ), 59–83. 10.1037/a0014086 19159148
Reagh Z. M. , & Ranganath C. (2018). What does the functional organization of cortico-hippocampal networks tell us about the functional organization of memory? Neurosci Lett, 680 , 69–76. 10.1016/j.neulet.2018.04.050 29704572
Reber P. J. (2013). The neural basis of implicit learning and memory: a review of neuropsychological and neuroimaging research. Neuropsychologia, 51 (10 ), 2026–2042. 10.1016/j.neuropsychologia.2013.06.019 23806840
Reber T. P. , DoLam A. T. A. , Axmacher N. , Elger C. E. , Helmstaedter C. , Henke K. , & Fell J. (2015). Intracranial EEG Correlates of Implicit Relational Inference within the Hippocampus. Hippocampus, n/a–n/a. 10.1002/hipo.22490
Reber T. P. , Luechinger R. , Boesiger P. , & Henke K. (2012). Unconscious relational inference recruits the hippocampus. J Neurosci, 32 (18 ), 6138–6148. 10.1523/JNEUROSCI.5639-11.2012 22553020
Rhodes S. M. , & Donaldson D. I. (2007). Electrophysiological evidence for the influence of unitization on the processes engaged during episodic retrieval: enhancing familiarity based remembering. Neuropsychologia, 45 (2 ), 412–424. 10.1016/j.neuropsychologia.2006.06.022 16930640
Rivas-Fernández M. Á. , Galdo-Álvarez S. , Zurrón M. , Díaz F. , & Lindín M. (2020). Spatiotemporal pattern of brain electrical activity related to immediate and delayed episodic memory retrieval. Neurobiology of Learning and Memory, 175 , 107309. 10.1016/j.nlm.2020.107309 32890759
Roberts B. M. , Clarke A. , Addante R. J. , & Ranganath C. (2018). Entrainment enhances theta oscillations and improves episodic memory. Cogn Neurosci, 9 (3–4 ), 181–193. 10.1080/17588928.2018.1521386 30198823
Rolls E. T. (2013). The mechanisms for pattern completion and pattern separation in the hippocampus [Review]. Front Syst Neurosci, 7 , 74. 10.3389/fnsys.2013.00074 24198767
Rouder J. N. , Speckman P. L. , Sun D. , Morey R. D. , & Iverson G. (2009). Bayesian t tests for accepting and rejecting the null hypothesis [Research Support, N.I.H., Extramural Research Support, U.S. Gov’t, Non-P.H.S.]. Psychon Bull Rev, 16 (2 ), 225–237. 10.3758/PBR.16.2.225 19293088
Rugg M. D. (2009). Functional neuroimaging and cognitive theory. In Rösler F. , Ranganath C. , Röder B. , & Kluwe R. (Eds.), Neuroimaging of Human Memory: Linking cognitive processes to neural systems (pp. 0). Oxford University Press. 10.1093/acprof:oso/9780199217298.003.0024
Rugg M. D. , & Coles M. G. H. (1995). The ERP and cognitive psychology: conceptual issues. In Electrophysiology of Mind: Event-Related Brain Potentials and Cognition. Oxford University Press.
Rugg M. D. , & Curran T. (2007). Event-related potentials and recognition memory. Trends Cogn Sci, 11 (6 ), 251–257. 10.1016/j.tics.2007.04.004 17481940
Rugg M. D. , Mark R. E. , Walla P. , Schloerscheidt A. M. , Birch C. S. , & Allan K. (1998). Dissociation of the neural correlates of implicit and explicit memory. Nature, 392 (6676 ), 595–598. 10.1038/33396 9560154
Rugg M. D. , Walla P. , Schloerscheidt A. M. , Fletcher P. C. , Frith C. D. , & Dolan R. J. (1998). Neural correlates of depth of processing effects on recollection: evidence from brain potentials and positron emission tomography. Exp Brain Res, 123 (1–2 ), 18–23. http://www.ncbi.nlm.nih.gov/pubmed/9835388 9835388
Rugg M. D. , & Yonelinas A. P. (2003). Human recognition memory: a cognitive neuroscience perspective. Trends in Cognitive Sciences, 7 (7 ), 313–319. 10.1016/s1364-6613(03)00131-1 12860190
Ryals A. J. , Yadon C. A. , Nomi J. S. , & Cleary A. M. (2011). When word identification fails: ERP correlates of recognition without identification and of word identification failure. Neuropsychologia, 49 (12 ), 3224–3237. 10.1016/j.neuropsychologia.2011.07.027 21827776
Schönbrodt F. D. , & Wagenmakers E.-J. (2018). Bayes factor design analysis: Planning for compelling evidence. Psychonomic Bulletin & Review, 25 (1 ), 128–142. 10.3758/s13423-017-1230-y 28251595
Shallice T. (1988). From neuropsychology to mental structure [doi:10.1017/CBO9780511526817]. Cambridge University Press. 10.1017/CBO9780511526817
Simons J. S. (2009). Constraints on cognitive theory from neuroimaging studies of source memory. In Rösler F. , Ranganath C. , Röder B. , & Kluwe R. (Eds.), Neuroimaging of Human Memory: Linking cognitive processes to neural systems (pp. 0). Oxford University Press. 10.1093/acprof:oso/9780199217298.003.0022
Squire L. R. (2009). The legacy of patient H.M. for neuroscience. Neuron, 61 (1 ), 6–9. 10.1016/j.neuron.2008.12.023 19146808
Squire L. R. , Wixted J. T. , & Clark R. E. (2007). Recognition memory and the medial temporal lobe: a new perspective. Nat Rev Neurosci, 8 (11 ), 872–883. 10.1038/nrn2154 17948032
Squire L. R. , & Zola S. M. (1997). Amnesia, memory and brain systems. Philos Trans R Soc Lond B Biol Sci, 352 (1362 ), 1663–1673. 10.1098/rstb.1997.0148 9415918
Stangl M. , Maoz S. L. , & Suthana N. (2023). Mobile cognition: imaging the human brain in the ‘real world’. Nature Reviews Neuroscience, 24 (6 ), 347–362. 10.1038/s41583-023-00692-y 37046077
Staresina B. P. , Cooper E. , & Henson R. N. (2013). Reversible information flow across the medial temporal lobe: the hippocampus links cortical modules during memory retrieval. J Neurosci, 33 (35 ), 14184–14192. 10.1523/JNEUROSCI.1987-13.2013 23986252
Staresina B. P. , Fell J. , Do Lam A. T. , Axmacher N. , & Henson R. N. (2012). Memory signals are temporally dissociated in and across human hippocampus and perirhinal cortex. Nat Neurosci, 15 (8 ), 1167–1173. 10.1038/nn.3154 22751037
Staresina B. P. , Fell J. , Dunn J. C. , Axmacher N. , & Henson R. N. (2013). Using state-trace analysis to dissociate the functions of the human hippocampus and perirhinal cortex in recognition memory. Proc Natl Acad Sci U S A, 110 (8 ), 3119–3124. 10.1073/pnas.1215710110 23382181
Staresina B. P. , & Wimber M. (2019). A Neural Chronometry of Memory Recall. Trends in Cognitive Sciences, 23 (12 ), 1071–1085. 10.1016/j.tics.2019.09.011 31672429
Stróżak P. , Leynes P. A. , & Wojtasiński M. (2020). Distinct FN400/N400 memory effects for perceptually fluent and disfluent words. Brain Cogn, 147 , 105661. 10.1016/j.bandc.2020.105661 33360780
Suthana N. , Ekstrom A. , Moshirvaziri S. , Knowlton B. , & Bookheimer S. (2011). Dissociations within human hippocampal subregions during encoding and retrieval of spatial information. Hippocampus, 21 (7 ), 694–701. 10.1002/hipo.20833 20882543
Suthana N. , & Fried I. (2012). Percepts to recollections: insights from single neuron recordings in the human brain. Trends Cogn Sci, 16 (8 ), 427–436. 10.1016/j.tics.2012.06.006 22795560
Suthana N. , Haneef Z. , Stern J. , Mukamel R. , Behnke E. , Knowlton B. , & Fried I. (2012). Memory enhancement and deep-brain stimulation of the entorhinal area [Research Support, N.I.H., Extramural Research Support, Non-U.S. Gov’t]. New England Journal of Medicine, 366 (6 ), 502–510. 10.1056/NEJMoa1107212 22316444
Suthana N. A. , Donix M. , Wozny D. R. , Bazih A. , Jones M. , Heidemann R. M. , Trampel R. , Ekstrom A. D. , Scharf M. , Knowlton B. , Turner R. , & Bookheimer S. Y. (2015). High-resolution 7T fMRI of Human Hippocampal Subfields during Associative Learning. J Cogn Neurosci, 27 (6 ), 1194–1206. 10.1162/jocn_a_00772 25514656
Suthana N. A. , Parikshak N. N. , Ekstrom A. D. , Ison M. J. , Knowlton B. J. , Bookheimer S. Y. , & Fried I. (2015). Specific responses of human hippocampal neurons are associated with better memory. Proc Natl Acad Sci U S A, 112 (33 ), 10503–10508. 10.1073/pnas.1423036112 26240357
Titiz A. , Hill M. R. H. , Mankin E. A. , Z M. A. , Eliashiv D. , Tchemodanov N. , Maoz U. , Stern J. , Tran M. E. , Schuette P. , Behnke E. , Suthana N. A. , & Fried I. (2017). Theta-burst microstimulation in the human entorhinal area improves memory specificity. Elife, 6 . 10.7554/eLife.29515
Topalovic U. , Barclay S. , Ling C. , Alzuhair A. , Yu W. , Hokhikyan V. , Chandrakumar H. , Rozgic D. , Jiang W. , Basir-Kazeruni S. , Maoz S. L. , Inman C. S. , Stangl M. , Gill J. , Bari A. , Fallah A. , Eliashiv D. , Pouratian N. , Fried I. ,…Markovic D. (2023). A wearable platform for closed-loop stimulation and recording of single-neuron and local field potential activity in freely moving humans. Nature Neuroscience, 26 (3 ), 517–527. 10.1038/s41593-023-01260-4 36804647
Trott C. T. , Friedman D. , Ritter W. , Fabiani M. , & Snodgrass J. G. (1999). Episodic priming and memory for temporal source: event-related potentials reveal age-related differences in prefrontal functioning. Psychol Aging, 14 (3 ), 390–413. http://www.ncbi.nlm.nih.gov/pubmed/10509695 10509695
Tu H. W. , Alty E. E. , & Diana R. A. (2017). Event-related potentials during encoding: Comparing unitization to relational processing. Brain Res, 1667 , 46–54. 10.1016/j.brainres.2017.05.003 28495307
Urbach T. P. , & Kutas M. (2002). The intractability of scaling scalp distributions to infer neuroelectric sources [Research Support, U.S. Gov’t, P.H.S.]. Psychophysiology, 39 (6 ), 791–808. 10.1017/S0048577202010648 12462507
van Diepen R. M. , & Mazaheri A. (2018). The Caveats of observing Inter-Trial Phase-Coherence in Cognitive Neuroscience. Sci Rep, 8 (1 ), 2990. 10.1038/s41598-018-20423-z 29445210
van Dijk H. , van der Werf J. , Mazaheri A. , Medendorp W. P. , & Jensen O. (2010). Modulations in oscillatory activity with amplitude asymmetry can produce cognitively relevant event-related responses. Proc Natl Acad Sci U S A, 107 (2 ), 900–905. 10.1073/pnas.0908821107 20080773
Voss J. L. , & Federmeier K. D. (2011). FN400 potentials are functionally identical to N400 potentials and reflect semantic processing during recognition testing. Psychophysiology, 48 (4 ), 532–546. 10.1111/j.1469-8986.2010.01085.x 20701709
Voss J. L. , Lucas H. D. , & Paller K. A. (2010). Conceptual priming and familiarity: different expressions of memory during recognition testing with distinct neurophysiological correlates. J Cogn Neurosci, 22 (11 ), 2638–2651. 10.1162/jocn.2009.21341 19702474
Voss J. L. , Lucas H. D. , & Paller K. A. (2012). More than a feeling: Pervasive influences of memory without awareness of retrieval. Cogn Neurosci, 3 (3–4 ), 193–207. 10.1080/17588928.2012.674935 24171735
Voss J. L. , & Paller K. A. (2010). What makes recognition without awareness appear to be elusive? Strategic factors that influence the accuracy of guesses. Learn Mem, 17 (9 ), 460–468. https://doi.org/17/9/460 [pii] 10.1101/lm.1896010 20810621
Wagenmakers E. J. , Marsman M. , Jamil T. , Ly A. , Verhagen J. , Love J. , Selker R. , Gronau Q. F. , Šmíra M. , Epskamp S. , Matzke D. , Rouder J. N. , & Morey R. D. (2018). Bayesian inference for psychology. Part I: Theoretical advantages and practical ramifications. Psychon Bull Rev, 25 (1 ), 35–57. 10.3758/s13423-017-1343-3 28779455
Wais P. E. , Wixted J. T. , Hopkins R. O. , & Squire L. R. (2006). The hippocampus supports both the recollection and the familiarity components of recognition memory. Neuron, 49 (3 ), 459–466. 10.1016/j.neuron.2005.12.020 16446148
Watrous A. J. , & Buchanan R. J. (2020). The Oscillatory ReConstruction Algorithm (ORCA) adaptively identifies frequency bands to improve spectral decomposition in human and rodent neural recordings. J Neurophysiol. 10.1152/jn.00292.2020
Watrous A. J. , Deuker L. , Fell J. , & Axmacher N. (2015). Phase-amplitude coupling supports phase coding in human ECoG. Elife, 4 . 10.7554/eLife.07886
Watrous A. J. , Fell J. , Ekstrom A. D. , & Axmacher N. (2015). More than spikes: common oscillatory mechanisms for content specific neural representations during perception and memory. Curr Opin Neurobiol, 31 , 33–39. 10.1016/j.conb.2014.07.024 25129044
Watrous A. J. , Miller J. , Qasim S. E. , Fried I. , & Jacobs J. (2018). Phase-tuned neuronal firing encodes human contextual representations for navigational goals. Elife, 7 , e32554. 10.7554/eLife.32554 29932417
Whittlesea B. W. , & Williams L. D. (2000). The source of feelings of familiarity: the discrepancy-attribution hypothesis. J Exp Psychol Learn Mem Cogn, 26 (3 ), 547–565. http://www.ncbi.nlm.nih.gov/pubmed/10855417 10855417
Whittlesea B. W. A. , & Williams L. D. (2001). The Discrepancy-Attribution Hypothesis: II. Expectation, Uncertainty, Surprise, and Feelings of Familiarity [Article]. Journal of Experimental Psychology: Learning Memory and Cognition, 27 (1 ), 14–33. 10.1037/0278-7393.27.1.14 11204095
Wilding E. L. (1999). Separating retrieval strategies from retrieval success: an event-related potential study of source memory. Neuropsychologia, 37 (4 ), 441–454. 10.1016/s0028-3932(98)00100-6 10215091
Wilding E. L. (2006). The practice of rescaling scalp-recorded event-related potentials. Biol Psychol, 72 (3 ), 325–332. 10.1016/j.biopsycho.2005.12.002 16442202
Wilding E. L. , Doyle M. C. , & Rugg M. D. (1995). Recognition memory with and without retrieval of context: an event-related potential study. Neuropsychologia, 33 (6 ), 743–767. https://doi.org/0028-3932(95)00017-W [pii]7675165
Wilding E. L. , & Rugg M. D. (1996). An event-related potential study of recognition memory with and without retrieval of source. Brain, 119 ( Pt 3 ), 889–905. http://www.ncbi.nlm.nih.gov/pubmed/8673500 8673500
Wixted J. T. (2007). Dual-process theory and signal-detection theory of recognition memory. Psychol Rev, 114 (1 ), 152–176. 10.1037/0033-295X.114.1.152 17227185
Wixted J. T. , & Mickes L. (2010). A continuous dual-process model of remember/know judgments. Psychol Rev, 117 (4 ), 1025–1054. 10.1037/a0020874 20836613
Wixted J. T. , & Squire L. R. (2004). Recall and recognition are equally impaired in patients with selective hippocampal damage [Review]. Cogn Affect Behav Neurosci, 4 (1 ), 58–66. http://www.ncbi.nlm.nih.gov/pubmed/15259889 15259889
Wixted J. T. , & Squire L. R. (2010). The role of the human hippocampus in familiarity-based and recollection-based recognition memory. Behav Brain Res, 215 (2 ), 197–208. 10.1016/j.bbr.2010.04.020 20412819
Wixted J. T. , & Squire L. R. (2011a). The familiarity/recollection distinction does not illuminate medial temporal lobe function: response to Montaldi and Mayes. Trends Cogn Sci, 15 (8 ), 340–341. 10.1016/j.tics.2011.06.006 21763175
Wixted J. T. , & Squire L. R. (2011b). The medial temporal lobe and the attributes of memory. Trends Cogn Sci, 15 (5 ), 210–217. 10.1016/j.tics.2011.03.005 21481629
Wolk D. A. , Schacter D. L. , Lygizos M. , Sen N. M. , Chong H. , Holcomb P. J. , Daffner K. R. , & Budson A. E. (2007). ERP correlates of Remember/Know decisions: association with the late posterior negativity. Biol Psychol, 75 (2 ), 131–135. 10.1016/j.biopsycho.2007.01.005 17329007
Woodruff C. C. , Hayama H. R. , & Rugg M. D. (2006). Electrophysiological dissociation of the neural correlates of recollection and familiarity. Brain Res, 1100 (1 ), 125–135. 10.1016/j.brainres.2006.05.019 16774746
Yassa M. A. , & Stark C. E. (2011). Pattern separation in the hippocampus [Research Support, N.I.H., Extramural Review]. Trends in Neurosciences, 34 (10 ), 515–525. 10.1016/j.tins.2011.06.006 21788086
Yeh N. , & Koen J. D. (2023). Informative and uninformative prestimulus cues at encoding benefit familiarity and source memory. Memory, 31 (3 ), 367–379. 10.1080/09658211.2022.2156546 36546483
Yonelinas . (2001). Consciousness, control, and confidence: The 3 Cs of recognition memory. Journal of Experimental Psychology: General, 130 (3 ), 361–379. 10.1037//0096-3445.130.3.361 11561915
Yonelinas A. P. (1994). Receiver-operating characteristics in recognition memory: Evidence for a dual-process model. Journal of Experimental Psychology: Learning, Memory, and Cognition, 20 (6 ), 1341–1354. 10.1037/0278-7393.20.6.1341 7983467
Yonelinas A. P. (1999). The contribution of recollection and familiarity to recognition and source-memory judgments: a formal dual-process model and an analysis of receiver operating characteristics. J Exp Psychol Learn Mem Cogn, 25 (6 ), 1415–1434. http://www.ncbi.nlm.nih.gov/pubmed/10605829 10605829
Yonelinas A. P. (2001). Components of episodic memory: the contribution of recollection and familiarity. Philos Trans R Soc Lond B Biol Sci, 356 (1413 ), 1363–1374. 10.1098/rstb.2001.0939 11571028
Yonelinas A. P. (2002). The Nature of Recollection and Familiarity: A Review of 30 Years of Research. Journal of Memory and Language, 46 (3 ), 441–517. 10.1006/jmla.2002.2864
Yonelinas A. P. (2009). Developing theories that bridge mind and brain: Some thoughts of a cognitive psychologist. In Rösler F. , Ranganath C. , Röder B. , & Kluwe R. (Eds.), Neuroimaging of Human Memory: Linking cognitive processes to neural systems (pp. 0). Oxford University Press. 10.1093/acprof:oso/9780199217298.003.0019
Yonelinas A. P. , Aly M. , Wang W. C. , & Koen J. D. (2010). Recollection and familiarity: examining controversial assumptions and new directions. Hippocampus, 20 (11 ), 1178–1194. 10.1002/hipo.20864 20848606
Yonelinas A. P. , & Jacoby L. L. (1996). Noncriterial recollection: familiarity as automatic, irrelevant recollection. Conscious Cogn, 5 (1–2 ), 131–141. 10.1006/ccog.1996.0008
Yonelinas A. P. , Kroll N. E. , Quamme J. R. , Lazzara M. M. , Sauve M. J. , Widaman K. F. , & Knight R. T. (2002). Effects of extensive temporal lobe damage or mild hypoxia on recollection and familiarity. Nat Neurosci, 5 (11 ), 1236–1241. 10.1038/nn961 12379865
Yonelinas A. P. , & Parks C. M. (2007). Receiver operating characteristics (ROCs) in recognition memory: a review. Psychol Bull, 133 (5 ), 800–832. 10.1037/0033-2909.133.5.800 17723031
Yonelinas A. P. , Quamme J. R. , Widaman K. F. , Kroll N. E. , Sauve M. J. , & Knight R. T. (2004). Mild hypoxia disrupts recollection, not familiarity [Review]. Cogn Affect Behav Neurosci, 4 (3 ), 393–400; discussion 401–406. http://www.ncbi.nlm.nih.gov/pubmed/15535174 15535174
Yonelinas A. P. , Ranganath C. , Ekstrom A. D. , & Wiltgen B. J. (2019). A contextual binding theory of episodic memory: systems consolidation reconsidered. Nat Rev Neurosci. 10.1038/s41583-019-0150-4
Yong E. (2012). Replication studies: Bad copy. Nature, 485 (7398 ), 298–300. 10.1038/485298a 22596136
Yu S. S. , & Rugg M. D. (2010). Dissociation of the electrophysiological correlates of familiarity strength and item repetition. Brain Res, 1320 , 74–84. 10.1016/j.brainres.2009.12.071 20051232
Zhang Y. , Hedo R. , Rivera A. , Rull R. , Richardson S. , & Tu X. M. (2019). Post hoc power analysis: is it an informative and meaningful analysis? Gen Psychiatr, 32 (4 ), e100069. 10.1136/gpsych-2019-100069 31552383
