Abstract

Like yin and yang, novelty and familiarity are typically described as separate-yet-complementary aspects of an experience, two ends of a single continuum. However, novelty and familiarity are also multifaceted. For instance, novelty can sometimes result in enhanced mnemonic performance, whereas at other times familiarity is better remembered. As previous investigations focused primarily on the experimental aspect of novelty, the mechanisms supporting conceptual novelty (the novel combination of two previously unrelated existing concepts) remain unclear. Importantly, conceptual novelty can be recognized as such only when compared with preexperimental familiar knowledge, regardless of experimental status. Here we applied a combined repetition suppression/subsequent memory fMRI paradigm, focusing on the conceptual aspect of novelty and familiarity as the subject matter. Conceptual novelty was characterized by sustained neural activity; familiarity, on the other hand, exhibited repetition effects in multiple cortical regions, a subset of which was modulated by successful encoding. Subsequent memory of novelty was associated only with activation differences in a distinct set of regions, including the hippocampus and medial cortical regions. These results suggest that conceptual novelty (a) does not (easily) trigger the repetition suppression phenomenon but requires sustained neural recruitment and (b) activates dedicated encoding mechanisms. Conceptual familiarity, in contrast, allows rapid neural processing that depends upon existing neural representations. Overall, these findings challenge the definition of novelty as a unitary concept. Furthermore, they bear important implications for research into the neural bases of knowledge representation and recognition memory.

INTRODUCTION

Familiarity breeds contempt, while rarity winsadmiration

Apuleius

Our remarkable ability to rapidly categorize events in our environment as old or new underlies many of our daily actions. Novelty identification allows fast solicitation of novel opportunities that may result in beneficial outcomes or of new stimuli that may pose a threat (Schomaker & Meeter, 2015; Biederman & Vessel, 2006; Knight, 1996).Familiarity and repetition, on the other hand, facilitate perception and access to information (Zhao, Al-Aidroos, & Turk-Browne, 2013; Alba & Hasher, 1983; Bransford & Johnson, 1972). Interestingly, whereas familiarity and novelty are inherently opposing forces, both have been associated with improved memory (Schomaker & Meeter, 2015; van Kesteren, Ruiter, Fernández, & Henson, 2012; Tulving & Kroll, 1995; Alba & Hasher, 1983; Schulman, 1974). Here we propose that differences in operational definitions of familiarity and novelty have been underappreciated in past studies. These differences may have contributed to the conflicting behavioral results and the neural processes associated with novelty and familiarity. Focusing on the conceptual aspect of novelty—the combination of preexisting yet unrelated concepts—the current investigation seeks to identify the neural systems supporting processing and subsequent memory of familiarity and novelty.

Most previous studies of novelty and familiarity focused on experimental history as the defining characteristic, whereby items were deemed as familiar if they have been presented in the experiment or novel if they have not been presented (Kafkas & Montaldi, 2014; Tulving, Markowitsch, Craik, Habib, & Houle, 1996). However, experimental novelty refers only to an episodic recency manipulation; the intuitive definition of novelty as a new idea or combination or previous ideas is not captured by these studies. We argue that the latter notion of novelty can be investigated by conceptual novelty, formed by the combination of preexisting concepts that are unrelated or even incongruent in relation to each other (e.g., “bad forehead” or “purple banana”; van Kesteren et al., 2012; Wisniewski, 1997). We were specifically interested in the processes associated with initial exposure to conceptually novel (or familiar) information, rather than the prolonged gradual acquisition of new knowledge that has already been explored (Barron, Dolan, & Behrens, 2013; McClelland, 2013; Zeithamova, Dominick, & Preston, 2012; Kumaran, Summerfield, Hassabis, & Maguire, 2009; McClelland, McNaughton, & O'Reilly, 1995).

A straightforward account of conceptual novelty and familiarity processing would suggest that the two might be distinguished by the same neural processes as their experimental counterparts (e.g., Gilmore, Nelson, & McDermott, 2015; Kafkas & Montaldi, 2014). Briefly, novelty and familiarity can be represented as two ends of a monotonic continuum, with novelty inducing increased activity in the recently suggested “parietal memory network” and in additional cortical and subcortical regions (Gilmore et al., 2015; Kumaran & Maguire, 2009; Tulving et al., 1996). The single continuum mechanism fits nicely with the improved memory typically observed for experimental novelty, particularly in the medial temporal lobes (MTL; Schomaker & Meeter, 2015; Kirchhoff, Wagner, Maril, & Stern, 2000; Tulving et al., 1996).

However, some findings are difficult to integrate in a single continuum account. Specifically, behavioral studies demonstrate a long-term memory benefit for semantically familiar or repeating events, suggesting that a single monotonic continuum with novelty at its apex may not be the only mechanism at play (Bein et al., 2015; Reder et al., 2013; Kim, Yi, Raye, & Johnson, 2012; Craik & Tulving, 1975; Schulman, 1974; Bransford & Johnson, 1972; Ebbinghaus, 1885). Additionally, initial neural evidence hint that experimental novelty and familiarity elicit unique signals in separate brain regions, stressing the distinct roles of novelty detection and reactivation of existing representations (Kafkas & Montaldi, 2014). A parallel spatial distinction has been recently suggested for the neural substrates of conceptual novelty and familiarity (van Kesteren et al., 2012), but the relation of these unique signals to subsequent memory has received little systematic investigations (but see van Kesteren et al., 2013).

The scarcity of reported differences between neural activations for conceptual novelty and familiarity (Maril et al., 2011; Staresina, Gray, & Davachi, 2009) may be explained by the paradigms utilized. Typically, the neural responses for novelty and familiarity are examined as a function of a single presentation of each stimulus. However, novelty and, in particular, conceptual novelty might require the construction of new representations, a process that has been suggested to be evident only in repetition-related neural effects (Segaert, Weber, de Lange, Petersson, & Hagoort, 2013). Indeed, enhanced sustained neural recruitment for novel, devoid-of-meaning stimuli was demonstrated upon repeated presentations of a novel event (Gagnepain et al., 2008; Henson, Shallice, & Dolan, 2000). Similarly, semantically related word pairs (i.e., familiar associations) exhibited more repetition suppression (RS) than semantically unrelated pairs (Gold et al., 2006). Therefore, repetition effects may assist in identifying the distinct neural mechanisms associated with conceptual novelty and familiarity.

To test whether conceptually novel and familiar events are associated with unique neural signals and to characterize the relation of such signals to subsequent long-term memory, we utilized a combined RS/subsequent memory (difference in memory [DM]) paradigm (Manelis, Wheeler, Paynter, Storey, & Reder, 2011; Turk-Browne, Yi, & Chun, 2006). DM analyses bin study events based on whether they were later remembered or forgotten in a subsequent memory test (Wagner et al., 1998). RS paradigms are characterized by a relative suppression of the activity of neural ensembles in response to repetitions of a stimulus or a process to which they recently responded. Despite ambiguity in the biophysical mechanism that underlie fMRI RS (Gotts, Chow, & Martin, 2012; Grill-Spector, Henson, & Martin, 2006), a careful experimental design allows for inferences to be made about the underlying neuronal representations; in the current case, the use of the RS paradigm enabled us to (a) test the effects of conceptual novelty and familiarity while controlling for experimental novelty, (b) examine the dynamics of sustained activation associated with a repeated presentation of events (RS effects), and (c) examine the relation of this sustained activation to subsequent memory. Importantly, this design allowed us to orthogonally manipulate the experimental episodic aspect and the conceptual aspects of stimuli; both conceptual novelty and familiarity could be either experimentally novel (first presentation) or familiar (second presentation).

To manipulate conceptual familiarity and novelty, we utilized a modified version of the “semantic congruency paradigm.” Familiarity was operationalized as events that are congruent with previous knowledge (repetition of prior experiences; e.g., “yellow banana”), and novelty as events that are incongruent with it (no previous exposure to the specific word combination; e.g., “purple banana”; Bein et al., 2015; Craik & Tulving, 1975; Schulman, 1974). Previous neuroimaging studies of this paradigm have generally demonstrated reduced memory for conceptual novel (vs. familiar) events, which may have resulted in weaker statistical power. To allow for an unbiased interpretation of the results, we modified the encoding task to produce similar memory rates for familiar and novel trials. This modification additionally equated the behavioral responses associated with novelty and familiarity during the encoding task; thus, we were able to rule out alternative explanations attributing different neural patterns to different behavioral responses (different behavioral responses have been demonstrated to result in different neural and behavioral effects in congruency and RS paradigms; e.g., Bein et al., 2015; Moore, Yi, & Chun, 2013; Henson, Shallice, Gorno-Tempini, & Dolan, 2002).

The combined RS/DM paradigm enabled us to explore three distinct neural effects that might differentiate conceptual novelty and familiarity: (1) activation effects (differences in neural activation between conceptual familiarity and novelty), (2) repetition effects (differences between conceptual familiarity and novelty in the amount of RS they induce), (3) successful encoding effects, manifested either as (3a) a difference in activation between subsequently remembered and forgotten events (DM analysis) or (3b) a difference in repetition effects between subsequently remembered and forgotten events (RS–DM analysis, which is obtained by the combination of RS and subsequent memory performance; Turk-Browne et al., 2006). Beyond differences between conceptual novelty and familiarity, this design allowed the disentanglement of conceptual and experimental novelty: If novelty is a unitary construct (i.e., conceptual and experimental novelties are processed similarly), we should observe enhanced activation for both conceptual novelty (regardless of repetitions) and experimental novelty (first presentation of events compared with their second presentation; an RS effect). However, if conceptual novelty taps a unique mechanism, there should be neural effects associated specifically with conceptual novelty. Altogether, these measures allowed us to thoroughly investigate multiple aspects of the neural phenomena that underlie the human capacity to represent and learn conceptually novel and familiar stimuli.

METHODS

Participants

Nineteen right-handed native Hebrew speakers participated in this study (seven women; mean age = 25.9 years, range = 21–31 years). Six additional participants were excluded because they remembered or forgot an insufficient number of trials (less than 10 items; four participants), exhibited excessive movement (>2 mm; one participant), or exhibited significant benign anatomical irregularity that distorted acquisition (one participant). All participants had normal vision and were screened to ensure that they had no neurological conditions or any other contraindications for MRI. The participants were paid for their participation in the experiment. Informed consent was obtained before participation in a manner approved by the Tel-Aviv Sourasky Medical Center Ethics Committee and the Hebrew University institutional review board.

Stimulus Materials

Overall, 352 items (Hebrew nouns) were paired with contexts (adjectives) to form target word pairs. Conceptual familiarity was operationalized as semantically congruent item–context pairs (e.g., “yellow banana”). Novel pairs were constructed by re-pairing the nouns and adjectives from the familiar pairs to form novel incongruous or unrelated combinations (i.e., each adjective that appeared in a familiar pair with one noun, e.g., “yellow BANANA,” was also used to construct a novel pair with another noun, e.g., “yellow ZEBRA”). To allow a sufficient number of trials, we included item pairs that were either strongly or weakly congruent or incongruent (e.g., strong: “yellow/purple banana”; weak: “wrinkled/bad forehead”; for a recent indication of similarity in their neural underpinnings, see Jackson, Hoffman, Pobric, & Lambon Ralph, 2015). This procedure resulted in a total of 352 familiar and 352 novel word pairs with each word (item or context) included once in a familiar and once in a novel word pair. The conceptual status (familiar/novel) of each pair was assessed in pilot studies using independent groups of participants. In these studies, each participant provided congruency ratings for about 120 unique noun–adjective pairs (each participant saw a specific noun only once), using a 1–4 scale (1 = totally incompatible, 2 = somewhat compatible, 3 = highly compatible, 4 = automatically associated, with the instruction to choose the last category whenever “reading the noun leads to the automatic retrieval of the adjective”). Each unique word pair was assessed by an average of 14.26 raters (SD = 0.82; range = 12–16). Noun–adjective pairs were classified as conceptually familiar only if the two highest ratings (3–4) constituted more than 70% of the responses (average ± SD in final experimental list: 85.15 ± 13.12%); items were classified as conceptually novel if more than 70% of the responses were of the lowest rating (1; proportions in final experimental list: 89.11 ± 11.01%). The final experimental list included only nouns for which both familiar and novel pairings exceeded the criteria. See Appendix A for representative stimuli. Full stimuli list is available upon request. The nouns and adjectives included two to seven letters (mean: nouns, 4.16 letters; adjectives, 4.83 letters).

Across the four encoding sessions (see Procedure), each participant was presented with 176 unique events (word pairs), 88 familiar and 88 novel; each individual word (item/context) was presented only in one specific condition. One hundred seventy-six additional items served as foils in the recognition test. Across all participants, we counterbalanced items and contexts such that each word was assigned to each of the conditions (familiar and novel pairs at encoding and foils at recognition). Care was taken to ensure that the word length and single-word frequency (assessed by previously established norms; Frost & Plaut, 2005) did not significantly differ between the familiar and novel pairs used for each individual participant.

Two additional sets of noncounterbalanced stimuli were constructed for experimental control purposes and were not used in any of the analyses. The first set consisted of additional 72 noun–adjective pairs, 36 familiar and 36 novel, all of which were presented in reverse order. The second set consisted of 144 noun–noun pairs that were constructed to serve as fillers (for details, see Behavioral Procedure section). None of the nouns or adjectives included in these additional sets was used to construct the target stimuli.

Behavioral Procedure

See Figure 1 for schematic overview of task design. Before entering the MRI scanner, the participants received the task instructions and performed a brief practice session (the stimuli presented in this session were not presented in any of the experimental conditions). Each 2-sec trial included the verbal pair in white letters on a black background for 1.5 sec, followed by a 0.5-sec fixation cross. The words were presented one above the other, noun above adjective (following the Hebrew syntax, in which nouns precede adjectives). The participants were asked to decide whether two nouns appeared on the screen and to indicate their response via one of two possible response buttons (“yes”/“no”) located under the index and middle fingers of their left hands. This task was chosen to ensure processing of both words while reducing the attention given to the congruency status of the stimuli (as such allocation of attention might disproportionally enhance existing relations in the semantic network; Cohen, Rissman, Suthana, Castel, & Knowlton, 2014). Noticeably, this paradigm also dissociated response (“yes/no”) from the conceptual status of the stimuli (a confound typically found in investigations of the congruency effect and other novelty effects) using a manipulation orthogonal to the (novel/familiar) conceptual status.

Figure 1. 

A schematic overview of task design. Top: Incidental encoding phase. Each conceptually novel/familiar noun–adjective word pair was presented for 2 sec; participants were asked to indicate whether two nouns are presented on screen (separate two-noun trials were presented but not included in any of the analyses). Eighty-eight unique familiar and unique novel noun–adjective pairs were presented, with each pair repeated once (lag: 2–36.5 sec). Baseline crosshair fixation trials (mean length = 1.38 sec, range = 0.5–6 sec) were interspersed between trials. Bottom: At the recognition memory test, participants had to indicate for each of the 176 studied nouns and for additional 176 unstudied nouns whether it was previously studied or not. HC old = high confidence old; LC old = low confidence old.

Figure 1. 

A schematic overview of task design. Top: Incidental encoding phase. Each conceptually novel/familiar noun–adjective word pair was presented for 2 sec; participants were asked to indicate whether two nouns are presented on screen (separate two-noun trials were presented but not included in any of the analyses). Eighty-eight unique familiar and unique novel noun–adjective pairs were presented, with each pair repeated once (lag: 2–36.5 sec). Baseline crosshair fixation trials (mean length = 1.38 sec, range = 0.5–6 sec) were interspersed between trials. Bottom: At the recognition memory test, participants had to indicate for each of the 176 studied nouns and for additional 176 unstudied nouns whether it was previously studied or not. HC old = high confidence old; LC old = low confidence old.

The main experiment consisted of a single fMRI protocol separated into four scanning sessions. Each of the four sessions consisted of 142 trials, including 44 unique item–context pairs (22 familiar, 22 novel), each presented twice (for a total of 88 trials). The between-repetition lag varied between 2 and 36.5 sec (mean = 17.65 sec) and consisted of baseline fixations and experimental trials. Importantly, the correct response for all of these critical target trials was identical because they did not consist of two nouns. Noun–noun filler pairs (for which the correct response was “yes”) amounted to additional 36 trials per session. Each session also included 18 catch trial pairs that were presented in reversed order (relative to the target pairs), with the adjectives appearing above the nouns, to prevent the participants from basing their response only on the lower word (which differed between the experimental and filler trials).

Shortly after leaving the scanner (up to 15 min delay), the participants completed a self-paced surprise recognition test. The participants were presented with 352 nouns (half of them incidentally encoded as items paired with a novel or a familiar context adjective). For each item, the participants had to decide whether they were highly confident that the item appeared at encoding (HC), or they had low confidence it appeared (LC), or none at all (Forget trials; F), in accord with previous congruency and previous RS studies (Bein et al., 2015; Krebs, Boehler, De Belder, & Egner, 2015; Manelis et al., 2011; Staresina et al., 2009). The results of this test were used to classify both appearances of each item–context pair in the incidental encoding phase as subsequently remembered or forgotten (for a similar approach, see Kremers et al., 2014; Manelis et al., 2011; Turk-Browne et al., 2006).

Imaging Procedure

The participants were scanned in a 3T General Electric Signa Horizon LX 9.1 echo speed scanner (Milwaukee, WI). The experiment included four consecutive sessions in which whole-brain T2*-weighted EPI functional images were acquired, each consisting of 166 volumes (repetition time = 2500 msec, 200-mm field of view, 64 × 64 matrix, echo time = 35, flip angle = 81, 42 slices acquired perpendicular to the fourth ventricle collected in an interleaved manner starting from the uppermost slice, 3.125 × 3.125 × 3.1-mm [width × length × thickness] voxel size, no gap). In each session, experimental trials were intermixed with baseline fixation trials (mean length = 1.38 sec, range = 0.5–6 sec). The baseline trials comprised approximately 30% of each run. The order of the stimulus and baseline trials was determined by a sequencing algorithm (OptSeq; surfer.nmr.mgh.harvard.edu/optseq/) designed to maximize the efficiency of the event-related design (Dale, 1999). High-resolution anatomical images (SPGR; 156 [1-mm] slices) were obtained for each participant. Head motion was minimized using cushions arranged around each participant's head and by explicitly instructing the participants to remain still while in the scanner.

The data were preprocessed and analyzed using the SPM8 software (Wellcome Department of Cognitive Neurology, London, UK) for MATLAB (The MathWorks, Natick, MA). The images were corrected for differences in slice acquisition timing and realigned to the first slice image using second-degree B-spline interpolation. The data were spatially normalized to an EPI template in MNI305 stereotactic space using 16 nonlinear iterations and trilinear interpolation; the images were subsequently resampled into 2-mm cubic voxels. Next, the images were spatially smoothed using an 8-mm FWHM isotropic Gaussian kernel. The first four volumes from each run were discarded to allow for T1 equilibrium.

fMRI Analyses

For the purpose of the current investigation, subsequently remembered or forgotten trials were defined by HC and F responses at the recognition test. Our primary interest lay in eight conditions (2 × 2 × 2): Presentation (first/second) × Conceptual relation (familiar/novel) × Memory (remembered/forgotten). Statistical analyses of the imaging data were performed using a general linear model in which the onsets of (correctly classified) trials for each of these eight conditions were used to construct stick functions for eight regressors. Three additional regressors included noun–noun pairs, reversed noun–adjective pairs, and erroneous trials (trials that elicited no response or a wrong response). These 11 regressors and their temporal derivatives were convolved with the canonical hemodynamic response function for each session, and the four scanning sessions were combined at the participant level using SPM's first-level multisession protocol. The effects of the conditions of interest were estimated using a subject-specific fixed-effects model, with session-specific effects entered as additional regressors of no interest. Adding RTs to the first-level model (by modeling events as epochs whose durations were determined by RTs instead of zero duration stick functions; the variable epoch model; Grinband, Wager, Lindquist, Ferrera, & Hirsch, 2008)did not significantly alter the results.

For the purpose of the current analyses, subject-specific estimates of each of the eight main conditions of interest (2 [Presentation] × 2 [Conceptual relation] × 2 [Memory]) were derived and entered into a second-level analysis, in which the participants were treated as a random effect in a full factorial three-way ANOVA. For all analyses, we used a whole-brain level approach as previous implementations of similar congruency paradigms provided nonconvergent results in terms of the precise spatial loci of activations (van Kesteren et al., 2013; Maril et al., 2011; Naghavi, Eriksson, Larsson, & Nyberg, 2011; van Kesteren, Rijpkema, Ruiter, & Fernández, 2010; Staresina et al., 2009). Unless otherwise noted, all reported results include only activations that survived a threshold of p < .001 (uncorrected) at the voxel level and (cluster size) corrected to p < .05 at the cluster level. A corrected cluster size of 50 contiguous voxels was determined by Monte Carlo simulations (n iterations = 1000) with the current imaging and analysis parameters, as recommended by Slotnick, Moo, Segal, and Hart (2003). Activations in MTL regions were considered significant when eight or more contiguous voxels were active at p < .001, equivalent to a corrected probability level of p < .05 for the volume of the MTL (as available in AAL toolbox for SPM; Tzourio-Mazoyer et al., 2002) using another Monte Carlo simulation (implemented in AFNI's AlphaSim tool as available in REST toolbox; Song et al., 2011).

To test for similarities and differences in the processing and encoding of conceptual novelty and familiarity, two complementary approaches were employed (similarly toKafkas & Montaldi, 2014). First, repetition effects were examined by the (p < .05 cluster level-corrected, as described above) F contrasts comparing first and second exposure (collapsed across subsequent memory). To characterize possible similarities in repetition effects, we conjoined the repetition F contrasts for familiar and novel information with the conjunction null hypothesis (Friston, Penny, & Glaser, 2005).

To test for possible differences in repetition effects, a masking approach was employed. The cluster level-corrected results of each repetition F contrast (e.g., familiarity) were exclusively masked with the complementary repetition F contrast (e.g., novelty), thresholded at p < .05. This procedure resulted only in suprathreshold clusters for one repetition F contrast that did not exhibit activity at the complementary F contrast even at the uncorrected lenient threshold of p < .05. The results of this masking procedure were subsequently inclusively masked with the F contrast of the interaction between repetition (first vs. second presentation) and conceptual statues (familiarity vs. novelty), thresholded at p < .05. This masking procedure was elected because it enabled identification of voxels whose activity was significantly associated with one condition but did not overlap with those sensitive to the parallel contrast, indicating specificity of the effects elicited for one type of information.

These two approaches were also used to examine differences in encoding. To test for classic DM encoding effects (Wagner et al., 1998), we examined t contrasts between the first presentation of a remembered event and the first presentation of a forgotten event separately for familiar and novel events (classic differential memory [DM] analysis). In this analysis, we focused on first exposures for consistency with previous studies (Maril et al., 2011; Staresina et al., 2009; Turk-Browne et al., 2006); a parallel analysis conducted on second exposures revealed no statistically significant activations (thresholded at p < .005, uncorrected). Subsequently, we employed the conjunction analysis approach to explore similarities in encoding. To test for differences in encoding, we utilized the aforementioned masking approach.

Finally, to characterize differential repetition effects for subsequently remembered and forgotten events (RS–DM analysis), we limited the analysis (via masking) to regions that exhibited repetition effects (removing the mask did not significantly alter the results) and examined whether any clusters exhibited repetition-related effects for subsequently remembered familiar events with no repetition-related effects for subsequently forgotten familiar events (via exclusive masking at p < .05). To complement this analysis, we employed, separately for familiar and novel events, another inclusive masking in which the results of the classic DM analysis were inclusively masked with the results of the RS–DM analysis. This inclusive masking procedure was performed at p < .005 for each of the analyses, thereby yielding a conjoined probability (in the sense of the logical operator AND) of 0.000025. This procedure allowed us to identify regions that exhibited both DM and RS–DM encoding effects for familiar (or novel) events.

Parameter estimates, used for illustrative purposes, were subsequently extracted from the clusters that survived these procedures using a custom MATLAB script that defined clusters as all statistically significant voxels within a sphere with a radius of 8 mm surrounding the peak voxel identified in the analysis of interest (with a minimum distance of 8 mm between cluster peaks). The parameter estimates were then averaged across all voxels within a cluster for each individual for each experimental condition.

RESULTS

Nineteen participants were scanned in four sessions during which they were presented with a total of 176 unique noun–adjective combinations that were equally divided between familiar and novel pairs across sessions. Each combination was repeated twice (mean lag = 17.65 sec) in a rapid event-related design, and participants were required to process both words to successfully complete the behavioral task for each combination (see Behavioral Procedure section). A subsequent recognition test for the noun words was taken immediately after the participants left the scanner.

Behavioral

The summary of accuracy rates and RTs for the encoding task (“whether two nouns appear on the screen”) are presented in Table 1. Subsequent memory rates, including proportion and raw scores for the memory test, are included in Table 2. Across all analyses, classification of items to conceptually familiar and novel events was determined by the context that had been associated with specific items at encoding.

Table 1. 

Accuracy Rates and RT at the Incidental Encoding Task for Experimental Conditions of Interest

FamiliarNovel
Accuracy (%) 
First presentation 0.96 (0.01) 0.94 (0.01) 
Second presentation 0.98 (0.003) 0.97 (0.005) 
 
RT (msec) 
First presentation 1102 (38) 1169 (35) 
Second presentation 941 (32) 970 (35) 
FamiliarNovel
Accuracy (%) 
First presentation 0.96 (0.01) 0.94 (0.01) 
Second presentation 0.98 (0.003) 0.97 (0.005) 
 
RT (msec) 
First presentation 1102 (38) 1169 (35) 
Second presentation 941 (32) 970 (35) 

SEM in parentheses.

Table 2. 

Mean, Range, and Proportion of Response Scores Provided at the Subsequent Memory Test

HCLCNew
Familiar (88) 
Mean 32 (2.81) 11.95 (1.78) 41.42 (3.52) 
Range 14–59 1–27 17–70 
Proportion 0.37 (0.03) 0.14 (0.02) 0.48 (0.04) 
 
Novel (88) 
Mean 30.58 (3.03) 12 (1.86) 41.26 (3.58) 
Range 13–59 0–25 14–64 
Proportion 0.36 (0.04) 0.14 (0.02) 0.49 (0.04) 
 
Foil (176) 
Mean 25.89 (3.82) 24 (4.5) 123.64 (6.52) 
Range 7–71 1–63 65–155 
Proportion 0.15 (0.02) 0.14 (0.03) 0.71 (0.05) 
HCLCNew
Familiar (88) 
Mean 32 (2.81) 11.95 (1.78) 41.42 (3.52) 
Range 14–59 1–27 17–70 
Proportion 0.37 (0.03) 0.14 (0.02) 0.48 (0.04) 
 
Novel (88) 
Mean 30.58 (3.03) 12 (1.86) 41.26 (3.58) 
Range 13–59 0–25 14–64 
Proportion 0.36 (0.04) 0.14 (0.02) 0.49 (0.04) 
 
Foil (176) 
Mean 25.89 (3.82) 24 (4.5) 123.64 (6.52) 
Range 7–71 1–63 65–155 
Proportion 0.15 (0.02) 0.14 (0.03) 0.71 (0.05) 

Response scores are divided by conditions at encoding (familiar/novel) and test (foil). HC = high confidence old responses; LC = low confidence old responses; New = new responses. Parentheses for condition labels indicate the maximal number of items included in each condition (see Results section text for details). SEM in parentheses.

Accuracy

A 2 (Familiarity/novelty) by 2 (Repetition) repeated-measures ANOVA with the accuracy rate in the incidental encodingtask as the dependent variable revealed significant effects of repetition, indicating improved accuracy on the second presentation, F(1, 18) = 11.88, p < .003, ηp2 = 0.4, and of conceptual relation (more accurate responses for familiar pairs), F(1, 18) = 14.83, p < .001, ηp2 = 0.45, along with a significant interaction, F(1, 18) = 6.94, p < .02, ηp2 = 0.28. Although significant differences were observed, the performance in all conditions of interest was at ceiling (see Table 1). Because erroneous trials were omitted from subsequent analyses, these differences did not affect the imaging results.

RT

Both familiar and novel pairs were associated with significant repetition priming (familiar: t(18) = 12; novel: t(19) = 15.32; both p values < .001). Novel pairs were associated with increased priming (161 and 199 msec for familiar and novel events, respectively). There was a significant interaction between conceptual relation (familiarity/novelty) and repetition, F(1, 18) = 9.11, ηp2 = 0.34, p < .007. Main effects for conceptual relation and repetition were also observed, F(1, 19) = 27.81, ηp2 = 0.61, p < .001 and F(1, 19) = 238.2, ηp2 = 0.93, p < .001, respectively (see Table 1). Follow-up contrasts revealed, as expected, faster RTs for familiar stimuli during both the first (t(19) = 5.62, p < .001) and second (t(19) = 2.88, p < .01) presentations. Importantly, if repetition priming indicated neural RS (Schacter, Wig, & Stevens, 2007), novel pairs should be associated with increased RS. As will be evident from the imaging results presented below, this was not the case: Novel pairs were not associated with increased RS. Thus, the RT data cannot explain the obtained neuroimaging results.

Subsequent Memory

Memory performance was assessed via a recognition test for items (encoded nouns); this analysis included only items that were correctly identified at encoding (on average, 84.75 of 88 items per condition; range: 79–88). High-confidence (HC) recognition responses for old items (novel and familiar) significantly differed from new items (false alarms; 36.8% vs. 14.7%, t(18) = 11.12, p < .001), with response distribution similar to the that obtained in previous investigations of combined RS and DM effects (Manelis et al., 2011; Turk-Browne et al., 2006). Novel and familiar items did not differ in terms of their recognition scores (36% vs. 37%, corresponding to 30.58 and 32 items, respectively, t(19) < 1; see Table 2 for full data). The similar memory rates for novel and familiar items were expected, given the relatively shallow encoding task (Bein et al., 2015); importantly, this similarity in memory scores enabled unbiased interpretation of the subsequent imaging results.

Imaging

Overall, familiar events were associated with substantial reductions in neural activation upon their second presentation that were nearly absent for novel events (see Figure 2 for an overview of activity for first and second presentations compared with crosshair baseline fixations). This result was obtained in the face of similar activation for familiar and novel events in the first presentation (no cluster survived the statistical threshold for the comparison, similar to previous studies). In the following paragraphs, we provide a formal statistical analysis of these differences; pertinent to the current investigation, we also examine how successful encoding was associated with these differences. We start by describing the repetition and RS–DM effects for familiarity, follow with a description of the parallel effects for novelty (including effects common to both conceptual statuses), and conclude with the results of the classic DM analyses.

Figure 2. 

An overview of all activations (p < .05, cluster level-corrected, contrasted with crosshair baseline) for the first presentation of familiar events (top left), first presentation of novel events (top right), second presentation of familiar events (bottom left), and second presentation of novel events (bottom right). Left–right orientation in this and the following figures is displayed in neurological convention.

Figure 2. 

An overview of all activations (p < .05, cluster level-corrected, contrasted with crosshair baseline) for the first presentation of familiar events (top left), first presentation of novel events (top right), second presentation of familiar events (bottom left), and second presentation of novel events (bottom right). Left–right orientation in this and the following figures is displayed in neurological convention.

To identify differences in familiarity and novelty processing, we applied a masking procedure in which repetition (or encoding) effects for one type of conceptual relation (e.g., novelty: “purple banana”) were exclusively masked with effects for the other type of conceptual relation (familiarity; e.g., “yellow banana”; for details, see fMRI Analyses section).

Familiarity: Repetition and RS–DM Effects in Multiple Regions

Repetition-related effects were uniquely demonstrated for familiar word pairs in multiple regions. The left posterior inferior temporal gyrus and left anterior-ventral inferior frontal gyrus (avLIFG) exhibited RS, whereas a swath of regions, including the right parahippocampal gyrus, ventral and dorsal medial prefrontal cortices, and retrosplenial complex, was associated with repetition enhancement (RE; see Figure 3 and Table 3A). Importantly, there was a significant interaction between repetition and conceptual status in the vast majority of these regions (compare Table 3A with Table 4).

Figure 3. 

Clusters that exhibited repetition effects only for familiar events (at p < .05, cluster level-corrected) and parameter estimates extracted from the (A) left inferior temporal gyrus [−50, −48, −14], (B) left anterior temporal lobe [−56, −4, −30], (C) posterior cingulate cortex [8, −46, 36], and (D) right PHC [30, −20, −26]. The coordinates are reported in MNI space. Error bars in this and the following figures denote the SEM.

Figure 3. 

Clusters that exhibited repetition effects only for familiar events (at p < .05, cluster level-corrected) and parameter estimates extracted from the (A) left inferior temporal gyrus [−50, −48, −14], (B) left anterior temporal lobe [−56, −4, −30], (C) posterior cingulate cortex [8, −46, 36], and (D) right PHC [30, −20, −26]. The coordinates are reported in MNI space. Error bars in this and the following figures denote the SEM.

Table 3. 

Gray Matter Regions that Exhibited Significant RS and RS–DM Effects for Familiar but Not for Novel Events

RegionMNI CoordinatesZ ValueNo. of Voxels
xyz
A. Repetition Effects (RS or RE) 
Ventral and dorsal medial frontal cortices, superior frontal gyrus 12 62 28 5.61 3783 
L inferior frontal gyrus −46 28 4.86 988 
L angular gyrus −50 −76 34 4.83 352 
R angular/supramarginal gyrus, cuneus, precuneus, posterior cingulate gyrus 58 −54 26 4.55 3050 
L (anterior) inferior temporal gyrus −56 −4 −30 4.29 88 
L middle/superior frontal gyrus −30 36 50 4.26 119 
R cerebellum 38 −66 −30 4.04 110 
R cerebellum 22 −74 −54 3.88 83 
L lingual gyrus −18 −98 −10 3.99 149 
Dorsal medial frontal gyrus −4 18 48 3.81 73 
L (posterior) inferior temporal gyrus −50 −48 −16 3.71 79 
Middle occipital gyrus 32 −90 10 3.67 63 
R parahippocampal gyrus 30 −20 −26 3.57 
 
B. RS–DM Effects 
L inferior frontal gyrus −54 34 4.68 791 
R superior frontal gyrus 16 58 32 4.57 194 
R superior frontal gyrus 14 32 64 4.36 62 
R angular/supramarginal gyrus 58 −52 26 4.06 189 
 
C. Classic Subsequent Memory Effects 
L inferior frontal gyrus −56 30 10 3.34 63 
 
D. Conjunction of Repetition and DM Effects 
L inferior frontal gyrus −56 30 10 3.34 31 
RegionMNI CoordinatesZ ValueNo. of Voxels
xyz
A. Repetition Effects (RS or RE) 
Ventral and dorsal medial frontal cortices, superior frontal gyrus 12 62 28 5.61 3783 
L inferior frontal gyrus −46 28 4.86 988 
L angular gyrus −50 −76 34 4.83 352 
R angular/supramarginal gyrus, cuneus, precuneus, posterior cingulate gyrus 58 −54 26 4.55 3050 
L (anterior) inferior temporal gyrus −56 −4 −30 4.29 88 
L middle/superior frontal gyrus −30 36 50 4.26 119 
R cerebellum 38 −66 −30 4.04 110 
R cerebellum 22 −74 −54 3.88 83 
L lingual gyrus −18 −98 −10 3.99 149 
Dorsal medial frontal gyrus −4 18 48 3.81 73 
L (posterior) inferior temporal gyrus −50 −48 −16 3.71 79 
Middle occipital gyrus 32 −90 10 3.67 63 
R parahippocampal gyrus 30 −20 −26 3.57 
 
B. RS–DM Effects 
L inferior frontal gyrus −54 34 4.68 791 
R superior frontal gyrus 16 58 32 4.57 194 
R superior frontal gyrus 14 32 64 4.36 62 
R angular/supramarginal gyrus 58 −52 26 4.06 189 
 
C. Classic Subsequent Memory Effects 
L inferior frontal gyrus −56 30 10 3.34 63 
 
D. Conjunction of Repetition and DM Effects 
L inferior frontal gyrus −56 30 10 3.34 31 

(A) Repetition effects unique to familiar pairs. (B) Repetition effects unique to subsequently remembered familiar events. (C) Classic subsequent memory effects unique to familiar pairs. (D) Conjunctions of repetition and classic subsequent memory effects for familiar pairs (see also Figure 6). Z values are reported for repetition or subsequent memory effects for familiar stimuli. The coordinates in this and the following tables are reported in the MNI space.

Table 4. 

Coordinates and Z Values for Gray Matter Regions that Exhibited Significant Interaction between Repetition Effects and Conceptual Status (Familiarity vs. Novelty)

RegionMNI CoordinatesZ ValueNo. of Voxels
xyz
Ventral and dorsal medial frontal cortices, superior frontal gyrus 12 62 30 4.09 2134 
L inferior frontal gyrus −46 24 2.65 186 
L angular gyrus −48 −78 26 2.48 83 
R angular/supramarginal gyrus 58 −42 46 3.61 807 
R cuneus, precuneus 16 −78 32 3.51 539 
L precuneus −16 −94 36 3.45 82 
Posterior cingulate gyrus −42 24 2.25 121 
L (anterior) inferior temporal gyrus −54 −10 −30 2.82 53 
L lingual gyrus −26 −92 −8 2.29 59 
L (posterior) inferior temporal gyrus −50 −48 −16 3.71 79 
RegionMNI CoordinatesZ ValueNo. of Voxels
xyz
Ventral and dorsal medial frontal cortices, superior frontal gyrus 12 62 30 4.09 2134 
L inferior frontal gyrus −46 24 2.65 186 
L angular gyrus −48 −78 26 2.48 83 
R angular/supramarginal gyrus 58 −42 46 3.61 807 
R cuneus, precuneus 16 −78 32 3.51 539 
L precuneus −16 −94 36 3.45 82 
Posterior cingulate gyrus −42 24 2.25 121 
L (anterior) inferior temporal gyrus −54 −10 −30 2.82 53 
L lingual gyrus −26 −92 −8 2.29 59 
L (posterior) inferior temporal gyrus −50 −48 −16 3.71 79 

Interaction effects for subsequent memory resulted in regions identical to those reported in the main text; therefore, they are not reported.

Notably, a subset of regions, including the left inferior pFC and right angular gyrus, exhibited repetition effects for subsequently remembered but not for subsequently forgotten familiar events (RS–DM effects; Table 3B). In these regions, the successful encoding of familiar events, but not encoding failure, was associated with repetition effects. An inspection of the parameter estimates revealed that the first presentations of subsequently remembered and forgotten familiar events did not typically differ (i.e., no DM effect was observed; all p values > .01). Thus, subsequently forgotten familiar events were not significantly different at first presentation from subsequently remembered events, and the lack of repetition effects for subsequently forgotten events could not be attributed to floor effects.

Novelty: Sustained Activation—Nearly No Repetition Effects

The second presentation of a novel event was associated with similar levels of activation as its first presentation (Figure 2); no cluster emerged with repetition effects unique to novelty. Consequently, no region was detected in the RS–DM analysis for novel events.

To complement this result and examine whether novelty was responsible for any repetition effects, we employed a conjunction analysis to identify regions that exhibited repetition effects for both familiarity and novelty. This analysis revealed a single cluster in the left mid-dorsal inferior frontal gyrus, in which both conceptually familiar and novel word pairs elicited RS at the corrected statistical threshold (Figure 4 and Table 5A). Post hoc examination of the parameter estimates extracted from this cluster revealed a marginally significant difference in activation between familiar and novel items for both the first and second presentations (t(19) = 2.28 and t(19) = 2.26, respectively; all p values < .04). There was no interaction between repetition and conceptual status (F < 1), thereby indicating that conceptually familiar and novel stimuli were indeed processed similarly in this cluster. Critically, no region emerged in the RS–DM analysis when taking this cluster into account, suggesting that successful novelty encoding does not rely on repetition-related neural mechanisms.

Figure 4. 

Regions that exhibited repetition effects for familiar and novel events (conjunction analysis conducted across subsequent memory status; see fMRI Analyses section and Table 5 for details).

Figure 4. 

Regions that exhibited repetition effects for familiar and novel events (conjunction analysis conducted across subsequent memory status; see fMRI Analyses section and Table 5 for details).

Table 5. 

Gray Matter Regions that Exhibited RS and DM Effects for Novel Events

RegionMNI CoordinatesZ ValueNo. of Voxels
xyz
A. Conjunction: Repetitions Suppression for Familiar and Novel Pairs 
L inferior frontal gyrus −58 18 24 3.54 61 
 
B. Subsequent Memory Effects for Novel Pairs (Remembered > Forgotten) 
L anterior hippocampus −24 −6 −22 3.41 36 
 
C. Inverse Subsequent Memory Effects for Novel Pairs (Remembered < Forgotten) 
Ventral medial prefrontal cortex 12 46 3.73 363 
Dorsal medial prefrontal cortex 44 32 3.47 107 
Posterior cingulate −2 −26 36 3.43 209 
RegionMNI CoordinatesZ ValueNo. of Voxels
xyz
A. Conjunction: Repetitions Suppression for Familiar and Novel Pairs 
L inferior frontal gyrus −58 18 24 3.54 61 
 
B. Subsequent Memory Effects for Novel Pairs (Remembered > Forgotten) 
L anterior hippocampus −24 −6 −22 3.41 36 
 
C. Inverse Subsequent Memory Effects for Novel Pairs (Remembered < Forgotten) 
Ventral medial prefrontal cortex 12 46 3.73 363 
Dorsal medial prefrontal cortex 44 32 3.47 107 
Posterior cingulate −2 −26 36 3.43 209 

(A) The conjunction of repetition effects for familiar and novel events. (b) Positive subsequent memory effects unique to novel pairs. (C) Negative subsequent memory effects unique to novel pairs. Z values in B and C are reported for subsequent memory effects for novel stimuli.

Subsequent Memory Activation Effects: Novelty and Familiarity

Given that conceptually novel events were (episodically) encoded, the null results of the RS–DM analysis for novel events provided additional motivation for a classic DM analysis, in which the difference in activation for subsequently remembered and forgotten events was analyzed. This analysis revealed multiple regions that were exclusively involved in the encoding of novel word pairs (as indicated by the masking procedure). The classic DM effect (in which activation of the subsequently remembered event exceeds that of the subsequently forgotten event) was observed in the left anterior hippocampus (HPC; Figure 5A and Table 5B). An inspection of the parameter estimates for the HPC indicated that only the first presentation of subsequently remembered novel events was associated with significant activation (p < .001); activations for all other event types did not differ from the baseline (all p values > .2). Additional inverse DM effects were observed in the posterior cingulate cortex and medial prefrontal cortex (dorsal and ventral; Figure 5B–D and Table 5C). Note that inspection of Figure 5 might suggest that novel RS–DM effects (i.e., repetition effects for remembered but not forgotten novel effects) also exist. However, simple effects for the repetition of subsequently remembered novel events did not approach significance (p > .03, uncorrected, for all regions).

Figure 5. 

Clusters that exhibited DM effects at the first presentation only for novel events (at p < .05, cluster level-corrected) and parameter estimates extracted from (A) left HC [−24, −4, −22], (B) posterior cingulate cortex [−2, −26, 36], (C) dorsal medial prefrontal cortex [4, 44, 32], and (D) ventral medial prefrontal cortex [12, 46, 2]. Coordinates reported in MNI space.

Figure 5. 

Clusters that exhibited DM effects at the first presentation only for novel events (at p < .05, cluster level-corrected) and parameter estimates extracted from (A) left HC [−24, −4, −22], (B) posterior cingulate cortex [−2, −26, 36], (C) dorsal medial prefrontal cortex [4, 44, 32], and (D) ventral medial prefrontal cortex [12, 46, 2]. Coordinates reported in MNI space.

Interestingly, there was spatial dissociation between the encoding of novel and familiar events. Evidence for this dissociation was obtained from the conjunction of the two classic DM analyses, which yielded null results (even at a reduced threshold of p < .01).

The classic DM analysis for familiar events indicated that a single region was uniquely associated with DM for familiar events: a specific cluster in the avLIFG (Figure 6 and Table 3C), in agreement with previous investigations (Maril et al., 2011; Staresina et al., 2009). The parameter estimates extracted from this region revealed that the DM effect was evident only during the first presentation of familiar pairs (all other DM p values > .2). A conjunction of the classic DM analysis with the repetition effects analysis confirmed that a subset of voxels within this cluster avLIFG demonstrated significant RS and DM effects for familiar word pairs. These results indicate that activation in the avLIFG during the first presentation of an event predicted subsequent memory (for familiar events): If the activation differed significantly from the baseline (as in remembered events), both DM and RS–DM effects would ensue; if not (as in forgotten events), no behavioral or neural effect would be evident (see Table 3D).

Figure 6. 

Results of the conjunction of the classic DM and repetition effects for familiar events and parameter estimates extracted from the avLIFG (see also Table 3D). Parameter estimates for familiar forgotten events did not demonstrate repetition effects (t values < 1), and there were no effects of DM or repetition for novel word pairs in this regions (p values > .1). Thus, activation at the first presentation of an event in the avLIFG was uniquely associated with subsequent memory for familiar events.

Figure 6. 

Results of the conjunction of the classic DM and repetition effects for familiar events and parameter estimates extracted from the avLIFG (see also Table 3D). Parameter estimates for familiar forgotten events did not demonstrate repetition effects (t values < 1), and there were no effects of DM or repetition for novel word pairs in this regions (p values > .1). Thus, activation at the first presentation of an event in the avLIFG was uniquely associated with subsequent memory for familiar events.

DISCUSSION

This study was designed to investigate the neural underpinnings of conceptual novelty and familiarity. For this purpose, we used word combinations that were either conceptually familiar (i.e., congruous, such as “yellow banana”) or novel (i.e., unrelated or incongruous, such as “bad forehead” or “purple banana”) in a combined RS/DM paradigm. This paradigm allowed us to manipulate conceptual as well as experimental novelty and familiarity. We found that conceptual novelty was generally associated with sustained activation; it exhibited almost no RS effects, in contrast to conceptual familiarity, which was accompanied by repetition effects in multiple regions. Successful encoding was dissociated as well: For conceptual familiarity, we observed (1) elevated activation at first exposure of subsequently remembered events (DM effects) and (2) repetition-related effects specific to successfully encoded events (RS–DM effects). Subsequent memory for novelty was evident only in DM effects that were spatially distinct from familiarity effects. Importantly, these differences were observed in the face of similar behavioral outcomes; thus, they cannot be attributed to differences in statistical power. Overall, these results suggest that conceptual familiarity and novelty are processed and encoded using distinct neural mechanisms. In the following sections, we highlight several innovative implications of these findings. We conclude by outlining possible directions for future investigations.

Novelty: Sustained Activation

Several recent theoretical and empirical frameworks have suggested that processing familiar and novel events requires partially distinct neural systems, in which a neural “hardwired” preference is specifically dedicated to novelty and its consequential prediction errors (Gilmore et al., 2015; Hawco & Lepage, 2014; Kafkas & Montaldi, 2014; Kumaran & Maguire, 2009; Henson et al., 2000; Tulving et al., 1996). Our findings differ from these previous investigations, which focused primarily on experimentally defined novelty—the lack of previous experimental exposure. At the first exposure of conceptual familiarity and novelty (i.e., when both were experimentally novel) no extensive differences in neural activation at were observed. At the second exposure (in which both conditions were experimentally familiar), we have found evidence for sustained activation of neural ensembles in response to conceptual novelty. No cluster demonstrated exclusive novelty-related repetition (see Figures 2 and 4), suggesting that conceptual novelty but not familiarity requires ongoing recruitment of resources (Segaert et al., 2013; Gotts et al., 2012; Grill-Spector et al., 2006). Given that RS is a ubiquitous phenomenon, the near absence of repetition effects specifically for conceptual novelty—operationalized here as two previously unrelated familiar words bound together—is an important contribution of this study. This result hints that conceptual novelty requires bypassing or aborting the ubiquitous processing principles of RS triggered by (conceptual) familiarity (for related findings, see Liu, Grady, & Moscovitch, 2016; Valyear, Gallivan, McLean, & Culham, 2012; Gold et al., 2006).

The difference in repetition effects may result from successful encoding of the conceptually familiar pair during the first presentation and its easier subsequent retrieval at the second presentation; the lack of repetition effects for conceptually novel items could thus reflect a failure to encode the conceptually novel items upon first presentation. However, two result patterns render this hypothesis unlikely. First, RT facilitation was evident for novel stimuli and to a larger extent than for familiar ones (Table 1), hinting that familiarity was not accompanied by easier and faster retrieval compared with novelty. Second, subsequent memory neural effects were observed only at the first and not at the second presentation of events, suggesting that the lack of repetition effects for novelty did not result from failure of encoding of novel items upon their first presentation.

Interestingly, some previous studies have found RE for novel stimuli; specifically, RE was demonstrated for novel stimuli that were completely devoid of meaning (abstract symbols or pseudowords; Gagnepain et al., 2008; Soldan, Zarahn, Hilton, & Stern, 2008; Henson et al., 2000). This difference from our findings sits well with the role ascribed to RE in the construction of new representations (Segaert et al., 2013): When no previous experience can support the processing of an event, as in stimuli devoid of meaning, RE is expected to occur. In the case of conceptually novel events, preexisting knowledge pertaining to the individual words of the pair does exist—it is the combination of these words that induces novelty. Thus, repetition effects might manifest in a gradual manner: In the absence of meaning, RE effects are predicted. When meaning is altogether familiar, RS occurs. In between—when previous knowledge can be recruited but through a novel relation-sustained activation will dominate.

Dual Route Encoding for Familiarity, but Not for Novelty

Successful encoding of familiar pairs was associated with both DM and RS–DM effects. Furthermore, DM effects for familiar items were evident only in a subset of the regions that exhibited RS–DM effects (Figure 6). This partial conjunction suggests that repetition and DM effects (only for familiar events) might be intertwined (see also Kremers et al., 2014; Turk-Browne et al., 2006). It is possible that an initial presentation provides a boost to an already existing representation, which is sufficient for the subsequent memory of familiarity to emerge; this boost would drive enhanced activation and repetition effects. According to this interpretation, novelty would require more than just a boost because it has no initial trace to begin with; thus, a more effortful process should occur, most likely via hippocampal-based mechanisms supporting the combination of two distinct items (Davachi, 2006), as evidenced by the hippocampal DM effect for novel events in the current study (see below). Regardless of the specific mechanisms, a single presentation of a familiar event triggered a cascade of RS and RS–DM effects; novelty encoding was unable to benefit from the potential contribution of the RS–DM supporting mechanism.

Parahippocampal Cortex Activation Profile: Potential Effects of Experimental Context on Familiarity

Our results also included RE only for familiar stimuli in the right parahippocampal cortex (PHC). The PHC is associated with RS for a wide variety of tasks (Kremers et al., 2014; Diana, Yonelinas, & Ranganath, 2012), although RE has also been observed (Greene & Soto, 2012; Düzel et al., 2003). Specifically, in experimental settings that include repetitions of old and new (spatial/conceptual) configurations, old configurations are associated with RE (Düzel et al., 2003). Our experiment included a similar setting of old and new (conceptual) configurations. The RE for familiar events could indicate sensitivity to the specific experimental context in which the familiar events were presented, thus allowing them to be differentiated from previous exposures; novel conceptual events might not require a similar mechanism because they have not been processed and do not require a similar differentiation according to context. This interpretation is in agreement with recent models that have assigned a contextual processing role to the PHC (Diana, Yonelinas, & Ranganath, 2007) and with recent behavioral studies that have demonstrated enhanced source memory for familiar items (Dewitt, Knight, Hicks, & Ball, 2012). Furthermore, this explanation suggests that previous PHC findings (observed in approximately similar coordinates) from paradigms that stress semantic relations might represent a general principle rather than a manifestation of the specific utilized stimuli or task, as has been suggested (see also Bein, Reggev, & Maril, 2014; van Kesteren et al., 2013).

Successful Encoding of Conceptual Novelty

Subsequent memory for novel events was associated with differential responses in the left anterior HPC (Figure 5). The HPC activity pattern is in agreement with previous evidence that the hippocampal complex participates in encoding and binding of novel items and associations (Kafkas & Montaldi, 2014; Kumaran & Maguire, 2009; Davachi, 2006; Pariente et al., 2005; Tulving et al., 1996), with its anterior portion indicated as particularly important for novelty encoding (Spaniol et al., 2009; Bernard et al., 2004; for a recent review, see Moscovitch, Cabeza, Winocur, & Nadel, 2016). Critically, the activations in our study cannot be attributed to the experimental novelty of an event because conceptually familiar events (although experimentally novel at their first presentation in our task) did not exhibit a similar DM effect. To the best of our knowledge, this is the first time that an HPC proper DM signal is observed for conceptually novel events, lending support to recent theories (Preston & Eichenbaum, 2013; van Kesteren et al., 2012).

The difference between conceptually novel and familiar events was also evident in cortical midline regions, where a reverse DM effect was observed primarily for conceptually novel events. Interestingly, reverse DM effects in these regions were previously demonstrated using familiar stimuli (Huijbers et al., 2012; Kim, Daselaar, & Cabeza, 2010). Using only familiar stimuli in previous investigations may have inadvertently enhanced the effect of built-in experimental novelty, which accompanies the (initial) presentation of any stimulus in an experimental context. The current experimental context (which included both conceptually familiar and novel events) may have diminished the importance of this experimental novelty in favor of its conceptual counterpart (Vannini, Hedden, Sullivan, & Sperling, 2013), resulting in DM effects which were limited in the midline and parietal regions to conceptually novel events.

Conclusions

Prior knowledge plays a key role in many cognitive processes. It directs allocation of attention (Zhao et al., 2013) and enables rapid processing of relevant information and encoding, retrieval, and integration of information (Schlichting & Preston, 2015; Ghosh & Gilboa, 2014). Importantly, prior knowledge serves as the background against which novel information can be understood; without this background, the resources required to process novelty (Kafkas & Montaldi, 2014; Kumaran & Maguire, 2009; Tulving et al., 1996) could not be easily allocated. Applying a careful operationalization of conceptual familiarity and novelty revealed that, whereas familiarity was associated with neural sensitivity in many cortical regions, no clusters demonstrated reliably exclusive repetition-related effects for novel events. We propose that designating conceptual novelty as the subject matter, rather than the typical episodically contextualized novelty, was the factor responsible for our novel findings. Our data contribute to the growing understanding that novelty is a multifaceted, nonunitary concept. The results strengthen our belief that conceptual novelty provides a vital perspective that is complementary to the other forms of novelty queried in previous studies.

Our results additionally suggest that encoding conceptually familiar events benefits from two neural mechanisms: activation differences at first exposure (DM effects) and differences that are evident only upon subsequent presentation (RS–DM effects). Novel events, which exhibited no evidence of RS–DM effects, require a neural route that is largely distinct from the processing route of an already established concept. The exact nature of this mnemonic trace, its temporal characteristics and persistence, and whether it represents the creation of a new unique entity that exceeds the sum of its components rather than simply an update of existing semantic knowledge remain topics for future research.

APPENDIX A

Representative list of 25 nouns presented in the current study alongside with the congruent (familiar) and in congruent (novel) adjectives with which they were presented at encoding (counterbalanced across participants; each participant was presented with a specific noun only once and only in a single conceptual condition).

NounCongruent AdjectiveIncongruent Adjective
Size Appropriate Winding 
Pea Green Appropriate 
Music Harmonic Green 
Shack Unstable Harmonic 
Snake Venomous Unstable 
Reception Bad Venomous 
Forehead Wrinkled Bad 
Lamp Shiny Wrinkled 
Pasta Curly Shiny 
Chicken Roasted Curly 
Tie Ironed Roasted 
Armor Metallic Ironed 
Pillow Fluffy Metallic 
Road Jammed Fluffy 
Husband Considerate Jammed 
Sand Rough Considerate 
Store Rented Rough 
Angle Sharp Rented 
Disc Scratched Sharp 
Lottery Biased Scratched 
Food Bland Biased 
Candle Lit Bland 
Percentage Minimal Lit 
Wine Intoxicating Minimal 
Path Winding Intoxicating 
NounCongruent AdjectiveIncongruent Adjective
Size Appropriate Winding 
Pea Green Appropriate 
Music Harmonic Green 
Shack Unstable Harmonic 
Snake Venomous Unstable 
Reception Bad Venomous 
Forehead Wrinkled Bad 
Lamp Shiny Wrinkled 
Pasta Curly Shiny 
Chicken Roasted Curly 
Tie Ironed Roasted 
Armor Metallic Ironed 
Pillow Fluffy Metallic 
Road Jammed Fluffy 
Husband Considerate Jammed 
Sand Rough Considerate 
Store Rented Rough 
Angle Sharp Rented 
Disc Scratched Sharp 
Lottery Biased Scratched 
Food Bland Biased 
Candle Lit Bland 
Percentage Minimal Lit 
Wine Intoxicating Minimal 
Path Winding Intoxicating 

Acknowledgments

This work was supported by Israeli Science Foundation (350/10) and National Institute for Psychobiology in Israel (232-13-14, both to A. M.). N. R. was supported by the Samuel and Lottie Rudin foundation, The Hoffman leadership and responsibility program, and the President of Israel's scholarship for scientific innovation.

Reprint requests should be sent to Anat Maril, Department of Psychology, The Hebrew University, Jerusalem, Israel, 91905, or via e-mail: anat.maril@mail.huji.ac.il.

REFERENCES

Alba
,
J. W.
, &
Hasher
,
L.
(
1983
).
Is memory schematic?
Psychological Bulletin
,
93
,
203
231
.
Barron
,
H. C.
,
Dolan
,
R. J.
, &
Behrens
,
T. E. J.
(
2013
).
Online evaluation of novel choices by simultaneous representation of multiple memories
.
Nature Neuroscience
,
16
,
1492
1498
.
Bein
,
O.
,
Livneh
,
N.
,
Reggev
,
N.
,
Gilead
,
M.
,
Goshen-Gottstein
,
Y.
, &
Maril
,
A.
(
2015
).
Delineating the effect of semantic congruency on episodic memory: The role of integration and relatedness
.
PloS One
,
10
,
e0115624
.
Bein
,
O.
,
Reggev
,
N.
, &
Maril
,
A.
(
2014
).
Prior knowledge influences on hippocampus and medial prefrontal cortex interactions in subsequent memory
.
Neuropsychologia
,
64
,
320
330
.
Bernard
,
F. A.
,
Bullmore
,
E. T.
,
Graham
,
K. S.
,
Thompson
,
S. A.
,
Hodges
,
J. R.
, &
Fletcher
,
P. C.
(
2004
).
The hippocampal region is involved in successful recognition of both remote and recent famous faces
.
Neuroimage
,
22
,
1704
1714
.
Biederman
,
I.
, &
Vessel
,
E. A.
(
2006
).
Perceptual pleasure and the brain
.
American Scientist
,
94
,
249
255
.
Bransford
,
J. D.
, &
Johnson
,
M. K.
(
1972
).
Contextual prerequisites for understanding: Some investigations of comprehension and recall
.
Journal of Verbal Learning and Verbal Behavior
,
11
,
717
726
.
Cohen
,
M. S.
,
Rissman
,
J.
,
Suthana
,
N. A.
,
Castel
,
A. D.
, &
Knowlton
,
B. J.
(
2014
).
Value-based modulation of memory encoding involves strategic engagement of fronto-temporal semantic processing regions
.
Cognitive, Affective & Behavioral Neuroscience
,
14
,
578
592
.
Craik
,
F. I. M.
, &
Tulving
,
E.
(
1975
).
Depth of processing and the retention of words in episodic memory
.
Journal of Experimental Psychology: General
,
104
,
268
294
.
Dale
,
A. M.
(
1999
).
Optimal experimental design for event-related fMRI
.
Human Brain Mapping
,
8
,
109
114
.
Davachi
,
L.
(
2006
).
Item, context and relational episodic encoding in humans
.
Current Opinion in Neurobiology
,
16
,
693
700
.
Dewitt
,
M. R.
,
Knight
,
J. B.
,
Hicks
,
J. L.
, &
Ball
,
B. H.
(
2012
).
The effects of prior knowledge on the encoding of episodic contextual details
.
Psychonomic Bulletin & Review
,
19
,
251
257
.
Diana
,
R. A.
,
Yonelinas
,
A. P.
, &
Ranganath
,
C.
(
2007
).
Imaging recollection and familiarity in the medial temporal lobe: A three-component model
.
Trends in Cognitive Sciences
,
11
,
379
386
.
Diana
,
R. A.
,
Yonelinas
,
A. P.
, &
Ranganath
,
C.
(
2012
).
Adaptation to cognitive context and item information in the medial temporal lobes
.
Neuropsychologia
,
50
,
3062
3069
.
Düzel
,
E.
,
Habib
,
R.
,
Rotte
,
M.
,
Guderian
,
S.
,
Tulving
,
E.
, &
Heinze
,
H.-J.
(
2003
).
Human hippocampal and parahippocampal activity during visual associative recognition memory for spatial and nonspatial stimulus configurations
.
Journal of Neuroscience
,
23
,
9439
9444
.
Ebbinghaus
,
H.
(
1885
).
Memory: A contribution to experimental psychology (translated 1913)
. In
H. A.
Ruger
&
C. E.
Bussenius
(Eds.).
New York, NY
:
Teachers College, Columbia University
.
Friston
,
K. J.
,
Penny
,
W. D.
, &
Glaser
,
D. E.
(
2005
).
Conjunction revisited
.
Neuroimage
,
25
,
661
667
.
Frost
,
R.
, &
Plaut
,
D.
(
2005
).
The word frequency database for printed Hebrew
. .
Gagnepain
,
P.
,
Chételat
,
G.
,
Landeau
,
B.
,
Dayan
,
J.
,
Eustache
,
F.
, &
Lebreton
,
K.
(
2008
).
Spoken word memory traces within the human auditory cortex revealed by repetition priming and functional magnetic resonance imaging
.
Journal of Neuroscience
,
28
,
5281
5289
.
Ghosh
,
V. E.
, &
Gilboa
,
A.
(
2014
).
What is a memory schema? A historical perspective on current neuroscience literature
.
Neuropsychologia
,
53
,
104
114
.
Gilmore
,
A. W.
,
Nelson
,
S. M.
, &
McDermott
,
K. B.
(
2015
).
A parietal memory network revealed by multiple MRI methods
.
Trends in Cognitive Sciences
,
19
,
534
543
.
Gold
,
B. T.
,
Balota
,
D. A.
,
Jones
,
S. J.
,
Powell
,
D. K.
,
Smith
,
C. D.
, &
Andersen
,
A. H.
(
2006
).
Dissociation of automatic and strategic lexical-semantics: Functional magnetic resonance imaging evidence for differing roles of multiple frontotemporal regions
.
Journal of Neuroscience
,
26
,
6523
6532
.
Gotts
,
S. J.
,
Chow
,
C. C.
, &
Martin
,
A.
(
2012
).
Repetition priming and repetition suppression: A case for enhanced efficiency through neural synchronization
.
Cognitive Neuroscience
,
3
,
227
237
.
Greene
,
C. M.
, &
Soto
,
D.
(
2012
).
Neural repetition effects in the medial temporal lobe complex are modulated by previous encoding experience
.
PloS One
,
7
,
e40870
.
Grill-Spector
,
K.
,
Henson
,
R.
, &
Martin
,
A.
(
2006
).
Repetition and the brain: Neural models of stimulus-specific effects
.
Trends in Cognitive Sciences
,
10
,
14
23
.
Grinband
,
J.
,
Wager
,
T. D.
,
Lindquist
,
M.
,
Ferrera
,
V. P.
, &
Hirsch
,
J.
(
2008
).
Detection of time-varying signals in event-related fMRI designs
.
Neuroimage
,
43
,
509
520
.
Hawco
,
C.
, &
Lepage
,
M.
(
2014
).
Overlapping patterns of neural activity for different forms of novelty in fMRI
.
Frontiers in Human Neuroscience
,
8
,
699
.
Henson
,
R. N. A.
,
Shallice
,
T.
, &
Dolan
,
R. J.
(
2000
).
Neuroimaging evidence for dissociable forms of repetition priming
.
Science
,
287
,
1269
1272
.
Henson
,
R. N. A.
,
Shallice
,
T.
,
Gorno-Tempini
,
M. L.
, &
Dolan
,
R. J.
(
2002
).
Face repetition effects in implicit and explicit memory tests as measured by fMRI
.
Cerebral Cortex
,
12
,
178
186
.
Huijbers
,
W.
,
Vannini
,
P.
,
Sperling
,
R. A.
,
Pennartz
,
C. M.
,
Cabeza
,
R.
, &
Daselaar
,
S. M.
(
2012
).
Explaining the encoding/retrieval flip: Memory-related deactivations and activations in the posteromedial cortex
.
Neuropsychologia
,
50
,
3764
3774
.
Jackson
,
R. L.
,
Hoffman
,
P.
,
Pobric
,
G.
, &
Lambon Ralph
,
M. A.
(
2015
).
The nature and neural correlates of semantic association versus conceptual similarity
.
Cerebral Cortex
,
25
,
4319
4333
.
Kafkas
,
A.
, &
Montaldi
,
D.
(
2014
).
Two separate, but interacting, neural systems for familiarity and novelty detection: A dual-route mechanism
.
Hippocampus
,
24
,
516
527
.
Kim
,
H.
,
Daselaar
,
S. M.
, &
Cabeza
,
R.
(
2010
).
Overlapping brain activity between episodic memory encoding and retrieval: Roles of the task-positive and task-negative networks
.
Neuroimage
,
49
,
1045
1054
.
Kim
,
K.
,
Yi
,
D.-J.
,
Raye
,
C. L.
, &
Johnson
,
M. K.
(
2012
).
Negative effects of item repetition on source memory
.
Memory & Cognition
,
40
,
889
901
.
Kirchhoff
,
B. A.
,
Wagner
,
A. D.
,
Maril
,
A.
, &
Stern
,
C. E.
(
2000
).
Prefrontal-temporal circuitry for episodic encoding and subsequent memory
.
Journal of Neuroscience
,
20
,
6173
6180
.
Knight
,
R. T.
(
1996
).
Contribution of human hippocampal region to novelty detection
.
Nature
,
383
,
256
259
.
Krebs
,
R. M.
,
Boehler
,
C. N.
,
De Belder
,
M.
, &
Egner
,
T.
(
2015
).
Neural conflict-control mechanisms improve memory for target stimuli
.
Cerebral Cortex
,
25
,
833
843
.
Kremers
,
N. A. W.
,
Deuker
,
L.
,
Kranz
,
T. A.
,
Oehrn
,
C.
,
Fell
,
J.
, &
Axmacher
,
N.
(
2014
).
Hippocampal control of repetition effects for associative stimuli
.
Hippocampus
,
24
,
892
902
.
Kumaran
,
D.
, &
Maguire
,
E. A.
(
2009
).
Novelty signals: A window into hippocampal information processing
.
Trends in Cognitive Sciences
,
13
,
47
54
.
Kumaran
,
D.
,
Summerfield
,
J. J.
,
Hassabis
,
D.
, &
Maguire
,
E. A.
(
2009
).
Tracking the emergence of conceptual knowledge during human decision making
.
Neuron
,
63
,
889
901
.
Liu
,
Z.-X.
,
Grady
,
C.
, &
Moscovitch
,
M.
(
2016
).
Effects of prior-knowledge on brain activation and connectivity during associative memory encoding
.
Cerebral Cortex
.
doi:10.1093/cercor/bhw047
.
Manelis
,
A.
,
Wheeler
,
M. E.
,
Paynter
,
C. A.
,
Storey
,
L.
, &
Reder
,
L. M.
(
2011
).
Opposing patterns of neural priming in same-exemplar vs. different-exemplar repetition predict subsequent memory
.
Neuroimage
,
55
,
763
772
.
Maril
,
A.
,
Avital
,
R.
,
Reggev
,
N.
,
Zuckerman
,
M.
,
Sadeh
,
T.
,
Sira
,
L. B.
, et al
(
2011
).
Event congruency and episodic encoding: A developmental fMRI study
.
Neuropsychologia
,
49
,
3036
3045
.
McClelland
,
J. L.
(
2013
).
Incorporating rapid neocortical learning of new schema-consistent information into complementary learning systems theory
.
Journal of Experimental Psychology. General
,
142
,
1190
1210
.
McClelland
,
J. L.
,
McNaughton
,
B. L.
, &
O'Reilly
,
R. C.
(
1995
).
Why there are complementary learning systems in the hippocampus and neocortex: Insights from the successes and failures of connectionist models of learning and memory
.
Psychological Review
,
102
,
419
457
.
Moore
,
K. S.
,
Yi
,
D.
, &
Chun
,
M.
(
2013
).
The effect of attention on repetition suppression and multivoxel pattern similarity
.
Journal of Cognitive Neuroscience
,
25
,
1305
1314
.
Moscovitch
,
M.
,
Cabeza
,
R.
,
Winocur
,
G.
, &
Nadel
,
L.
(
2016
).
Episodic memory and beyond: The hippocampus and neocortex in transformation
.
Annual Review of Psychology
,
67
,
105
134
.
Naghavi
,
H. R.
,
Eriksson
,
J.
,
Larsson
,
A.
, &
Nyberg
,
L.
(
2011
).
Cortical regions underlying successful encoding of semantically congruent and incongruent associations between common auditory and visual objects
.
Neuroscience Letters
,
505
,
191
195
.
Pariente
,
J.
,
Cole
,
S.
,
Henson
,
R. N. A.
,
Clare
,
L.
,
Kennedy
,
A.
,
Rossor
,
M.
, et al
(
2005
).
Alzheimer's patients engage an alternative network during a memory task
.
Annals of Neurology
,
58
,
870
879
.
Preston
,
A. R.
, &
Eichenbaum
,
H.
(
2013
).
Interplay of hippocampus and prefrontal cortex in memory
.
Current Biology
,
23
,
R764
R773
.
Reder
,
L. M.
,
Victoria
,
L. W.
,
Manelis
,
A.
,
Oates
,
J. M.
,
Dutcher
,
J. M.
,
Bates
,
J. T.
, et al
(
2013
).
Why it's easier to remember seeing a face we already know than one we don't: Preexisting memory representations facilitate memory formation
.
Psychological Science
,
24
,
363
372
.
Schacter
,
D. L.
,
Wig
,
G. S.
, &
Stevens
,
W. D.
(
2007
).
Reductions in cortical activity during priming
.
Current Opinion in Neurobiology
,
17
,
171
176
.
Schlichting
,
M. L.
, &
Preston
,
A. R.
(
2015
).
Memory integration: Neural mechanisms and implications for behavior
.
Current Opinion in Behavioral Sciences
,
1
,
1
8
.
Schomaker
,
J.
, &
Meeter
,
M.
(
2015
).
Short- and long-lasting consequences of novelty, deviance and surprise on brain and cognition
.
Neuroscience & Biobehavioral Reviews
,
55
,
268
279
.
Schulman
,
A. I.
(
1974
).
Memory for words recently classified
.
Memory & Cognition
,
2
,
47
52
.
Segaert
,
K.
,
Weber
,
K.
,
de Lange
,
F. P.
,
Petersson
,
K. M.
, &
Hagoort
,
P.
(
2013
).
The suppression of repetition enhancement: A review of fMRI studies
.
Neuropsychologia
,
51
,
59
66
.
Slotnick
,
S. D.
,
Moo
,
L. R.
,
Segal
,
J. B.
, &
Hart
,
J.
(
2003
).
Distinct prefrontal cortex activity associated with item memory and source memory for visual shapes
.
Cognitive Brain Research
,
17
,
75
82
.
Soldan
,
A.
,
Zarahn
,
E.
,
Hilton
,
H. J.
, &
Stern
,
Y.
(
2008
).
Global familiarity of visual stimuli affects repetition-related neural plasticity but not repetition priming
.
Neuroimage
,
39
,
515
526
.
Song
,
X.-W.
,
Dong
,
Z.-Y.
,
Long
,
X.-Y.
,
Li
,
S.-F.
,
Zuo
,
X.-N.
,
Zhu
,
C.-Z.
, et al
(
2011
).
REST: A toolkit for resting-state functional magnetic resonance imaging data processing
.
PloS One
,
6
,
e25031
.
Spaniol
,
J.
,
Davidson
,
P. S. R.
,
Kim
,
A. S. N.
,
Han
,
H.
,
Moscovitch
,
M.
, &
Grady
,
C. L.
(
2009
).
Event-related fMRI studies of episodic encoding and retrieval: Meta-analyses using activation likelihood estimation
.
Neuropsychologia
,
47
,
1765
1779
.
Staresina
,
B. P.
,
Gray
,
J. C.
, &
Davachi
,
L.
(
2009
).
Event congruency enhances episodic memory encoding through semantic elaboration and relational binding
.
Cerebral Cortex
,
19
,
1198
1207
.
Tulving
,
E.
, &
Kroll
,
N.
(
1995
).
Novelty assessment in the brain and long-term memory encoding
.
Psychonomic Bulletin & Review
,
2
,
387
390
.
Tulving
,
E.
,
Markowitsch
,
H. J.
,
Craik
,
F. E.
,
Habib
,
R.
, &
Houle
,
S.
(
1996
).
Novelty and familiarity activations in PET studies of memory encoding and retrieval
.
Cerebral Cortex
,
6
,
71
79
.
Turk-Browne
,
N. B.
,
Yi
,
D.-J.
, &
Chun
,
M. M.
(
2006
).
Linking implicit and explicit memory: Common encoding factors and shared representations
.
Neuron
,
49
,
917
927
.
Tzourio-Mazoyer
,
N.
,
Landeau
,
B.
,
Papathanassiou
,
D.
,
Crivello
,
F.
,
Etard
,
O.
,
Delcroix
,
N.
, et al
(
2002
).
Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain
.
Neuroimage
,
15
,
273
289
.
Valyear
,
K. F.
,
Gallivan
,
J. P.
,
McLean
,
D. A.
, &
Culham
,
J. C.
(
2012
).
fMRI repetition suppression for familiar but not arbitrary actions with tools
.
Journal of Neuroscience
,
32
,
4247
4259
.
van Kesteren
,
M. T. R.
,
Beul
,
S. F.
,
Takashima
,
A.
,
Henson
,
R. N.
,
Ruiter
,
D. J.
, &
Fernández
,
G.
(
2013
).
Differential roles for medial prefrontal and medial temporal cortices in schema-dependent encoding: From congruent to incongruent
.
Neuropsychologia
,
51
,
2352
2359
.
van Kesteren
,
M. T. R.
,
Rijpkema
,
M.
,
Ruiter
,
D. J.
, &
Fernández
,
G.
(
2010
).
Retrieval of associative information congruent with prior knowledge is related to increased medial prefrontal activity and connectivity
.
Journal of Neuroscience
,
30
,
15888
15894
.
van Kesteren
,
M. T. R.
,
Ruiter
,
D. J.
,
Fernández
,
G.
, &
Henson
,
R. N.
(
2012
).
How schema and novelty augment memory formation
.
Trends in Neurosciences
,
35
,
211
219
.
Vannini
,
P.
,
Hedden
,
T.
,
Sullivan
,
C.
, &
Sperling
,
R. A.
(
2013
).
Differential functional response in the posteromedial cortices and hippocampus to stimulus repetition during successful memory encoding
.
Human Brain Mapping
,
34
,
1568
1578
.
Wagner
,
A. D.
,
Schacter
,
D. L.
,
Rotte
,
M.
,
Koutstaal
,
W.
,
Maril
,
A.
,
Dale
,
A. M.
, et al
(
1998
).
Building memories: Remembering and forgetting of verbal experiences as predicted by brain activity
.
Science
,
281
,
1188
1191
.
Wisniewski
,
E. J.
(
1997
).
When concepts combine
.
Psychonomic Bulletin & Review
,
4
,
167
183
.
Zeithamova
,
D.
,
Dominick
,
A. L.
, &
Preston
,
A. R.
(
2012
).
Hippocampal and ventral medial prefrontal activation during retrieval-mediated learning supports novel inference
.
Neuron
,
75
,
168
179
.
Zhao
,
J.
,
Al-Aidroos
,
N.
, &
Turk-Browne
,
N. B.
(
2013
).
Attention is spontaneously biased toward regularities
.
Psychological Science
,
24
,
667
677
.