Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Nov 1.
Published in final edited form as: Psychophysiology. 2016 Jul 27;53(11):1669–1678. doi: 10.1111/psyp.12730

FN400 and LPC memory effects for concrete and abstract words

Paweł Stróżak a, Christopher W Bird b, Krystin Corby b, Gwen Frishkoff c, Tim Curran b
PMCID: PMC5061608  NIHMSID: NIHMS800039  PMID: 27463978

Abstract

According to dual-process models, recognition memory depends on two neurocognitive mechanisms: familiarity, which has been linked to the "frontal N400" (FN400) effect in studies using event-related potentials (ERPs), and recollection, which is reflected by changes in the late positive complex (LPC). Recently, there has been some debate over the relationship between FN400 familiarity effects and N400 semantic effects. According to one view, these effects are one and the same. Proponents of this view have suggested that the frontal distribution of the FN400 could be due to stimulus concreteness: recognition memory experiments commonly use highly imageable or concrete words (or pictures), which elicit semantic ERPs with a frontal distribution. In the present study we tested this claim using a recognition memory paradigm in which subjects memorized concrete and abstract nouns; half of the words changed font color between study and test. FN400 and LPC old/new effects were observed for abstract, as well as concrete words, and were stronger over right hemisphere electrodes for concrete words. However, there was no difference in anteriority of the FN400 effect for the two word types. These findings challenge the notion that the frontal distribution of the FN400 old/new effect is fully explained by stimulus concreteness.

Keywords: Memory, Language/Speech, ERPs

Dual-process models assume the existence of two distinct processes in recognition memory — familiarity and recollection (Atkinson & Juola, 1974; Mandler, 1980; Yonelinas, 1999, 2002). Familiarity is fast-acting, automatic, and associated with a generic "feeling" that an item has previously been encountered. Recollection is slower and is accompanied by a more detailed memory for past events, such as when or where the item was previously encountered, or whether the item had a particular characteristic.

Studies in cognitive neuroscience have provided converging evidence for the dual process model (Aggleton et al., 2005; Rugg & Curran, 2007; Yonelinas, Otten, Shaw, & Rugg, 2005). In particular, event-related potentials (ERPs) have revealed neurophysiological indices of familiarity and recollection — the mid-frontal FN400 and the parietal late positive complex (LPC) — which appear to be functionally, as well as electrophysiologically, distinct (Rugg & Curran, 2007; Friedman & Johnson, 2000; Tsivilis, Otten, & Rugg, 2001). In a one study, Curran (2000) examined recognition memory for words that were inflected as singular (e.g., "table") or plural (e.g., "spoons"). At test, old words (e.g., "table", "spoons"), new words (e.g., "forks"), and plurality-reversed lures (e.g., "tables", "spoon") were presented. Subjects were instructed to classify as "studied" only the same-plurality old words. During the test phase, the FN400 component had a more negative amplitude for words that were correctly rejected (e. g., "spoon" or “forks”), as compared with those that were correctly recognized (e.g., “table”) as well as false alarms (e.g., “tables”). Because there was no difference in the FN400 response to correctly recognized items and false alarms, Curran (2000) concluded that the FN400 effect was due to familiarity-based memory, as opposed to recollection. By contrast, the LPC was recorded from 400 to 800 ms over parietal electrodes and was more positive in response to words that were correctly recognized, as compared with correct rejections and false alarms. Based on this pattern of results, the LPC was proposed as a specific index of recollection-based memory. In the subsequent studies, ERPs indices of familiarity and recollection were tested with regard to other characteristics of the stimuli used, such as picture orientation (Curran & Cleary, 2003; Nyhus & Curran, 2012) or color (Cycowicz, Friedman, & Snodgrass, 2001; Ecker, Zimmer, & Groh-Bordin, 2007a, 2007b; Groh-Bordin, Zimmer, & Ecker, 2006).

Paller, Voss and Boehm (2007) challenged the claim that the FN400 component is a specific index of familiarity-based memory. Instead, they suggested that it could reflect conceptual-semantic processing. They argued that the repetition of stimuli between study and test phases in recognition memory experiments enhances the fluency of conceptual processing. This enhanced fluency is associated with electrophysiological responses that co-occur and contribute to the feeling of familiarity. Thus, the relative increase in FN400 negativity to correct rejections (i.e., correct responses to "new" stimuli) versus items perceived as "old" (i.e., hits and false alarms) might reflect the increased effort needed to retrieve their associated meanings, rather than the absence of familiarity-based memory. Conceptual-semantic processing of meaningful stimuli is indexed by the standard N400 potential, which was firstly described as having more pronounced, centroparietal negativity for semantic anomalies than for semantically congruent control sentences (Kutas & Hillyard, 1980). The N400 is attenuated when conceptual processing is facilitated, as in the case of repetition or conceptual priming (Kutas & Federmeier, 2011). If the account of Paller et al. (2007) is correct, then the so-called frontal N400, or "FN400", effect could be regarded as functionally identical to the centroparietal N400 repetition effect. This proposal is reasonable, given that both ERPs occur within ~300–500 ms after stimulus onset and that familiarity-based memory may be closely related to conceptual-semantic access (Paller et al., 2007). Electrophysiologically, one apparent difference is that the FN400 is typically maximal over mid-frontal electrodes, whereas the classical semantic N400 is typically more posterior (Kutas & Hillyard, 1980). However, the spatial distribution of the N400 effect varies considerably across studies (Kutas & Federmeier, 2011). Therefore, it is possible that the frontal distribution of the FN400 reflects modulation of the semantic N400 by specific stimulus or task-related variables.

One well-known source of variation in N400 topography is stimulus concreteness: N400 priming effects are typically maximal over mid-frontal, as opposed to parietal, sites for concrete words (Kounios & Holcomb, 1994; Adorni & Proverbio, 2012). Given this finding, Voss and Federmeier (2011) and others have argued that the frontal distribution of the FN400 might reflect the concreteness of word stimuli, rather than a process — such as familiarity-based memory — that can be dissociated from conceptual-semantic access. They suggested that concrete stimuli might facilitate semantic processing relative to abstract stimuli, thus shifting the N400 repetition effect for concrete stimuli towards frontal areas of the scalp. More generally, the variability in N400 topography is consistent with the view that the N400 is not a unitary component, but reflects the activation of multiple neural generators (Frishkoff, Tucker, Davey, & Scherg, 2004; Frishkoff, 2007; Kutas & Federmeier, 2011).

In the present study, we considered word concreteness and familiarity-based memory effects within the same paradigm. Previous studies of episodic memory effects have typically used words that are medium-to-high in concreteness. However, there are no reports of FN400 familiarity effects when comparing concrete versus abstract words. We suggest that word concreteness is not sufficient to account for the frontal distribution of the FN400 and that this effect might be also observed for abstract words. To test this idea, we used a modified version of Curran’s (2000) experiments. Words appeared in different colored fonts in both the study and test phases. At test, participants were instructed to respond "old" only for items presented in the same color during the study and test phases. Importantly, words included abstract as well as concrete nouns. We hypothesized that frontal old/new effects would be elicited for both types of words from ~300 to 500 ms. If the old/new effect for abstract words was found to be significant over frontal electrodes, this would challenge the claim that the anteriority of the FN400 is reducible to stimulus concreteness. We also compared the topographic distributions of old/new effects by concreteness level, using range normalization to rescale the difference waves prior to statistical analysis (Wilding, 2006). Any topographic differences between familiarity effects elicited by concrete and abstract words would indicate that stimulus concreteness can exert some influence on the FN400 topography. Finally, we expected to observe an LPC effect: that is, greater positivity for old responses versus new responses. We had no predictions regarding the influence of concreteness on the topography of the LPC.

Method

Participants

Thirty-four right-handed, native speakers of English (17 females, 17 males) participated in the experiment. The mean age of participants was M=23.0 years (SD=3.14; Range=18–29 years). All subjects had normal or corrected-to-normal vision, with no known neurological problems. Data from two subjects were excluded from the final analysis because there were too few good EEG trials (less than 36 in any category) after rejection of trials with electrophysiological artifacts. Therefore, final results are reported for 32 subjects (16 females, 16 males). All procedures were approved by University of Colorado Institutional Review Board. Subjects were paid $15 per hour or received course credit for their participation.

Materials and Procedure

Word Stimuli

Stimuli included 864 words collected from the MRC Psycholinguistic Database (Wilson, 1988). Half of them were concrete nouns (concreteness ratings: M=570.69, SD=34.15, Range=501–670; imageability ratings: M=569.89, SD=42.30, Range=432–667; e.g., horse, bowl, chair). The other half were abstract nouns (concreteness ratings: M=305.40, SD=28.30, Range=217–349; imageability ratings: M=375.92, SD=68.17, Range=101–578; e.g., trouble, justice, origin). Concreteness was significantly different for concrete and abstract nouns, [t(862) = 124.31, p < 0.001], and imageability also differed between the two word types [t(862) = 50.25, p < 0.001]. There was no difference in written word frequency (Kučera & Francis, 1967) for concrete words (M=69.98, SD=103.31, Range=2–967) versus abstract words (M=81.79, SD=129.12, Range=1–683, [t(862) = 1.48, p = 0.14]. The two word types were balanced in length (M=5.32, SD=1.22, Range=3–7): i.e., there were similar distributions of shorter (3-letter), mid-length (4–5 letter), and longer (6–7 letter) words.

For the experiment, words were divided into 12 blocks of 72 words each, counterbalanced by concreteness and word length. Each block consisted of three subsets of 24 nouns (12 concrete, 12 abstract). The first subset was presented during study and test (“old words”). The second subset was presented only during test (“new words”). Finally, the third subset was presented in one font color during study and in a different font color at test (“similar words”). Half of the words within each subset (6 concrete, 6 abstract) were presented in blue (0, 114, 255 RGB values); the other half were presented in orange (255, 140, 0 RGB values). Half of the “similar” words changed from blue (in the study phase) to orange (in the test phase); the other half changed from orange to blue. The color manipulation was intended to approximate Curran’s (2000)plurality manipulation without the restriction of using pluralizable words. The subsets were counterbalanced across subjects, such that each subset served equally as often as "old", "new", and "similar" within each block. Words were randomly ordered within each block, subject to the above constraints. An additional 48 words served as buffers: two were presented at the beginning and two were presented at the end of each study phase within each block. Another 36 nouns (half abstract, half concrete) were used as practice stimuli.

All stimuli were presented at the center of a 17-inch LCD computer monitor with a display resolution of 1280 × 1024 pixels and refresh rate of 60 Hz. Words were presented in Arial font, with a point size of 30.

Task Procedure

Subjects were seated in a dimly lit, electrically shielded and acoustically isolated chamber, at a viewing distance of 50 cm from the computer monitor. Participants completed a short practice session, followed by 12 experimental (study–test) blocks.

In each study phase 48 words (24 "old" and 24 "similar") and 4 buffer words were presented (2 in the beginning and 2 at the end of each list). At the beginning of each study trial, a fixation cross appeared for a random duration (between 1400 and 1600 ms) and was immediately followed by a word, which was presented for 1000 ms. Subjects were instructed to study each word and to memorize its font color. No overt response was required during the study phase.

An old/new recognition test, or test phase, followed each study phase. Each of 72 test trials began with a fixation cross, which appeared for a random duration (between 1400 and 1600 ms) and was immediately replaced by a word, which was presented for 1000 ms. The word was immediately followed by a response probe (a question mark, "?"). Subjects were asked to indicate whether the word was "old" (studied) or "new." They were told to respond as quickly and accurately as possible, by pressing one of two buttons on the Serial Response Box using their (right or left) index finger. They were also informed that they should answer "old" only in response to words that appeared in the same font color during the study and test phases; otherwise, they should respond "new." Assignment of left and right index fingers to “old” and “new” responses was counterbalanced across subjects. There were two subject-paced blink breaks during each test phase, one every 24 trials.

Subjects took self-paced breaks between the study and test phases and between blocks. Electrode impedances were checked before the start of the experiment and after every three blocks, lasting at least 5 minutes each time.

EEG data collection and preprocessing

An electroencephalogram (EEG) was recorded continuously from 128 scalp locations using a HydroCel Geodesic Sensor Net (Tucker, 1993) connected to a DC-coupled, 128-channel, high-input-impedance amplifier (200 MΩ, NetAmps 200, Electrical Geodesics, Inc., Eugene, OR). The EEG was referenced to the vertex electrode, band-pass filtered (0.1 to 100 Hz) and digitized at a sampling rate of 250 Hz. Electrode impedances were kept below 50 kΩ, an acceptable range for this system.

NetStation (NetStation 4, Electrical Geodesics, Inc., Eugene, OR) and EEGLAB version 13 (Delorme & Makeig, 2004) software was used for EEG signal processing and analysis. Continuous EEG was filtered offline, using a digital low-pass filter (40 Hz). The EEG was segmented into 1200 ms epochs, beginning 200 ms prior to stimulus onset. Eye-movements and eye-blinks were corrected using Independent Component Analysis (Delorme, Sejnowski, & Makeig, 2007). Bad channels were manually marked as bad before ICA and interpolated after that using the spherical option of the EEGLAB function for channel interpolation (Bigdely-Shamlo et al., 2015). Remaining artifacts were rejected using a moving window peak-to-peak amplitude method with window width 200 ms, window step 100 ms, and threshold ±100 µV (Luck, 2014). Artifact-free trials were averaged and re-referenced to the average of all 128 channels. The resulting ERPs were baseline-corrected relative to the 200-ms pre-stimulus interval.

The number of trials per subject was comparable across the four experimental conditions: (1) "old" responses to concrete words, collapsing across hits and false alarms (M=115.09, SD=39.73, Range=38–201); (2) "old" responses to abstract words, collapsing across hits and false alarms (M=111.28, SD=40.26, Range=36–185), (3) correct rejections of new concrete words (M=106.63, SD=23.94, Range=41–139), and (4) correct rejections of new abstract words (M=106.44, SD=22.99, Range=44–138).

Data analysis

Behavioral Data Analysis

We compared the mean proportion of correct rejections, hits (correct recognition of old words), and false alarms (incorrect endorsement of color-changed words as "old"). Data were entered into a two-way repeated-measures analysis of variance (ANOVA) with Condition (correct rejection/hit/false alarm) and Concreteness (concrete/abstract) as repeated measures. Greenhouse-Geisser corrected p-values are reported where appropriate.

EEG Data Analysis

Our strategy for ERP analysis was similar to that of previous study in which FN400 and N400 components were recorded (Stróżak, Abedzadeh, & Curran, 2016). We analyzed ERP data from twelve electrode clusters: LF (left frontal), MF (midline frontal), RF (right frontal), LC (left central), MC (midline central), RC (right central), LP (left parietal), MP (midline parietal), RP (right parietal), LO (left occipital), MO (midline occipital), and RO (right occipital). Each cluster consisted of six electrodes, including one International 10–20 channel (F3, Fz, F4, C3, Cz, C4, P3, Pz, P4, O1, Oz, or O2), and five adjacent electrodes (see Figure 1). Mean amplitude values were computed across the seven electrodes within each region for both the FN400 (300–500 ms) and LPC (500–800 ms) time windows.

Figure 1.

Figure 1

Open in a new tab

Sensor locations on the 128-channel HydroCel Geodesic Sensor Net. Twelve electrode clusters included in ANOVAs are denoted: LF (left frontal), MF (midline frontal), RF (right frontal), LC (left central), MC (midline central), RC (right central), LP (left parietal), MP (midline parietal), RP (right parietal), LO (left occipital), MO (midline occipital), and RO (right occipital)

ERP analyses examined responses during the test phase of the experiment, focusing on old/new effects: that is, differences in response to studied words perceived as old (collapsing across hits to old words and false alarms to color-changed words) and words correctly perceived as new (i.e., correct rejections). We disregarded the accuracy of color judgments because there was little evidence that color-correct and color-incorrect old responses in our study indexed distinct neurocognitive processes.1

Separate analyses were conducted for mid-latency (300–500 ms) and late (500–800 ms) time windows, corresponding to the FN400 and LPC effects. In each case, we conducted repeated-measures analyses of variance (ANOVA) with four factors: Condition (old/new), Concreteness (concrete/abstract), Region (frontal/central/parietal/occipital), and Laterality (left/midline/right). A Greenhouse-Geisser correction was applied when appropriate, and Bonferroni corrections were used to correct for multiple comparisons. Effects due to Region and Laterality are noted only if they interact with Condition and/or Concreteness.

Topographic analyses were conducted on rescaled data, using the range normalization procedure described in Wilding (2006). The vector-length procedure (McCarthy & Wood, 1985) — though widely used — has been criticized for failing to address condition differences due to baseline effects and noise (Urbach & Kutas, 2002). According to Wilding (2006), the range normalization procedure is not susceptible to these criticisms.

Results presented below were obtained from average-referenced ERPs. The same analyses were conducted on mastoid-referenced ERPs (re-referencing to the average of the left and right mastoids) to allow for comparisons with results from other studies in which a mastoid reference was used (e.g., Voss and Federmeier, 2011; Bridger et al., 2012). The results of these analyses were consistent with the initial average-referenced analysis, so only average-reference results are reported.2 ERP waveform plots and topographic maps based on mastoid-referenced data are presented in Supplementary Figures.

Results

Behavioral results

Mean proportions and standard errors of the means for recognition judgments are summarized in Table 1. Analysis of mean accuracy revealed a main effect of Condition [F(2,62) = 61.76, p < .001, η2 = .67]. Bonferroni-adjusted post-hoc comparisons showed that participants were more accurate for correct rejections (M = .85, SEM = .03) versus hits (M = .55, SEM = .04; p < .001), and correct recognitions were more prevalent than false alarms (M = .34, SEM = .03; p < .001). A Condition × Concreteness interaction was also observed [F(2,62) = 17.36, p < .001, η2 = .36]. Bonferroni corrected post-hoc tests showed that subjects were more accurate in recognizing old concrete words (M = .58, SEM = .04) versus old abstract words (M = .53, SEM = .03; p < .001). Subjects were equally good at rejecting new concrete words (M = .85, SEM = .03) and new abstract words (M = .84, SEM = .03; p = .59). Also, the mean proportion of false alarms (to color-changed words) was larger for abstract words (M = .35, SEM = .03), as compared with concrete words (M = .32, SEM = .03; p < .05).

Table 1.

Mean proportions and standard errors of the means for recognition judgments

Correct rejection of
new words		 Correct recognition of
old words		 False recognition of
similar words
M SEM M SEM M SEM
Concrete .85 .03 .58 .04 .32 .03
Abstract .84 .03 .53 .03 .35 .03
Open in a new tab

ERP results

FN400 old/new effects (300–500 ms)

Figures 2 and 3 show grand-averaged ERPs elicited during the test phase for each condition. Results for concrete nouns are depicted in Figure 2; results for abstract nouns are shown in Figure 3. In both figures, there appear to be differences in ERP responses to "old" versus "new" words in the FN400 (300–500 ms) and LPC (500–800 ms) windows.

Figure 2.

Figure 2

Open in a new tab

Grand-averaged ERPs for concrete words elicited by old responses (collapsed across hits and false alarms) and new responses (correct rejections)

Figure 3.

Figure 3

Open in a new tab

Grand-averaged ERPs for abstract words elicited by old responses (collapsed across hits and false alarms) and new responses (correct rejections)

Within the 300–500 ms time window, statistical analyses showed a main effect of Condition [F(1,31) = 26.57, p < .001, η2 = .46]. ERPs were more negative for correct rejections (M = −.14 µV, SEM = .09) than for "old" responses (M = .001 µV, SEM = .09). The Condition effect was clarified by Condition × Region [F(1.63,50.66) = 10.59, p < .001, η2 = .26] and Condition × Region × Laterality [F(3.42,106.14) = 6.17, p < .001, η2 = .17] interactions (see Figures 4A and 4B). Bonferroni-corrected post-hoc comparisons for the Condition × Region interaction showed that the old/new effect was most pronounced in frontal [p < .001, η2 = .41] and central [p < .001, η2 = .51] regions, and was also significant in the parietal region [p < .05, η2 = .18]. The old/new effect was reversed in the occipital region, such that ERPs were more negative for "old" responses than for correct rejections [p < .05, η2 = .16], Bonferroni-corrected post-hoc comparisons for the Condition × Region × Laterality interaction showed that the old/new effect was most prominent in the midline central (MC) cluster [p < .001, η2 = .57].

Figure 4.

Figure 4

Open in a new tab

Topographic maps in the 300–500 ms time window for FN400 effect for concrete words (A), FN400 effect for abstract words (B), and in the 500–800 ms time window for LPC effect for concrete words (C), and LPC effect for abstract words (D)

There was also a main effect of Concreteness [F(1,31) = 4.84, p < .05, η2 = .14]. ERPs were more negative for concrete nouns (M = −.10 µV, SEM = .10) than for abstract nouns (M = −.04 µV, SEM = .09). The Concreteness effect was clarified by Concreteness × Region [F(1.5,46.6) = 29.7, p < .001, η2 = .49], Concreteness × Laterality [F(2,62) = 5.19, p < .01, η2 = .14], and Concreteness × Region × Laterality [F(4.2,130.14) = 10.34, p < .001, η2 = .25] interactions. Bonferroni-corrected post-hoc comparisons for the Concreteness × Region interaction showed that ERPs to concrete words were more negative than ERPs to abstract words in frontal [p < .001, η2 = .50] and central [p < .001, η2 = .53] regions. The concreteness effect was reversed in occipital regions, such that ERPs were more negative for abstract versus concrete nouns [p < .001, η2 = .52]. Bonferroni-corrected post-hoc comparisons for the Concreteness × Laterality interaction showed that ERPs to concrete words were more negative than ERPs to abstract words only in the right hemisphere [p < .01, η2 = .23; p = .62 in the left hemisphere; p = .065 in the midline]. Bonferroni-corrected post-hoc comparisons for the Concreteness × Region × Laterality interaction showed that the concreteness effect was most prominent in the right frontal (RF) cluster [p < .001, η2 = .58], in the midline central (MC) cluster [p < .001, η2 = .61], and in the left occipital (LO) cluster [p < .001, η2 = .53].

Importantly, there was no two-way Condition × Concreteness interaction [F(1,31) = .004, p = .95, η2 = .00] and no three-way Condition × Concreteness × Region interaction [F(1.47,45.49) = 1.49, p = .24, η2 = .05]. Additional analyses were conducted, focusing on FN400 old/new effects for concrete and abstract words within the frontal region. Separate ANOVAs were conducted for concrete and abstract words, including ERPs in frontal regions only (i.e., data from left, midline, and right frontal clusters). Each analysis included two factors: Condition (old/new) and Laterality (left/midline/right). For concrete words, there was a main effect of Condition [F(1,31) = 18.22, p < .001, η2 = .37]: ERPs were more negative for correct rejections (M = −.92 µV, SEM = .23) than for "old" responses (M = −.56 µV, SEM = .26). A main effect of Condition was also observed for abstract words [F(1,31) = 9.19, p < .01, η2 = .23]: again, ERPs were more negative for correct rejections (M = −.52 µV, SEM = .25) than for "old" responses (M = −.25 µV, SEM = .23).

Topographic comparisons of FN400 effects for concrete and abstract words

Additional analyses wehre conducted, using range-normalized difference waves (Wilding, 2006), to evaluate topographic differences in old/new effects for concrete versus abstract nouns. The old/new difference wave for concrete nouns was created by subtracting correct rejections to concrete words from "old" responses to concrete words (collapsing across hits and false alarms). The old/new difference wave for abstract nouns was created following the same procedure. Topographic maps corresponding to the difference waves are depicted in Figures 4A and 4B. These figures show FN400 effects (i.e., increased positivity for "old" versus "new" responses from 300 to 500 ms) for abstract and concrete words.

Range-normalized data were entered into a repeated-measures analysis of variance (ANOVA), with three factors: Contrast (old/new effect for concrete nouns versus old/new effect for abstract nouns), Region (frontal/central/parietal/occipital), and Laterality (left/midline/right). Results showed a Contrast × Laterality interaction [F(2,62) = 3.30, p < .05, η2 = .10]. This interaction was deconstructed by conducting Bonferroni-corrected post-hoc tests that compared old/new effects for concrete versus abstract words at each level of laterality (left hemisphere, midline and right hemisphere clusters). Post-hoc comparisons showed that the difference in the distribution of the old/new effects for concrete and abstract words was found only in the right hemisphere [p < .05]: within this region, the old/new effect was larger for concrete than for abstract words. Importantly, there was no Contrast × Region interaction [F(1.43,44.47) = .26, p = .70, η2 = .008] and no Contrast × Region × Laterality interaction [F(4.21,130.59) = 1.3, p = .27, η2 = .04].

LPC old/new effects (500–800 ms)

Within the 500–800 ms time window, there was a main effect of Condition [F(1,31) = 19.19, p < .001, η2 = .38]. Consistent with previous studies, the LPC was more positive for "old" responses (M = .41 µV, SEM = .10) than for correct rejections (M = .20 µV, SEM = .08). The Condition effect was clarified by a marginally significant Condition × Region interaction [F(1.39,42.98) = 3.42, p = .058, η2 = .10] and by a significant Condition × Region × Laterality interaction [F(3.68,114.20) = 4.68, p < .01, η2 = .13] (see Figures 4C and 4D). Bonferroni-corrected post-hoc comparisons for the Condition × Region interaction showed that "old" responses elicited more positive ERPs than correct rejections within the central [p < .01, η2 = .29] and parietal [p < .01, η2 = .29] regions (p = .057 in frontal regions; p = .37 in occipital regions). Bonferroni-corrected post-hoc comparisons for the Condition × Region × Laterality interaction showed that the old/new effect was most prominent within the midline central (MC) cluster [p < .001, η2 = .44] and within the left parietal (LP) cluster [p < .001, η2 = .37].

There was also a main effect of Concreteness [F(1,31) = 6.07, p < .05, η2 = .16]. ERPs were more positive for concrete nouns (M = .35 µV, SEM = .09) than for abstract nouns (M = .26 µV, SEM = .09). The Concreteness effect was clarified by a Concreteness × Region interaction [F(1.68,52.16) = 17.88, p < .001, η2 = .37], and by a Concreteness × Region × Laterality interaction [F(4.02,124.56) = 2.78, p < .05, η2 = .08]. Bonferroni-corrected post-hoc comparisons for the Concreteness × Region interaction showed that ERPs to concrete words were more negative than ERPs to abstract words in frontal regions [p < .001, η2 = .37]. This effect was reversed over posterior sites: concrete words elicited more positive ERPs than abstract words over parietal sites [p < .01, η2 = .29] and also over occipital sites [p < .001, η2 = .40]. The concreteness effect was non-significant in central regions [p = .17]. Bonferroni-corrected post-hoc comparisons for the Concreteness × Region × Laterality interaction showed the most prominent concreteness effects within the right frontal (RF) cluster [p < .001, η2 = .38], the midline parietal (MP) cluster [p < .01, η2 = .31], and the right occipital (RO) cluster [p < .001, η2 = .43].

Finally, a significant Condition × Concreteness × Region [F(1.44,44.52) = 4.21, p < .05, η2 = .12] interaction was observed. For concrete words, Bonferroni-corrected post-hoc comparisons revealed that "old" responses elicited more positive ERPs than correct rejections in frontal [p < .05, η2 = .16], central [p < .01, η2 = .24] and parietal [p < .05, η2 = .19] regions. For abstract words, "old" responses elicited more positive ERPs than correct rejections within central [p < .01, η2 = .21] and parietal [p < .01, η2 = .28] regions, but the old/new effect for abstract words failed to reach significance within frontal regions [p = .34]. For both word types, there was no difference in old versus new responses within occipital regions [p = .07 for concrete words; p = .75 for abstract words].

Topographic comparisons of LPC effects for concrete and abstract words

To further probe the topographic differences in LPC effects for concrete versus abstract words, we performed additional, unplanned comparisons using range-normalized difference waves (following Wilding, 2006), similar to the FN400 topographic comparisons. ERP difference waves for concrete and abstract words were computed within the LPC (500–800 ms) time window. The topographic maps corresponding to these difference waves are depicted in Figures 4C and 4D.

A repeated-measures analysis of variance (ANOVA) was conducted, with three factors: Contrast (old/new effect for concrete nouns versus old/new effect for abstract words), Region (frontal/central/parietal/occipital), and Laterality (left/midline/right). Analysis of range-normalized data showed a significant Contrast × Laterality interaction [F(2,62) = 3.48, p < .05, η2 = .10]. This interaction was deconstructed by conducting Bonferroni-corrected post-hoc tests that compared old/new effects for concrete versus abstract words at each level of laterality (left hemisphere, midline, right hemisphere). Post-hoc comparisons showed that the difference in the distribution of the old/new effects for concrete and abstract words was significant only in the right hemisphere [p < .05], with larger old/new effects for concrete versus abstract words within this region. There was no Contrast × Region interaction [F(1.33,41.27) = 1.99, p = .16, η2 = .06] and no Contrast × Region × Laterality interaction [F(4.15,128.56) = 0.9, p = .47, η2 = .03].

Discussion

In the present study, our main goal was to compare the regional (i.e., frontal versus non-frontal) distribution of FN400 old/new effects for concrete and abstract words. Below we discuss our findings for the FN400, as well as LPC effects and right hemisphere differences for concrete versus abstract words. We consider our findings with respect to prior work on episodic memory for meaningful stimuli (e.g., words, pictures). We also discuss how familiarity and conceptual-semantic processes may contribute to episodic memory for words. Finally, we discuss some limitations and suggestions for future studies.

The present study shows FN400 responses that are modulated by memory for recently presented words. ERP responses were more negative in response to new words that were correctly rejected than for words that were correctly or incorrectly endorsed as old within the FN400 (300–500 ms) time window, replicating the original findings of Curran (2000). Importantly, the FN400 memory effect was observed for abstract, as well as concrete, words, and there were no differences in regional distribution of the effect for the two word types. In addition, topographic analyses using range-normalized data (Wilding, 2006) failed to show any difference in regional distribution of the FN400 effect for concrete and abstract words. These findings are inconsistent with the claim that the anteriority of the FN400 is due to stimulus concreteness.

Although the FN400 old/new effects did not differ for concrete and abstract words, concrete words did elicit a more right-lateralized old/new memory effect (compare Figures 4A and 4B). This finding is significant for two reasons. First, it suggests that the topographic analysis had enough statistical power to detect interactions. Thus, the failure to observe an interaction between Contrast (concrete versus abstract old/new effect) and Region (frontal, central, parietal, occipital) is unlikely due to the lack of statistical power. Second, the larger old/new effect for concrete versus abstract words within the right-lateral regions is consistent with previous ERP studies, which have shown larger responses for concrete versus abstract words over the right hemisphere (e.g., Kounios & Holcomb, 1994). Although it is risky to assume that right-lateral ERPs are due to underlying generators in the right cortex, researchers have noted that the right-lateralization of concreteness effects are consistent with dual-coding theory (Holcomb, Kounios, Anderson, & West, 1999; West & Holcomb, 2000). According to this theory (Paivio, 1991), abstract words are represented using left-hemisphere verbal codes, whereas concrete words also activate right-hemisphere mechanisms that support imagery. On the other hand, Figures 4A and 4B show that concrete words elicit a larger overall effect, as compared with abstract words. This finding is compatible with the context availability model, which suggests that concrete words activate more contextual information than abstract words and should therefore engage a larger set of brain areas (Fiebach & Friederici, 2003). In either case, the difference in old/new effects for concrete versus abstract words can inform our understanding of the FN400 potential.

With regard to the late positive complex (LPC), there is wide agreement that the LPC old/new effect is associated with recollection-based memory (Paller et al., 2007; Rugg & Curran, 2007). As expected, the present results indicated that LPC amplitudes were larger for old responses than for new responses. We also found that the LPC effect, like the FN400 effect, was more right-lateralized for concrete than for abstract words and, in fact, was larger overall for concrete words. Thus, it seems that retrieval of detailed episodic information associated with concrete words activates more brain regions in the right hemisphere than the retrieval of such information associated with abstract words. Again, this finding may be consistent with dual-coding theory (Paivio, 1991) or context availability (Fiebach & Friederici, 2003), or both.

There is active debate about the functional significance of the FN400 effect. In particular, some researchers (e.g., Voss & Federmeier, 2012) have suggested that the FN400 old/new effect reflects the same set of cognitive processes as the N400 semantic effect. Indeed, given that both effects are elicited by meaningful stimuli — including pictures (Ganis, Kutas, & Sereno, 1996; Olivares, Iglesias, & Bobes, 1999) and words — it may be difficult to separate these two effects when subjects are engaged in semantic processing. In support of this idea, Stróżak, Abedzadeh, and Curran (2016) found that FN400 and N400 effects were topographically indistinguishable when semantic priming was embedded within a recognition test. However, when the semantic (study) and memory (test) phases were separated, the FN400 memory effect was found to have a more frontal distribution than the N400 semantic effect. This finding suggests that it may be possible to distinguish between FN400 familiarity and N400 semantic effects, even within the same experiment and in response to the same stimuli (also see Bridger, et al., 2012).

Although the present results do not directly address differences between the view the FN400 effects reflect familiarity (Groh-Bordin, Zimmer, & Ecker, 2006; Rugg & Curran, 2007) vs. conceptual priming (Voss & Paller, 2007; Voss, Lucas, & Paller, 2010), they do suggest some ways to think about the inter-relationship of these processes. Our finding that the FN400 old/new effect for concrete words was more right-lateralized than for abstract words suggests that the electrophysiological underpinnings of familiarity judgments are not unitary, but vary according to the type of information being evaluated (here: concrete or abstract words). If so, familiarity might be explained in terms of fluency (or ease of processing) to prior stimulus exposure (Jacoby & Whitehouse, 1989), which would be more right-lateralized for concrete than for abstract words because concrete concepts are represented more bilaterally than the abstract concepts. From this perspective, FN400 old/new effects could reflect familiarity or conceptual-semantic processing or both. This is also consistent with a finding of occipital familiarity-related ERPs for words that were orthographically, but not semantically similar to studied words (Lucas & Paller, 2013), which is compatible with the notion of representationally-specific mechanisms of fluency that contributes to familiarity. Thus, it is possible that in our study, as well as in other recognition memory experiments, the FN400 and N400 components are intermixed and produce broad topography on the scalp.

It is important to acknowledge that conclusions based on the present findings may be limited in some respects. First, the task that we used in our procedure (remembering words and their font colors) is likely to have focused subjects' attention on nonsemantic (perceptual) features of the stimuli. It is therefore possible that our results would have been different if we had used a different task. Previous ERP studies have reported larger FN400 old/new effects following a semantic, as opposed to a perceptual, encoding task (Nyhus & Curran, 2009) and larger following deep than shallow encoding conditions (Rugg, Allan, & Birch, 2000). In line with this finding, participants in the present study had relatively low accuracy. It is also worth noting that participants made simple binary responses ("old" versus "new"). It is therefore possible that the old/new contrast was contaminated by guesses based on font color, especially in the case of "false alarms" to color-altered words. In future studies, it may be possible to address this issue by using more fine-grained measures of episodic memory, such as confidence ratings, in addition to the old/new task.

Second, it is important to exercise caution when drawing inferences from null effects, such as the absence of interactions between old/new or concreteness effects and Region, especially for the topographic comparisons. The range-normalization procedure that we used is more conservative than the alternative vector-length procedure and is therefore subject to Type II errors (Haig, Gordon, & Hook, 1997). However, we did observe a significant interaction between Contrast (old/new effect for concrete words vs. old/new effect for abstract words) and Laterality for this same analysis. This finding makes it less unlikely that the null finding is simply due to a lack of statistical power (also see Voss & Federmeier, 2011).

Finally, it should be acknowledged that amplitude rescaling (by whatever means) does not guarantee that topographic differences are driven by functional differences associated with experimental condition, rather than by nuisance factors, such as baseline differences or noise (Urbach & Kutas, 2006). Therefore, although we performed topographic analyses on amplitude-normalized data, using range-normalization procedure, the present findings should be replicated, to minimize the possibility that topographic differences in the present study are due to chance factors. A few observations suggest that conditions were well matched at baseline within the present study and did not differ in terms of noise. For example, we did not observe any differences in the rate of rejection between concrete and abstract trials from the EEG analyses (see EEG data collection and preprocessing section for details). Thus, ERPs for concrete and abstract words were derived from roughly the same number of trials (although this does not guarantee that the signal-to-noise ratios are equivalent). Moreover, the directionality of the observed effects — including larger differences for concrete versus abstract words in right-lateral areas — are consistent with previous findings. Nonetheless, future studies should aim to replicate these findings and to explore other factors, such as task and ease of stimulus encoding and recall, to clarify the cognitive bases of these effects.

In conclusion, we observed FN400 and LPC old/new effects for abstract, as well as concrete, words. Both effects were stronger over right hemisphere electrodes in response to concrete versus abstract words, consistent with dual coding theory (Paivio, 1991). However, it is also important to note that concrete words elicited larger overall effects, which could be explained by activation of additional contextual information (including more semantic features), as hypothesized by the context availability model (Fiebach & Friederici, 2003). Importantly, there was no difference in anteriority of the FN400 effect for concrete and abstract words. This finding suggests that the frontal distribution of the FN400 old/new effect does not merely reflect an anterior shift of the semantic N400 due to stimulus concreteness (Voss & Federmeier, 2011).

Supplementary Material

Supp Fig S1. Figure S1.

Mastoid-referenced grand-averaged ERPs for concrete words elicited by old responses (collapsed across hits and false alarms) and new responses (correct rejections)

Supp Fig S2. Figure S2.

Mastoid-referenced grand-averaged ERPs for abstract words elicited by old responses (collapsed across hits and false alarms) and new responses (correct rejections)

Supp Fig S3. Figure S3.

Mastoid-referenced topographic maps in the 300–500 ms time window for FN400 effect for concrete words (A), FN400 effect for abstract words (B), and in the 500–800 ms time window for LPC effect for concrete words (C), and LPC effect for abstract words (D)

Acknowledgments

Funding for this research was provided by NIH Grants MH64812 and MH096698, and NSF Grant #SMA-1041755 to the Temporal Dynamics of Learning Center, an NSF Science of Learning Center. We are grateful to Ila Goldanloo, Kale Hubert, and Katie Cording for research assistance.

Footnotes

We disclose no actual or potential conflict of interest.

1

This was due to the fact that trials endorsed as false alarms were trials on which the word itself was correctly recognized, but for which only one specific detail (font color) was made incorrectly. That allows a large amount of non-criterial recollection (not font-color related; e.g., contextual details) to remain in false alarms trials (we thank two anonymous reviewers for converging suggestions on this topic). Moreover, 300–500 ms ERPs for correct recognition and false alarms to similar lures were virtually identical, as is typically observed (e.g., Curran, 2000).

2

There were two reasons for not reporting mastoid-referenced data as the main results. First, for range-normalized topographic contrasts, there was greater error variance for mastoid-referenced data than for average-referenced data, as previously demonstrated by Dien (1998). This reduced the test statistics and resulted in non-significant Bonferroni corrected post-hoc p values for mastoid-referenced topographic interactions. Second, the average reference technique is preferable when analyzing the N400 potential because it produces topography that is more readily interpretable in terms of known physiology and anatomy in comparison to alternative references (Johnson & Hamm, 2000).

References

  1. Adorni R, Proverbio AM. The neural manifestation of the word concreteness effect: An electrical neuroimaging study. Neuropsychologia. 2012;50:880–891. doi: 10.1016/j.neuropsychologia.2012.01.028. [DOI] [PubMed] [Google Scholar]
  2. Aggleton JP, Vann SD, Denby C, Dix S, Mayes AR, Roberts N, Yonelinas AP. Sparing the familiarity component of recognition memory in a patient with hippocampal pathology. Neuropsychologia. 2005;43:1810–1823. doi: 10.1016/j.neuropsychologia.2005.01.019. [DOI] [PubMed] [Google Scholar]
  3. Atkinson RC, Juola JF. Search and decision processes in recognition memory. In: Krantz DH, Atkinson RC, Luce RD, Suppes P, editors. Contemporary developments in mathematical psychology: Vol. 1. Learning, memory & thinking. San Francisco: Freeman; 1974. [Google Scholar]
  4. Bridger EK, Bader R, Kriukova O, Unger K, Mecklinger A. The FN400 is functionally distinct from the N400. NeuroImage. 2012;63:1334–1342. doi: 10.1016/j.neuroimage.2012.07.047. [DOI] [PubMed] [Google Scholar]
  5. Bigdely-Shamlo N, Mullen T, Kothe C, Su KM, Robbins KA. The PREP pipeline: standardized preprocessing for large-scale EEG analysis. Frontiers in Neuroinformatics. 2015;9:16. doi: 10.3389/fninf.2015.00016. http://dx.doi.org/10.3389/fninf.2015.00016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Curran T. Brain potentials of recollection and familiarity. Memory & Cognition. 2000;28:923–938. doi: 10.3758/bf03209340. [DOI] [PubMed] [Google Scholar]
  7. Curran T, Cleary AM. Using ERPs to dissociate recollection from familiarity in picture recognition. Cognitive Brain Research. 2003;15:191–205. doi: 10.1016/s0926-6410(02)00192-1. [DOI] [PubMed] [Google Scholar]
  8. Cycowicz YM, Friedman D, Snodgrass JG. Remembering the color of objects: An ERP investigation of source memory. Cerebral Cortex. 2001;11:322–334. doi: 10.1093/cercor/11.4.322. [DOI] [PubMed] [Google Scholar]
  9. Delorme A, Makeig S. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. Journal of Neuroscience Methods. 2004;134:9–21. doi: 10.1016/j.jneumeth.2003.10.009. [DOI] [PubMed] [Google Scholar]
  10. Delorme A, Sejnowski T, Makeig S. Enhanced detection of artifacts in EEG data using higher-order statistics and independent component analysis. Neuroimage. 2007;34:1443–1449. doi: 10.1016/j.neuroimage.2006.11.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Dien J. Issues in the application of the average reference: Review, critiques, and recommendations. Behavior Research Methods, Instruments, & Computers. 1998;30:34–43. [Google Scholar]
  12. Ecker UK, Zimmer HD, Groh-Bordin C. Color and context: An ERP study on intrinsic and extrinsic feature binding in episodic memory. Memory & Cognition. 2007a;35:1483–1501. doi: 10.3758/bf03193618. [DOI] [PubMed] [Google Scholar]
  13. Ecker UK, Zimmer HD, Groh-Bordin C. The influence of object and background color manipulations on the electrophysiological indices of recognition memory. Brain Research. 2007b;1185:221–230. doi: 10.1016/j.brainres.2007.09.047. [DOI] [PubMed] [Google Scholar]
  14. Fiebach CJ, Friederici AD. Processing concrete words: fMRI evidence against a specific right-hemisphere involvement. Neuropsychologia. 2004;42(1):62–70. doi: 10.1016/s0028-3932(03)00145-3. [DOI] [PubMed] [Google Scholar]
  15. Friedman D, Johnson R. Event-related potential (ERP) studies of memory encoding and retrieval: A selective review. Microscopy Research and Technique. 2000;51:6–28. doi: 10.1002/1097-0029(20001001)51:1<6::AID-JEMT2>3.0.CO;2-R. [DOI] [PubMed] [Google Scholar]
  16. Frishkoff GA. Hemispheric differences in strong versus weak semantic priming: Evidence from event-related brain potentials. Brain and Language. 2007;100:23–43. doi: 10.1016/j.bandl.2006.06.117. [DOI] [PubMed] [Google Scholar]
  17. Frishkoff GA, Tucker DM, Davey C, Scherg M. Frontal and posterior sources of event-related potentials in semantic comprehension. Cognitive Brain Research. 2004;20:329–354. doi: 10.1016/j.cogbrainres.2004.02.009. [DOI] [PubMed] [Google Scholar]
  18. Ganis G, Kutas M, Sereno MI. The search for "common sense": An electrophysiological study of the comprehension of words and pictures in reading. Journal of Cognitive Neuroscience. 1996;8:89–106. doi: 10.1162/jocn.1996.8.2.89. 10.1162/jocn.1996.8.2.89. [DOI] [PubMed] [Google Scholar]
  19. Groh-Bordin C, Zimmer HD, Ecker UK. Has the butcher on the bus dyed his hair? When color changes modulate ERP correlates of familiarity and recollection. NeuroImage. 2006;32:1879–1890. doi: 10.1016/j.neuroimage.2006.04.215. [DOI] [PubMed] [Google Scholar]
  20. Haig AR, Gordon E, Hook S. To scale or not to scale: McCarthy and Wood revisited. Electroencephalography & Clinical Neurophysiology. 1997;103:323–325. doi: 10.1016/s0013-4694(97)00009-6. [DOI] [PubMed] [Google Scholar]
  21. Holcomb PJ, Kounios J, Anderson JE, West WC. Dual-coding, context-availability, and concreteness effects in sentence comprehension: An electrophysiological investigation. Journal of Experimental Psychology: Learning, Memory, and Cognition. 1999;25:721–742. doi: 10.1037//0278-7393.25.3.721. [DOI] [PubMed] [Google Scholar]
  22. Jacoby LL, Whitehouse K. An illusion of memory: False recognition influenced by unconscious perception. Journal of Experimental Psychology: General. 1989;118:126–135. [Google Scholar]
  23. Johnson BW, Hamm JP. High-density mapping in an N400 paradigm: evidence for bilateral temporal lobe generators. Clinical Neurophysiology. 2000;111:532–545. doi: 10.1016/s1388-2457(99)00270-9. [DOI] [PubMed] [Google Scholar]
  24. Kounios J, Holcomb PJ. Concreteness effects in semantic processing: ERP evidence supporting dual-coding theory. Journal of Experimental Psychology: Learning, Memory, and Cognition. 1994;20:804–823. doi: 10.1037//0278-7393.20.4.804. [DOI] [PubMed] [Google Scholar]
  25. Kučera H, Francis WN. Computational analysis of present-day American English. Providence, RI: Brown University Press; 1967. [Google Scholar]
  26. Kutas M, Federmeier KD. Thirty years and counting: Finding meaning in the N400 component of the event-related brain potential (ERP) Annual Review of Psychology. 2011;62:621–647. doi: 10.1146/annurev.psych.093008.131123. 10.1146/annurev.psych.093008.131123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kutas M, Hillyard SA. Reading senseless sentences: Brain potentials reflect semantic incongruity. Science. 1980;207:203–205. doi: 10.1126/science.7350657. http://dx.doi.org/10.1126/science.7350657. [DOI] [PubMed] [Google Scholar]
  28. Lucas HD, Paller KA. Manipulating letter fluency for words alters electrophysiological correlates of recognition memory. NeuroImage. 2013;83:849–861. doi: 10.1016/j.neuroimage.2013.07.039. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Luck SJ. An introduction to the event-related potential technique. MIT press; 2014. [Google Scholar]
  30. Mandler G. Recognizing: The judgment of previous occurrence. Psychological Review. 1980;87:252–271. http://dx.doi.org/10.1037/0033-295X.87.3.252. [Google Scholar]
  31. McCarthy G, Wood C. Scalp distributions of event-related potentials: An ambiguity associated with analysis of variance models. Electroencephalography & Clinical Neurophysiology. 1985;62:203–208. doi: 10.1016/0168-5597(85)90015-2. [DOI] [PubMed] [Google Scholar]
  32. Nyhus E, Curran T. Semantic and perceptual effects on recognition memory: Evidence from ERP. Brain Research. 2009;1283:102–114. doi: 10.1016/j.brainres.2009.05.091. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Nyhus E, Curran T. Midazolam-induced amnesia reduces memory for details and affects the ERP correlates of recollection and familiarity. Journal of Cognitive Neuroscience. 2012;24:416–427. doi: 10.1162/jocn_a_00154. [DOI] [PubMed] [Google Scholar]
  34. Olivares EI, Iglesias J, Bobes MA. Searching for face-specific long latency ERPs: A topographic study of effects associated with mismatching features. Cognitive Brain Research. 1999;7:343–356. doi: 10.1016/s0926-6410(98)00038-x. [DOI] [PubMed] [Google Scholar]
  35. Paivio A. Dual coding theory: Retrospect and current status. Canadian Journal of Psychology. 1991;45:255–287. http://dx.doi.org/10.1037/h0084295. [Google Scholar]
  36. Paller KA, Voss JL, Boehm SG. Validating neural correlates of familiarity. Trends in Cognitive Sciences. 2007;11:243–250. doi: 10.1016/j.tics.2007.04.002. [DOI] [PubMed] [Google Scholar]
  37. Rugg MD, Allan K, Birch CS. Electrophysiological evidence for the modulation of retrieval orientation by depth of study processing. Journal of Cognitive Neuroscience. 2000;12:664–678. doi: 10.1162/089892900562291. 10.1162/089892900562291. [DOI] [PubMed] [Google Scholar]
  38. Rugg MD, Curran T. Event-related potentials and recognition memory. Trends in Cognitive Sciences. 2007;11:251–257. doi: 10.1016/j.tics.2007.04.004. [DOI] [PubMed] [Google Scholar]
  39. Stróżak P, Abedzadeh D, Curran T. Separating FN400 and N400 potentials across recognition memory experiments. Brain Research. 2016;1635:41–60. doi: 10.1016/j.brainres.2016.01.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Tsivilis D, Otten LJ, Rugg MD. Context effects on the neural correlates of recognition memory: An electrophysiological study. Neuron. 2001;31:497–505. doi: 10.1016/s0896-6273(01)00376-2. [DOI] [PubMed] [Google Scholar]
  41. Tucker DM. Spatial sampling of head electrical fields: The geodesic sensor net. Electroencephalography & Clinical Neurophysiology. 1993;87:154–163. doi: 10.1016/0013-4694(93)90121-b. [DOI] [PubMed] [Google Scholar]
  42. Urbach TP, Kutas M. The intractability of scaling scalp distributions to infer neuroelectric sources. Psychophysiology. 2002;39:791–808. doi: 10.1111/1469-8986.3960791. http://dx.doi.org/10.1111/1469-8986.3960791. [DOI] [PubMed] [Google Scholar]
  43. Urbach TP, Kutas M. Interpreting event-related brain potential (ERP) distributions: implications of baseline potentials and variability with application to amplitude normalization by vector scaling. Biological Psychology. 2006;72:333–343. doi: 10.1016/j.biopsycho.2005.11.012. [DOI] [PubMed] [Google Scholar]
  44. Voss JL, Federmeier KD. FN400 potentials are functionally identical to N400 potentials and reflect semantic processing during recognition testing. Psychophysiology. 2011;48:532–546. doi: 10.1111/j.1469-8986.2010.01085.x. http://dx.doi.org/10.1111/j.1469-8986.2010.01085.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Voss JL, Lucas HD, Paller KA. Conceptual priming and familiarity: different expressions of memory during recognition testing with distinct neurophysiological correlates. Journal of Cognitive Neuroscience. 2010;22:2638–2651. doi: 10.1162/jocn.2009.21341. [DOI] [PubMed] [Google Scholar]
  46. Voss JL, Paller KA. Neural correlates of conceptual implicit memory and their contamination of putative neural correlates of explicit memory. Learning & Memory. 2007;14(4):259–267. doi: 10.1101/lm.529807. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. West WC, Holcomb PJ. Imaginal, semantic, and surface-level processing of concrete and abstract words: An electrophysiological investigation. Journal of Cognitive Neuroscience. 2000;12:1024–1037. doi: 10.1162/08989290051137558. http://dx.doi.org/10.1162/08989290051137558. [DOI] [PubMed] [Google Scholar]
  48. Wilding EL. The practice of rescaling scalp-recorded event-related potentials. Biological Psychology. 2006;72:325–332. doi: 10.1016/j.biopsycho.2005.12.002. [DOI] [PubMed] [Google Scholar]
  49. Wilson MD. The MRC psycholinguistic database: Machine-usable Dictionary, version 2.00. Behavioural Research Methods, Instruments and Computers. 1988;20:6–10. http://dx.doi.org/10.3758/BF03202594. [Google Scholar]
  50. Yonelinas AP. The contribution of recollection and familiarity to recognition and source-memory judgments: A formal dual-process model and an analysis of receiver operating characteristics. Journal of Experimental Psychology: Learning, Memory, and Cognition. 1999;25:1415–1434. doi: 10.1037//0278-7393.25.6.1415. http://dx.doi.org/10.1037/0278-7393.25.6.1415. [DOI] [PubMed] [Google Scholar]
  51. Yonelinas AP. The nature of recollection and familiarity: A review of 30 years of research. Journal of Memory and Language. 2002;46:441–517. [Google Scholar]
  52. Yonelinas AP, Otten LJ, Shaw KN, Rugg MD. Separating the brain regions involved in recollection and familiarity in recognition memory. The Journal of Neuroscience. 2005;25:3002–3008. doi: 10.1523/JNEUROSCI.5295-04.2005. http://dx.doi.org/10.1523/JNEUROSCI.5295-04.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp Fig S1. Figure S1.

Mastoid-referenced grand-averaged ERPs for concrete words elicited by old responses (collapsed across hits and false alarms) and new responses (correct rejections)

NIHMS800039-supplement-Supp_Fig_S1.pdf (273KB, pdf)
Supp Fig S2. Figure S2.

Mastoid-referenced grand-averaged ERPs for abstract words elicited by old responses (collapsed across hits and false alarms) and new responses (correct rejections)

NIHMS800039-supplement-Supp_Fig_S2.pdf (271KB, pdf)
Supp Fig S3. Figure S3.

Mastoid-referenced topographic maps in the 300–500 ms time window for FN400 effect for concrete words (A), FN400 effect for abstract words (B), and in the 500–800 ms time window for LPC effect for concrete words (C), and LPC effect for abstract words (D)

NIHMS800039-supplement-Supp_Fig_S3.pdf (267.6KB, pdf)

RESOURCES