Memory‐guided perception is shaped by dynamic two‐stage theta‐ and alpha‐mediated retrieval

Abstract

How does memory influence auditory perception, and what are the underlying mechanisms that drive these interactions? Most empirical studies on the neural correlates of memory‐guided perception have used static visual tasks, resulting in a bias in the literature that contrasts with recent research highlighting the dynamic nature of memory retrieval. Here, we used electroencephalography to track the retrieval of auditory associative memories in a cue–target paradigm. Participants (N = 64) listened to real‐world soundscapes that were either predictive of an upcoming target tone or nonpredictive. Three key results emerged. First, targets were detected faster when embedded in predictive than in nonpredictive soundscapes (memory‐guided perceptual benefit). Second, changes in theta and alpha power differentiated soundscape contexts that were predictive from nonpredictive contexts at two distinct temporal intervals from soundscape onset (early—950 ms peak for theta and alpha, and late—1650 ms peak for alpha only). Third, early theta activity in the left anterior temporal lobe was correlated with memory‐guided perceptual benefits. Together, these findings underscore the role of distinct neural processes at different time points during associative retrieval. By emphasizing temporal sensitivity and by isolating cue‐related activity, we reveal a two‐stage retrieval mechanism that advances our understanding of how memory influences auditory perception.

Full text

DOI: 10.1111/nyas.15287
ORIGINAL ARTICLE
Memory-guided perception is shaped by dynamic two-stage
theta- and alpha-mediated retrieval
Manda Fischer1,2,3Morris Moscovitch2,3Claude Alain2,3
1The Brain and Mind Institute, University of
Western Ontario, London, Ontario, Canada
2Department of Psychology, University of
Toronto,Toronto, Ontario, Canada
3Rotman Research Institute, Baycrest Centre
for Geriatric Care, Toronto,Ontario, Canada
Correspondence
Manda Fischer, The Brain and Mind Institute,
University of Western Ontario, Western
Interdisciplinary Research Building, Perth
Drive, Rm. 4130, London, ON N6A 3K7,
Canada. Email: mfisch5@uwo.ca
Funding information
Natural Sciences and Engineering Research
Council of Canada, Grant/AwardNumbers:
Alexander Graham Bell Grant, Discovery
Grant/A8347,
DiscoveryGrant/RGPIN-2021-02721
Abstract
How does memory influence auditory perception, and what are the underlying mech-
anisms that drive these interactions? Most empirical studies on the neural correlates
of memory-guided perception have used static visual tasks, resulting in a bias in the
literature that contrasts with recent research highlighting the dynamic nature of mem-
ory retrieval. Here, we used electroencephalography to track the retrieval of auditory
associative memories in a cue–target paradigm. Participants (N=64) listened to
real-world soundscapes that were either predictive of an upcoming target tone or
nonpredictive. Three key results emerged. First, targets were detected faster when
embedded in predictive than in nonpredictive soundscapes (memory-guided percep-
tual benefit). Second, changes in theta and alpha power differentiated soundscape
contexts that were predictive from nonpredictive contexts at two distinct temporal
intervals from soundscape onset (early—950 ms peak for theta and alpha, and late—
1650 ms peak for alpha only). Third, early theta activity in the left anterior temporal
lobe was correlated with memory-guided perceptual benefits. Together, these findings
underscore the role of distinct neural processes at different time points during associa-
tive retrieval. By emphasizing temporal sensitivity and by isolating cue-related activity,
we reveal a two-stage retrieval mechanism that advances our understanding of how
memory influences auditory perception.
KEYWORDS
anterior temporal, associative memory, auditory, fronto-parietal, memory-guided, oscillations,
retrieval
INTRODUCTION
There is growing interest in understanding how memory and per-
ception interact.1–10 Most investigations, however, have focused on
paradigms that use static visual stimuli.11,12 This bias is at odds with
recent research emphasizing the dynamic nature of memory13 and
behavioral evidence demonstrating that memory can influence per-
ception in nonvisual domains. The current study addressed this gap
by using temporally sensitive measures to understand how memory
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any
medium, provided the original work is properly cited and is not used for commercial purposes.
© 2025 The Author(s). Annals of the New York Academy of Sciences published by Wiley PeriodicalsLLC on behalf of The New York Academy of Sciences.
retrieval dynamically operates to shape subsequent behavior on a
cue–target detection task.
Memory-guided perception likely unfolds in cascading yet tempo-
rally distinct stages, including memory retrieval and target processing.
Most studies, however, confound these stages by presenting the cue
and target simultaneously or by using blocked designs that do not
make it possible to analyze cue-related activity separately from target
processing.2–5,14–16 Consequently, these designs prioritize perceptual
processing while often masking important memory-related effects
AnnNYAcadSci. 2025;1544:159–171. wileyonlinelibrary.com/journal/nyas 159

160 ANNALS OF THE NEW YORK ACADEMY OF SCIENCES
occurring during the cue period, specifically impacting our understand-
ing of how different brain regions contribute to memory retrieval
versus target processing.
While fronto-parietal activity, which aids in focusing atten-
tion on anticipated targets, is reliably observed across studies,11
medial temporal lobe (MTL) activity—hypothesized to support
associative retrieval17,18—shows substantial variability, including
studies that report this activity19–23 and others that show no MTL
involvement.4,5,24–26 This inconsistency and lack of clarity in the field
may stem from the limited focus on cued retrieval specifically.
If memory guides perception, yet only effects related to perceptual
processing are robust and consistently observed, what is the nature
and timing of the source of this bias? Cue-related activity has been
directly linked to subsequent memory outcomes,27,28 and this process-
ing window likely plays a functional role in memory-guided perception.
We argue, therefore, that isolating the cued-retrieval processing stage
is essential in order to elucidate a comprehensive and time-sensitive
understanding of the neural processing that supports memory-guided
perception.
Current study
The present experiment aimed to elucidate the nature and timing of
auditory memory-guided perception by characterizing the relationship
between memory retrieval and subsequent real-time perceptual bene-
fits. We used a cue–target detection task to isolate cue-related activity
from target and response processing. Participants heard real-world
soundscapes (cues) that were either predictive of an upcoming tar-
get tone or nonpredictive. Importantly, in predictive trials, the target
tone was expected within a specific time window relative to cue onset,
but no physical tone was played during the cue period. This design
allowed us to dissociate cue-related activity from target-related activ-
ity, enabling us to focus on how predictions about an upcoming target
are formed rather than on the perception of the target itself. Instead
of comparing old and new trials, we contrasted predictive versus non-
predictive conditions. This approach helped control for nonassociative
memory effects, allowing us to better isolate the neural mechanisms
involved in associative memory retrieval.
The primary distinction between previous studies and the current
one is our examination of cue-related activity in isolation. While earlier
research has focused on perceptual processing21 and response-related
processing,29–32 the current study centers on cue-related processing
to better understand how memory retrieval contributes to percep-
tual benefits in reaction time (RT). Additionally, a key distinction of
our study is the focus on source-localized effects rather than tra-
ditional event-related potentials. Source-localized analyses require a
good signal-to-noise ratio (SNR); therefore, we pooled data across
studies to achieve this.
We hypothesized that response time to detect the target tone would
be faster for predictive than nonpredictive trials, in line with previ-
ous work that has demonstrated that auditory memory can enhance
perception.29,33 To understand how these perceptual benefits man-
ifest, we focused our analyses on changes in theta (4–7 Hz) and
alpha (8–12 Hz) power and source-localized these differences using a
multiple-source beamformer to complement and extend neurocogni-
tive models derived from the functional magnetic resonance imaging
(fMRI) literature in the field. Specifically, we predicted that changes
in theta and alpha power would distinguish between predictive and
nonpredictive conditions, indexing associative retrieval.34,35 Further,
because activity during memory retrieval has been shown to pre-
dict memory performance,27,28 we expected that cue-related changes
would predict real-time behavioral performance on our detection
task.
To preview our results, we found two distinct temporal inter-
vals (950 and 1650 ms peak relative to cue onset) during memory
retrieval. Early changes in theta power were source-localized to the
anterior temporal lobe (ATL) and scaled with the memory-guided RT
benefit afforded by predictive trials. Early and late changes in source-
localized fronto-parietal alpha power distinguished contexts that were
predictive from nonpredictive but did not scale with the memory-
guided RT benefit. These findings highlight the dynamic nature of
memory retrieval and the importance of cue-related activity during
memory-guided perception.
METHODS
Participants
Sixty-nine participants took part in one of two nearly identical
memory-guided electroencephalography (EEG) studies conducted by
our group.29 We pooled data across these studies to increase our sta-
tistical power for the present source analyses. One participant did not
have EEG data, and four participants were excludeddue to poor-quality
EEG recordings, resulting in a total of 64 participants (Mage =22.8
years; SDage =4.0; 41 female and 23 male). A sample size of 64 pro-
vides 80% power to detect a one-tailed paired effect of d≥0.3.37 The
memory-guided benefit found in previous studies has a medium effect
size (most conservative effect d=0.58 in Zimmermann et al.33), and
memory-guided effects of this size should be detectable with a power
of about 100% in the current design. Participants were between 18
and 35 years of age, had normal hearing as assessed using pure tone
audiometry, and had no history of psychiatric, neurological, or other
major illnesses. Participants provided informed written consent, and
the study was certified for ethical compliance by the Research Ethics
Board at Baycrest.
Stimuli
A total of 124 different real-world ambient soundscapes (e.g.,
Cityscape sound) (2500 ms) were selected from http://www.freesound.
org/. All soundscapes were down-sampled to a sampling rate of
44,100 Hz and were edited to have a 100 ms rise-and-fall time. Stim-
uli were presented binaurally using Presentation software (version

ANNALS OF THE NEW YORK ACADEMY OF SCIENCES 161
FIGURE 1 Overview of the paradigm, including the learning phase discrimination task and the test phase cue–target detection task. During
the learning phase, participants indicated whether the suprathreshold tone was high, low, or absent. During the test phase, participants detected a
faint lateralized target tone as quickly as possible by pressing the right or left arrow keys. Predictive trials contained a faint pure tone target within
the same time window and location as it appeared during learning, while nonpredictive and new trials contained a faint pure tone target presented
on a pseudo-randomized side. Note: Images are for visualization only. A blank screen was presented throughout the experiment and, occasionally,
written instructions were displayed between trials and blocks (see Methods for details). Abbreviations: ITI, intertrial interval; SPL, sound pressure
level.
13, Neurobehavioral Systems) via insert earphones (EARTONE 3a) at
an average listening volume of 65 decibels (dB) sound pressure level
(SPL) and with maximum peak amplitudes of 80 dB SPL. Sound levels
were measured using a Larson Davis System 824 and a 2CC coupler.
The pure tone (500 Hz or 1 kHz, 500 ms in duration, 50 ms rise/fall
time) was presented monaurally (left or right ear) at 80 dBA SPL dur-
ing the four learning phases and at 55 dBA SPL during the final test
phase.
Procedure
The study consisted of a learning phase of four learning blocks (80 real-
world ambient soundscapes repeated four times, once per block) and a
surprise test phase (104 real-world ambient soundscapes played once)
(Figure 1). During the learning phase, participants were exposed to
unique soundscape–tone pairings while performing an irrelevant dis-
crimination task. During the test phase, participants were tested on
these pairings, in a cue–target detection task. The order of the sound-
scapes was randomized within each block. Importantly, we controlled
for differences in the memorability of soundscape–tone pairings by
counterbalancing stimuli across conditions and participants. This type
of design ensured that the acoustics were matched for every condition
at the group level, ruling out the possibility that acoustics could drive
the observed effects.
Learning phase: Repeat exposure to
soundscape–tone pairings
During each of the four learning blocks, half of the 80 soundscapes
were paired with a pure tone (predictive condition), while the other half
did not contain a pure tone (nonpredictive condition). On predictive tri-
als, the pure tone was consistently presented in either the left or right
ear and embedded within a fixed time window (Exp 1: 2000 ms, Exp 2:
randomly varied between 1300, 1800, or 2300 ms relative to sound-
scape onset), such that an association could be formed between the
soundscape and the timing/location of the tone. Notably, the pure tone
level was high (80 dBA SPL, well above the hearing threshold), and we
made no explicit reference to the repeat timing and location of the pure
tone. These experimental manipulations aimed to promote incidental
learning of the association between the soundscape and the tone.38
Each trial began by presenting a soundscape that was either paired
with a tone (predictive) or not (nonpredictive). After the soundscape
finished playing, participants had 500 ms to indicate, by pressing the
left, right, or down arrow key, whether the pitch of the embedded
pure tone was high, low, or not present. We pseudo-randomized tone
frequency within and across learning blocks so that there was no con-
sistent pairing between a given soundscape and the frequency of the
tone. Immediately after making their response, participants were given
500 ms of visual feedback on their performance (green =correct; red =
incorrect; blue =too slow).

162 ANNALS OF THE NEW YORK ACADEMY OF SCIENCES
Test phase: Cue-target detection with
soundscape–tone pairings
After the learning phase was complete, participants took a short break
(30 min) before they underwent a surprise memory test. They were
presented with one block of 104 trials that included 80 previously
presented soundscapes (40 predictive and 40 nonpredictive) from the
learning phase and 24 new ones.
Each trial started by presenting a soundscape (cue), followed by its
replay. During the replay, a faint pure tone target (55 dBA SPL; near
hearing threshold) was embedded within the same time window used
during learning (1300–2300 ms relative to soundscape onset). During
the test phase, every trial included a pure tone target within the sound-
scape. For nonpredictive and new soundscapes, in which no tone was
embedded during learning, the location of the pure tone target was
pseudo-randomized: half of the trials had a tone on the left, while the
other half had a tone on the right. Participants indicated the location of
the pure tone target as quickly as possible by pressing the left or right
arrow keys.
On each trial after target detection, participants answered ques-
tions to probe their memory for the soundscape and soundscape–
tone association: (Q1) If the soundscape was old or new; (Q2)
When old, if, at learning, the tone paired with this soundscape
was on the left or right, or not present. The assignment of left
and right key presses corresponding to old and new responses
(Q1) was counterbalanced across participants to minimize response
bias.
Behavioral data analysis
We focused our analyses on the cue–target detection task. Contrasts
that compare activity for old and new stimuli, though commonly used
in the literature, inherently confound associative memory processes
with other cognitive functions that support successful task completion.
Therefore, we focused our analyses on the contrast between predictive
and nonpredictive conditions to control for nonassociative contribu-
tions to memory retrieval, allowing us to isolate the neural substrates
underlying associative retrieval that drive prediction-based benefits in
target detection. In other words, our contrast between predictive and
nonpredictive soundscape contexts compared retrievalevents in which
soundscape context is retrieved with or without retrieving the asso-
ciative information about the timing/location of the tone (target tone
prediction).
In line with previous work,1,33,39,40 we expected to see a memory-
guided benefit, in that target detection would be faster in the predictive
condition than in the nonpredictive condition. To assess the effect of
associative memory on target detection, we used a one-tail paired
t-test to compare target detection RTs between predictive and non-
predictive conditions. To assess whether the memory-cue benefit
(predictive vs. nonpredictive) was accompanied byrecognition memory
for the soundscape (Procedure—Q1) and target tone (Procedure—Q2),
we used independent t-tests.
We conducted Pearson’s correlations to assess brain–behavior
relationships, based on prior research demonstrating a relation
between hippocampal activity and memory-guided RT benefits14,21
and between posterior parietal activity and memory-guided RT
benefits.3We did not apply a correction for multiple comparisons
because these analyses were planned and motivated by established
findings. In contrast, any results related to brain regions outside
the hippocampus and posterior parietal cortex were considered
exploratory and were subject to Bonferroni correction to adjust for
multiple comparisons.
EEG recording and processing
EEG was recorded from 64 scalp electrodes using a BioSemi Active
Two acquisition system (BioSemi V.O.F.). The electrode montage was
based on the 10/20 system and included a Common Mode Sense active
electrode and a Driven Right Leg passive electrode serving as ground.
Six additional electrodes were placed below the hairline (mastoids,
pre-auricular points, and two facial electrodes) and extended to the
covered area. Four electrodes were placed at each eye’s outer canthus
and inferior orbit to monitor eye movements. All signals were bandpass
filtered between 0.01 and 100 Hz, sampled at a rate of 512 Hz, digi-
tized, and stored continuously. The EEG data were analyzed using the
Brain Electrical Source Analysis (BESA) software (BESA Research 7.1;
MEGIS).
All trials, regardless of behavioral accuracy, were included in the
EEG analysis. The EEG data were visually inspected to identify seg-
ments contaminated by defective electrodes. Noisy electrodes were
interpolated using data from the surrounding electrodes. The EEG was
then re-referenced to the average of all electrodes. For each partic-
ipant, ocular movements were identified from the continuous EEG
recording and then used to generate components that best accounted
for eye movements. The scalp projections of these components were
then subtracted from the EEG signal to minimize ocular contamination,
such as blinks and lateral and vertical eye movements. The continu-
ous EEG was parsed into epochs according to the onset of the first
presentation of the soundscape (cue). This cued–retrieval epoch com-
prised a 400 ms pre-stimulus interval (−600 to −200 ms) followed
by a 2500 ms cue interval. For trial rejection, the EEG was digi-
tally filtered with a 0.1 Hz high-pass filter (zero phase, 12 dB/octave)
and a 40 Hz low-pass filter (zero phase, 24 dB/octave). Epochs with
EEG signals exceeding ±120µV were marked and excluded from fur-
ther analysis. The processed data comprised at least 85% of the
epochs per experimental condition and participant. Each average
was baseline-corrected with respect to the pre-stimulus baseline
interval.
Time-frequency analysis
We used a complex demodulation method with 1 Hz-wide frequency
bins and 50 ms time resolution in the 2 and 50 Hz range for

ANNALS OF THE NEW YORK ACADEMY OF SCIENCES 163
decomposing the single-trial EEG data into a time-frequency represen-
tation. We focused the time-frequency analyses on theta (4–7 Hz) and
alpha (8–12 Hz) bands, as these two frequency bands have been asso-
ciated with memory retrieval34 and internal attention to the recovered
memory.41
Multiple-source beamformer in the time-frequency
domain
A multiple-source beamformer42 was applied to the time windows
identified from the time-frequency analysis. This technique spatially
filters scalp-recorded EEG data to estimate the source power of spe-
cific location(s) in the brain. Source power distributions in the 3D brain
were estimated by iteratively constructing the beamformer at each
voxel (7 mm3).43 For each channel, single-trial data were transformed
from the time domain to the time-frequency domain. This transforma-
tion yielded a cross-spectral density matrix comprising channel index,
frequency, and time information. Next, the output power of the beam-
former at each voxel was computed. We used the four-shell ellipsoidal
model provided by BESA Research 7.1, with a head radius of 85 mm
and thickness for scalp, bone, and cerebrospinal fluid of 6, 7, and 1 mm,
respectively. The relative conductivities were 0.33,0.33, 0.0042, and 1
S/m for brain, scalp, bone, and cerebrospinal fluid, respectively. A 1 Hz
high-pass filter was applied before analyzing source-localized theta
changes to remove excessdrift c omponents thatc ouldotherwise mask
the signal of interest.44,45
EEG data analysis
All analyses were performed on the EEG data recorded during the sur-
prise test phase to assess retrieval-related effects. We time-locked our
analyses to the onset of the soundscape (cue) (Figure 1) between 0
and 2000 ms relative to cue onset. The strength of this analysis design
is that it captures a prediction window during the cue period that is
unconfounded with later target and response processing when the
soundscape replays. The cue period during the test phase was iden-
tical in both studies, making it possible to examine memory-guided
preparatory neural activity.
Sensors and sources of oscillatory activity elicited by predictive
and nonpredictive trials from the test session were subjected to
nonparametric cluster-based permutation testing using BESA Statis-
tics software (Statistics 2.1, MEGIS GmbH). This analysis aimed to
identify neural correlates of auditory associative retrieval follow-
ing cue onset. A preliminary step identified clusters both in time
(adjacent time points) and space (adjacent electrodes), where the
time-frequency (electrode level) or beamformer (voxel level) represen-
tations differed between the conditions. A Monte-Carlo resampling
technique46 then identified clusters with higher t-values than 95%
of all clusters derived by random data permutation. This last step
ensured that this statistic was not subject to the multiple comparison
problem (for an in-depth overview of permutation statistics as imple-
mented in BESA Statistics, see Ref. 46). An alpha level of 0.01 was
used for cluster building, and the number of permutations was set to
5000.
RESULTS
Behavioral results
To assess participants’ recognition memory for the soundscape, we
calculated sensitivity (d’) using hits and false alarms derived from the
old–new judgments (Procedure Q1). Recognition memory for both pre-
dictive and nonpredictive soundscapes was above chance (d’=1.1, 95%
CI [0.94, 1.23], t(63) =15.1, p<0.0001, d=1.9) with lower recognition
memory for predictive soundscapes (mean d’=0.97, 95% CI [0.8, 1.1])
than nonpredictive soundscapes (mean d’=1.2, 95% CI [1.05, 1.36]),
t(63) =−6.0, p<0.0001, d=−0.75).
To test whether memory-guided perception depended on recogni-
tion memory for the presence of the tone in relation to the soundscape,
we ran a one-sample t-test comparing sensitivity (d’) for tone presence
or absence at learning to chance. Data did not meet the assumption
for normality (W=0.86, p<0.0001), so we conducted a Wilcoxon
signed-rank test. Participants performed above chance (W(63) =1495,
p<0.0001, rb=0.48), indicating that participants could discriminate
between trials that contained a tone at learning (predictive) and those
that did not (nonpredictive) (mean d’=0.27, 95% CI [0.05, 0.49]). Mem-
ory for the exact location of the tone, however, was no different from
chance (33% left, right, not present) (W(63) =914.5, p=0.40, rb=
−0.12; M=32.8, 95% CI [28.5, 37.2]).
RTs to detect the faint pure tone at test were computed on tri-
als in which participants correctly detected the pure tone target (M=
92.6%, 95% CI [91.5, 93.6]) and on trials in which they had success-
fully recalled the soundscape (M=54.5%, 95% CI [51.3, 57.7]). Trials
with RTs shorter than 100 ms or longer than the mean plus twice the
standard deviation were excluded from further analyses. Comparing
predictive and nonpredictive trials enabled us to test whether merely
pairing a real-world soundscape and a lateralized pure tone during
exposure afforded a benefit in RTs for detecting the pure tone target
at test. RT data were not normally distributed, so the RT data were
log10-transformed before analysis. Detection of the faint pure tone
target was faster for predictive (M=594 ms, 95% CI [561.0, 628.0])
than for nonpredictive trials (M=611 ms, 95% CI [575.1, 647.8]) (t(63)
=−2.3, p=0.01, d=−0.29), suggesting that participants had formed
context-to-target associations that aided them during detection
(Figure 2).
EEG results
Early and late changes in theta and alpha power
support auditory memory-guided perception
We compared oscillatory activity in the theta and alpha frequency
bands during the cue period at test. To have sufficient trials in
each condition for accurate source-localization, we included all trials

164 ANNALS OF THE NEW YORK ACADEMY OF SCIENCES
FIGURE 2 Group mean raw reaction time scores for predictive
and nonpredictive trials. Error bars represent the 95% within-subject
confidence interval. All analyses were conducted on trials in which
participants successfully (1) indicated the correct side of the target
tone (M=94.0%, 95% CI [93.0, 95.1]) and (2) categorized the
soundscape as old (all predictive and nonpredictive trials were
old—previously displayed during learning) (M=35.8%, 95% CI [31.2,
40.5]). *p<0.05.
regardless of participants’ recognition memory accuracy for the sound-
scape or tone pairing. Table 1presents a summary of the cluster-based
permutation statistics.
Early effects
An early cluster revealed that theta power was reduced in the pre-
dictive condition compared to the nonpredictive one (p=0.0036,
corrected) between 950 and 1200 ms over the right central and
central-parietal scalp area (Figure 3, top plot). Around this time, alpha
power was reduced for predictive trials compared to nonpredictive
ones (p=0.014, corrected) between 900 and 1050 ms over the right
frontal and frontocentral scalp area (Figure 3, middle plot).
Late effects
Three later clusters (1500–1900 ms) revealed that alpha power was
greater for predictive than nonpredictive trials over the left centraland
parietal scalp areas and right frontal area (ps<0.05) (Table 1, clusters
3–5). Cluster 4 from Table 1is depicted in the bottom plot of Figure 3.
Activity localized to fronto-parietal and ATL areas
supports auditory memory-guided perception
Early effects
The contrast between theta source-localized predictive and nonpredic-
tive activity during soundscape (cue) presentation yielded a difference
in source activity between predictive and nonpredictive trials in the left
ATL ( p=0.006, corrected, MNI coordinates: −40, 16, −24) (Table 2and
Figure 4, top plot).
Three clusters were identified when contrasting alpha source-
localized predictive and nonpredictive activity between 900 and
1300 ms relative to soundscape (cue) onset. The first two clusters
revealed a difference in source activity between predictive and non-
predictive trials in the right superior parietal lobe (p=0.026, corrected,
MNI coordinates: 31, −64, 58) and left premotor/supplementary motor
areas (p=0.028, corrected, MNI coordinates: −39, 1, 57). The third
cluster revealed a difference in source activity between predictive and
nonpredictive trials in the right anterior prefrontal cortex (p=0.033,
corrected, MNI coordinates: 33, 43, −7) (Table 2and Figure 4, middle
plot).
Late effects
Two clusters were identified when contrasting alpha source-localized
predictive and nonpredictive activity between 1500 and 1900 ms
relative to soundscape (cue) onset, with greater source activity for pre-
dictive trials than for nonpredictive trials in the right inferior parietal
lobe (p1=0.008, corrected, MNI coordinates: 31, −84, 32; p2=0.014,
corrected, MNI coordinates: 52, −68, 17) (Table 2and Figure 4, bottom
plot).
TAB L E 1 Time-frequency (channel-level) cluster results comparing predictive and nonpredictive theta and alpha band-pass filtered data.
Frequency
range (Hz)
Timing
effect Latency (ms) Cluster
Corrected
p-value
Mean
predictive
Mean non-
predictive Electrode
Theta 4 -7 Early 950 - 1200; 1000 max 10.0036 −0.05 0.04 P1 P3 POz CPz C2 C4
CP6CP4CP2P2P8
P10 PO8 CB2
Alpha 8 −12 Early 900 - 1050; 950 max 20.014 0.07 0.24 F2 F4 F6 FC2 FCz
Late 1500 - 1650; 1600 max 30.039 0.91 0.63 TP7 CP5 P5 P7
1600 - 1900; 1700 max 40.011 0.60 0.42 FT10 F10 LO2
1650 – 1900; 1750 max 50.022 1.08 0.78 CP1P1PzCPz

ANNALS OF THE NEW YORK ACADEMY OF SCIENCES 165
FIGURE 3 Time-frequency analyses in scalp space comparing
mean power in both conditions depicting both early and late clusters
in theta (4–7 Hz) and alpha (8–12 Hz) bands following soundscape
(cue) onset. The mean change in power relative to the pre-stimulus
baseline was calculated for both conditions and averaged over the
electrodes identified in clusters 1 (top plot), 2 (middle plot), and 4
(bottom plot) (see Table 1). Shaded regions represent the 95%
within-subject confidence interval. *p<0.05.
Brain−behavior correlations
We analyzed whether cue-related activity in posterior parietal regions
(three clusters) correlated with RT benefits as planned. None of the
posterior parietal clusters correlated with the transformed RT benefit
score (ps>0.05, uncorrected).
Next, we conducted exploratory analyses to investigate whether
differences in neural activity outside our predicted regions (hippocam-
pus and posterior parietal cortex) were associated with RT benefits.
We observed a positive correlation between theta activity in the ATL
and the transformed RT benefit score (r=0.33, p=0.008, corrected,
Figure 5). Alpha activity in the two frontal areas did not show a sig-
nificant correlation with the transformed RT benefit score (ps>0.05,
corrected). A Hotelling−Williams test comparing correlations between
early theta effects, and early and late alpha effects with the trans-
formed RT benefit score revealed no significant differences among the
six correlations (p>0.05, corrected).
Finally, we explored the relationship between theta and alpha
source-localized activity by performing five Pearson correlations
between the ATL cluster and three early and two late clusters
located in the frontal and posterior parietal cortex. This analysis
revealed a significant relationship between ATL activity and the pre-
motor/supplementary motor area (r=0.53, p<0.0001, corrected).
DISCUSSION
The current study aimed to characterize the temporal dynamics of
auditory memory-guided perception. Participants were first exposed
to the soundscape–tone pairings and then tested on these inciden-
tally formed associations in a cue–target detection task. Thus, our
task differs from and complements previous studies in the field by (1)
using auditory stimuli and (2) using a design and contrast that isolate
associative retrieval processes.
Our task was effective in assessing memory-guided perceptual
outcomes. Notably, targets were detected more quickly when embed-
ded in predictive versus nonpredictive soundscapes (memory-guided
detection benefit), demonstrating that the advantages of memory-
guided perception extend beyond visuospatial contexts to include
auditory associative memory. The RT benefit suggests that merely
presenting a soundscape with a tone only four times is sufficient to
improve target detection during retrieval. This benefit was accom-
panied by the above-chance recognition of the soundscape cue and
the presence/absence of the tone. Therefore, although participants
engaged in an incidental learning task, they may have adopted an
explicit retrieval strategy. Because there were not enough success-
ful memory trials to sort the EEG data based on memory judgments,
the neural results indicate the engagement of retrieval processes but
do not distinguish between successful or unsuccessful recall. Future
modifications to the experimental design, such as increasing the num-
ber of learning phases to increase the number of successful memory
trials per condition, would allow for data filtering based on mem-
ory performance, while maintaining an acceptable SNR for accurate

166 ANNALS OF THE NEW YORK ACADEMY OF SCIENCES
TAB L E 2 Beamformer (source-level) cluster results comparing predictive and nonpredictive theta and alpha band-pass filtered data.
Frequency
range (Hz)
Timing
effect Latency (ms)
Corrected
p-value
Mean
predictive
Mean non-
predictive
Cluster maxima
(X,Y,Z;MNI) Brodmannarea
Theta 4–7 Early 900–1300 0.006 −0.77 −3.50 −40, 16, −24 Left temporal pole (38)
Alpha 8–12 Early 900–1300 0.026 9.81 6.22 31, −64, 58 Right superior
parietal lobe (7)
0.028 9.04 4.89 −39, 1, 57 Left premotor/
supplementary motor areas
(6)
0.033 −2.90 −0.27 33, 43, −7Right anterior prefrontal
cortex (10)
Late 1500–1900 0.008 −62.16 −70.47 31, −84, 32 Right angular gyrus (39)
0.014 −64.55 −73.77 52, −68, 17 Right angular gyrus (39)
source-localization. This type of analysis could provide insights into
how implicit versus explicit associative retrieval may guide perception.
Two-stage associative retrieval
Soundscapes that were predictive of the target were neurally dis-
tinct from those that were nonpredictive within 950 ms after cue
onset. Importantly, theta power that was source-localized to the
left ATL (BA 38) distinguished predictive trials from nonpredictive
ones, and this activity scaled with target detection benefits, provid-
ing critical evidence that cue-related activity plays a functional role in
memory-guided perception. Further, differences in alpha power were
source-localized to frontal (BA 6 and BA 10) and parietal areas (BA 7
and BA 39), consistent with fMRI studies that implicate fronto-parietal
networks during visual memory-guided perception.
Theta activity localized to the ATL predicts auditory
memory-guided benefits in perception
Our cue–target detection task enabled us to isolate associative
retrieval from target and response processing. By isolating cue-related
activity, we were able to examine its role in biasing real-time per-
ception on a target-detection task. Theta changes in the left ATL
distinguished predictive from nonpredictive trials. This early stage of
associative retrieval may support the merging of retrieved information
with real-time processing47–49 akin to an early unconscious stage of
recollection.13
Because we used naturalistic soundscapes (e.g., cityscape or cof-
feeshop sound), the ATL activity we observed may reflect the role
of the ATL in semantic cognition.50 During the test, participants may
have drawn on conceptual connections between the soundscape and
the target to facilitate retrieval, which has been shown to improve
spatial memory.51 According to this prediction, differences in cue-
related activity between predictive and nonpredictive trials would be
more pronounced in regions linked to semantic processing, such as the
ATL. Alternatively, the ATL may play an important role in associative
retrieval regardless of semantic content. To this point, Chaumon et al.52
employed a memory-guided detection paradigm, using meaningless
geometric configurations as cues, and likewise observed activity in the
left ATL. These results, along with ours in the auditory domain, provide
complementary and converging evidence for a possible distinct neu-
ral correlate of memory-guided perception that facilitates the retrieval
of associative content to generate predictions for upcoming sensory
processing.
Critically, activity in the ATL predicted the magnitude of the
memory-guided benefit to perception at retrieval, demonstrating that
cue-related activity shapes memory-guided perception. This finding
highlights a functional role for cue-related activity during memory-
guided perception and is consistent with research linking cortical
activity during memory retrieval to performance on tasks of episodic
memory.27
Distinct contributions in time from alpha-localized
posterior parietal cortex during auditory
memory-guided perception
While previous research has highlighted the role of fronto-parietal net-
works in memory-guided perception, our study allows us to examine
the differential recruitment of these areas over time. Alpha oscillations
coordinate long-range structures53–55 and are associated with fronto-
parietal attention56 and control networks that support interactions
between long-term memory and attention.4Given the parietal cor-
tex’s involvement in memory tasks,57–59 the parietal activity observed
in the current study may index a state in which attention is used to
efficiently search through memory. The Attention to Memory model
posits that the posterior parietal cortex is sensitive to the goals and
content of attended representations.60,61 When confidence is low,
and more effort is required, top-down attention is employed by the
superior parietal lobe. In contrast, if an item in memory is easily
accessible, bottom-up attention may be employed to guide attention
automatically to the memorandum. Alpha power is linked to perceptual
sensitivity and the suppression of irrelevant information,62 although
it can be difficult to disentangle the direction of relative changes.63

ANNALS OF THE NEW YORK ACADEMY OF SCIENCES 167
FIGURE 4 Mean source-localized theta or alpha power for
predictive versus nonpredictive trials. The beamformer was
conducted between 900 and 1300 ms, as defined by the
time-frequency cluster. Plots show the difference between conditions
(nonpredictive minus predictive) in source-localized theta and alpha
power. Error bars represent the 95% within-subject confidence
interval (see Table 2). *p<0.05; **p<0.01. Abbreviation: BA,
Brodmann area.
In our study, the observed changes in source-localized alpha power
across two distinct temporal windows align with the Attention to
Memory model.57,61 Specifically, reduced alpha power in the superior
parietal lobe for predictive trials during the early time window may
indicate more efficient processing and attention toward the recovered
information.41 Following this, an increase in alpha power in the inferior
FIGURE 5 Memory-guided source activity in the anterior
temporal lobe plotted as a function of the memory-guided perceptual
benefit in reaction time (RT) at test phase. Shaded region represents
one standard error of the mean (SEM). Each dot represents the
average across trials for a given participant.
parietal lobe for predictive trials in the later time window may reflect
a shift toward greater cognitive control over the retrieved informa-
tion. Collectively,these findings highlight a possible neurophysiological
mechanism for temporally distinct contributions to memory retrieval in
service of perception.
While changes in alpha power did not correlate with RT benefits,
alpha in the premotor cortex correlated with theta-localized ATL activ-
ity. The premotor cortex, connected to fronto-parietal attention and
control networks,64,65 has been implicated in visual memory-guided
perception.3–5,20,25 Given that the premotor cortex is critical for motor
control and the implementation of goal-directed behavior,66 the pre-
motor cortex in this context may serve as a flexible control system for
translating goal-directed processing into action during memory-guided
behavior.
If the observed alpha activity is closely tied to motor control, its
influence on real-time processing during retrieval may have diminished
during the 3-s delay between cue onset and target detection, which
could explain the lack of correlation between cue-related alpha and
memory-guided RT benefits. Future modifications of the experimental
design could directly manipulate the duration between cue and target
to examine the potential relationship between cue-related alpha and
memory-guided benefits to clarify the role of alpha in memory-guided
auditory perception.
Toward a supramodal understanding of
memory-guided perception
Prevailing theories of memory-guided perception are generally
modality-agnostic, suggesting that the processes involved may not

168 ANNALS OF THE NEW YORK ACADEMY OF SCIENCES
FIGURE 6 Cluster maxima from the present empirical EEG study superimposed on fMRI meta-analysis coordinates from Fischer et al.11 and
intracranial EEG coordinates from Chaumon et al.52
be restricted to a single sensory modality.6,7 For instance, incidental
learning has been shown to significantly influence selective attention
in young, healthy adults,67,68 while predictive coding theory posits
that the brain’s predictive mechanisms operate in a domain-general
manner.69 This raises a critical question: do the neural correlates
observed during visuospatial memory-guided perception reflect
domain-general processes, or are they specific to the processing of
visuospatial mnemonic content?
While our four-shell head model achieves a reasonable level of pre-
cision in localizing sources70—approximately 1 cm average error71,72—
this resolution is still lower than the spatial accuracy typically provided
by fMRI. Nevertheless, we observed striking overlap across inde-
pendent datasets. Specifically, our EEG findings show remarkable
alignment with the results from an fMRI meta-analysis on visual
memory-guided perception,11 despite the fact that the datasets dif-
fer significantly in both the modalities tested (vision vs. audition)
and the measurement tools employed (fMRI vs. EEG). As well, all
datasets were analyzed without a priori model constraints, which
makes the observed overlap between studies even more remark-
able. This convergence suggests that there may be a supramodal
set of brain areas that support memory-guided perception, as illus-
trated in Figure 6. These findings open exciting avenues for further
investigation into whether the neural correlates observed during
visuospatial memory-guided perception reflect domain-general pro-
cesses or specific processing of visuospatial mnemonic content. Future
research should focus on quantitatively comparing neural activity dur-
ing memory-guided auditory and visual perceptual tasks. This work
could illuminate the potential supramodal nature of memory-guided
perception.
CONCLUSION
Emerging research highlights the complex relationship between mem-
ory and perception. Our results build on this exciting work and
lay a foundation for understanding how memory retrieval shapes
auditory perception. We have demonstrated that auditory asso-
ciative memory enhances perception, with neural changes at two
distinct temporal intervals following memory cue onset support-
ing this memory-guided effect. Specifically, changes in theta and
alpha power revealed both early and late effects, with early theta
power changes, localized to the ATL, correlating with the memory-
guided detection benefit. Our findings have two key implications:
First, they reveal the neural processes underlying memory-guided
auditory perception, aligning with current research using static visu-
ospatial stimuli. Second, they elucidate how memory retrieval unfolds
over time and demonstrate how cue-related activity shapes ongoing
perception.
AUTHOR CONTRIBUTIONS
M.F.: Conceptualization; methodology; software; investigation; for-
mal analysis; writing—original draft; visualization; project administra-
tion. M.M.: Conceptualization; writing—review and editing; supervi-
sion; funding acquisition. C.A.: Conceptualization; resources; writing—
review and editing; supervision; funding acquisition.
ACKNOWLEDGMENTS
The authors thank Mello Mo and Martin Soto for their help with data
collection. This work was supported by the Natural Sciences and Engi-
neering Research Council of Canada, including an Alexander Graham

ANNALS OF THE NEW YORK ACADEMY OF SCIENCES 169
Bell grant to M.F. and Discovery grants to M.M. [A8347] and C.A.
[RGPIN-2021-02721].
COMPETING INTERESTS
The authors declare that there were no conflicts of interest with
respect to the authorship or the publication of this article.
ORCID
Manda Fischer https://orcid.org/0000-0001-8353-9818
PEER REVIEW
The peer review history for this article is available at: https://publons.
com/publon/10.1111/nyas.15287
REFERENCES
1. Chun, M. M., & Jiang, Y.(1998). Contextual cueing: Implicit learning and
memory of visual context guides spatial attention. Cognitive Psychology,
36(1), 28–71. https://doi.org/10.1006/cogp.1998.0681
2. Giesbrecht, B., Sy, J. L., & Guerin, S. A. (2013). Both memory and atten-
tion systems contribute to visual search for targets cued by implicitly
learned context. Vision Research,85, 80–89. https://doi.org/10.1016/j.
visres.2012.10.006
3. Manginelli, A. A., Baumgartner, F., & Pollmann, S. (2013). Dorsal and
ventral working memory-related brain areas support distinct pro-
cesses in contextual cueing. Neuroimage,67, 363–374. https://doi.org/
10.1016/j.neuroimage.2012.11.025
4. Rosen, M. L., Stern, C. E., Michalka, S. W., Devaney, K. J., & Somers, D.
C. (2016). Cognitive control network contributions to memory-guided
visual attention. Cerebral Cortex,26(5), 2059–2073. https://doi.org/10.
1093/cercor/bhv028
5. Rosen, M. L., Stern, C. E., Devaney, K. J., & Somers, D. C. (2018). Cortical
and subcortical contributions to long-term memory-guided visuospa-
tial attention. Cerebral Cortex,28(8), 2935–2947. https://doi.org/10.
1093/cercor/bhx172
6. Peterson, M. A. (2019). Past experience and meaning affect object
detection: A hierarchical Bayesian approach. In K. D. Federmeier & D.
Beck (Eds.),Psychology of learning and motivation (Vol. 70, pp. 223–257).
Elsevier.
7. Barense, M. D., Ngo, J. K. W., Hung, L. H. T., & Peterson, M. A. (2012).
Interactions of memory and perception in amnesia: The figure–ground
perspective. Cerebral Cortex,22(11), 2680–2691. https://doi.org/10.
1093/cercor/bhr347
8. Sussman, E., Winkler, I., & Schröger, E. (2003). Top-down control
over involuntary attention switching in the auditory modality. Psy-
chonomic Bulletin & Review,10(3), 630–637. https://doi.org/10.3758/
BF03196525
9. Võ, M. L.-H., & Wolfe, J. M. (2013). The interplay of episodic and seman-
tic memory in guiding repeated search in scenes. Cognition,126(2),
198–212. https://doi.org/10.1016/j.cognition.2012.09.017
10. Nickel, A. E., Hopkins, L. S., Minor, G. N., & Hannula, D. E. (2020). Atten-
tion capture by episodic long-term memory. Cognition,201, 104312.
https://doi.org/10.1016/j.cognition.2020.104312
11. Fischer, M., Moscovitch, M., & Alain, C. (2021). A systematic review
and meta-analysis of memory-guided attention: Frontal and parietal
activation suggests involvement of fronto-parietal networks. Wiley
Interdisciplinary Reviews: Cognitive Science,12(1), e1546. https://doi.
org/10.1002/wcs.1546
12. Jiang, Y. V., Sisk, C. A., & Toh, Y. N. (2019). Implicit guidance of atten-
tion in contextual cueing: Neuropsychological and developmental
evidence. Neuroscience & Biobehavioral Reviews,105, 115–125. https://
doi.org/10.1016/j.neubiorev.2019.07.002
13. Moscovitch, M. (2008). The hippocampus as a “stupid,” domain-specific
module: Implications for theories of recent and remote memory,and of
imagination. Canadian Journal of Experimental Psychology /Revue Cana-
dienne De Psychologie Expérimentale,62(1), 62–79. https://doi.org/10.
1037/1196-1961.62.1.62
14. Greene, A. J., Gross, W. L., Elsinger, C. L., & Rao, S. M. (2007). Hip-
pocampal differentiation without recognition: An fMRI analysis of the
contextual cueing task. Learning & Memory,14(8), 548–553. https://doi.
org/10.1101/lm.609807
15. Preston, A. R., & Gabrieli, J. D. E. (2008). Dissociation between explicit
memory and configural memory in the human medial temporal lobe.
Cerebral Cortex,18(9), 2192–2207. https://doi.org/10.1093/cercor/
bhm245
16. Westerberg, C. E., Miller, B. B., Reber, P. J., Cohen, N. J., & Paller,
K. A. (2011). Neural correlates of contextual cueing are modulated
by explicit learning. Neuropsychologia,49(12), 3439–3447. https://doi.
org/10.1016/j.neuropsychologia.2011.08.019
17. Córdova, N. I., Turk-Browne, N. B., & Aly, M. (2019). Focusing on what
matters: Modulation of the human hippocampus by relational atten-
tion. Hippocampus,29(11), 1025–1037. https://doi.org/10.1002/hipo.
23082
18. Olsen, R. K., Moses, S. N., Riggs, L., & Ryan, J. D. (2012). The hippocam-
pus supports multiple cognitive processes through relational binding
and comparison. Frontiers in Human Neuroscience,6, 146. https://doi.
org/10.3389/fnhum.2012.00146
19. Goldfarb, E. V., Chun, M. M., & Phelps, E. A. (2016). Memory-guided
attention: Independent contributions of the hippocampus and stria-
tum. Neuron,89(2), 317–324. https://doi.org/10.1016/j.neuron.2015.
12.014
20. Stokes, M. G., Atherton, K., Patai, E. Z., & Nobre, A. C. (2012). Long-
term memory prepares neural activity for perception. Proceedings of
the National Academy of Sciences,109(6), E360–E367. https://doi.org/
10.1073/pnas.1108555108
21. Summerfield, J. J., Lepsien,J., Gitelman, D. R., Mesulam, M. M., & Nobre,
A. C. (2006). Orienting attention based on long-term memory experi-
ence. Neuron,49(6), 905–916. https://doi.org/10.1016/j.neuron.2006.
01.021
22. Manelis, A., & Reder, L. M. (2012). Procedural learning and associative
memory mechanisms contribute to contextual cueing: Evidence from
fMRI and eye-tracking. Learning & Memory,19(11), 527–534. https://
doi.org/10.1101/lm.025973.112
23. Günseli, E., & Aly, M. (2020). Preparation for upcoming attentional
states in the hippocampus and medial prefrontal cortex. eLife,9,
e53191.
24. Pollmann, S. (2009). Anterior prefrontal involvement in implicit con-
textual change detection. Frontiers in Human Neuroscience,3, 28.
https://doi.org/10.3389/neuro.09.028.2009
25. Pollmann, S., & Manginelli, A. A. (2010). Repeated contextual search
cues lead to reduced BOLD-onset times in early visual and left inferior
frontal cortex. Open Neuroimaging Journal,4, 9–15. https://doi.org/10.
2174/1874440001004010009
26. Pollmann, S., Eštoˇ
cinová, J., Sommer, S., Chelazzi, L., & Zinke, W.
(2016). Neural structures involved in visual search guidance by
reward-enhanced contextual cueing of the target location. Neuroim-
age,124, 887–897. https://doi.org/10.1016/j.neuroimage.2015.09.
040
27. Trelle, A. N., Carr, V. A., Guerin, S. A., Thieu, M. K., Jayakumar, M., Guo,
W., Nadiadwala, A., Corso, N. K., Hunt, M. P., Litovsky, C. P., Tanner, N.
J., Deutsch, G. K., Bernstein, J. D., Harrison, M. B., Khazenzon, A. M.,
Jiang, J., Sha, S. J., Fredericks, C. A., Rutt, B. K., .. . Wagner, A. D. (2020).
Hippocampal and cortical mechanisms at retrieval explain variability in

170 ANNALS OF THE NEW YORK ACADEMY OF SCIENCES
episodic remembering in older adults. eLife,9, e55335. https://doi.org/
10.7554/eLife.55335
28. Vivekananda, U., Bush, D., Bisby, J. A., Baxendale, S., Rodionov, R.,
Diehl, B., Chowdhury, F. A., McEvoy, A. W., Miserocchi, A., Walker,
M. C., & Burgess, N. (2021). Theta power and theta-gamma cou-
pling support long-term spatial memory retrieval. Hippocampus,31(2),
213–220. https://doi.org/10.1002/hipo.23284
29. Fischer, M., Moscovitch, M., Fukuda, K., & Alain, C. (2023). Ready
for action! When the brain learns, yet memory-biased action does
not follow. Neuropsychologia,189, 108660. https://doi.org/10.1016/j.
neuropsychologia.2023.108660
30. Kunar, M. A., Flusberg, S. J., & Wolfe, J. M. (2006). Contextual cuing by
global features. Perception & Psychophysics,68(7), 1204–1216. https://
doi.org/10.3758/BF03193721
31. Schankin, A., & Schubö, A. (2009). Cognitive processes facilitated by
contextual cueing: Evidence from event-related brain potentials. Psy-
chophysiology,46(3), 668–679. https://doi.org/10.1111/j.1469-8986.
2009.00807.x
32. Schankin, A., & Schubö, A. (2010). Contextual cueing effects despite
spatially cued target locations. Psychophysiology,47(4), 717–727.
https://doi.org/10.1111/j.1469-8986.2010.00979.x
33. Zimmermann, J., Moscovitch, M., & Alain, C. (2017). Long-term mem-
ory biases auditory spatial attention. Journal of Experimental Psychol-
ogy: Learning, Memory, and Cognition,43(10), 1602–1615. https://doi.
org/10.1037/xlm0000398
34. Herweg, N. A., Solomon, E. A., & Kahana, M. J. (2020). Theta oscillations
in human memory. Trendsin Cognitive Sciences,24(3), 208–227. https://
doi.org/10.1016/j.tics.2019.12.006
35. Lim, S.-J., Wöstmann, M., & Obleser, J. (2015). Selective attention
to auditory memory neurally enhances perceptual precision. Jour-
nal of Neuroscience,35(49), 16094–16104. https://doi.org/10.1523/
JNEUROSCI.2674-15.2015
36. Fischer, M., Moscovitch, M., & Alain, C. (2023). [Unpublished raw
data on Memory-guided auditory perception is robust to temporal jitter].
University of Toronto.
37. Faul, F., Erdfelder, E., Buchner, A., & Lang, A.-G. (2009). Statistical
power analyses using G*Power 3.1: Tests for correlation and regres-
sion analyses. Behavior Research Methods,41(4), 1149–1160. https://
doi.org/10.3758/BRM.41.4.1149
38. Shanks, D. R. (2005). Implicit learning. In K. Lamberts, & R. Goldstone
(Eds.), Handbook of cognition (pp. 202–220). Sage.
39. Olson, I. R., & Chun, M. M. (2001). Temporal contextual cuing of visual
attention. Journal of Experimental Psychology: Learning, Memory, and
Cognition,27(5), 1299–1313. https://doi.org/10.1037/0278-7393.27.
5.1299
40. Zimmermann, J., Ross, B., Moscovitch, M., & Alain, C. (2020). Neu-
ral dynamics supporting auditory long-term memory effects on tar-
get detection. Neuroimage,218, 116979. https://doi.org/10.1016/j.
neuroimage.2020.116979
41. Backer, K. C., Binns, M. A., & Alain, C. (2015). Neural dynamics under-
lying attentional orienting to auditory representations in short-term
memory.Journal of Neuroscience,35(3), 1307–1318. https://doi.org/10.
1523/JNEUROSCI.1487-14.2015
42. Gross, J., Kujala, J., Hämäläinen, M., Timmermann, L., Schnitzler, A.,
& Salmelin, R. (2001). Dynamic imaging of coherent sources: Study-
ing neural interactions in the human brain. Proceedings of the National
Academy of Sciences,98(2), 694–699. https://doi.org/10.1073/pnas.98.
2.694
43. Green, J. J., & McDonald, J. J. (2009). A practical guide to beamformer
source reconstruction for EEG. In T.C. Handy (Ed.), Brain signal analysis
(pp. 79–98). MIT Press.
44. Southwell, R., Baumann, A., Gal, C., Barascud, N., Friston, K., & Chait,
M. (2017). Is predictability salient? A study of attentional capture by
auditory patterns. Philosophical Transactions of the Royal Society B: Bio-
logical Sciences,372(1714), 20160105. https://doi.org/10.1098/rstb.
2016.0105
45. Vanhatalo, S., Voipio, J., & Kaila, K. (2005). Full-band EEG (FbEEG):
An emerging standard in electroencephalography. Clinical Neurophys-
iology,116(1), 1–8. https://doi.org/10.1016/j.clinph.2004.09.015
46. Maris, E., & Oostenveld, R. (2007). Nonparametric statistical test-
ing of EEG- and MEG-data. Journal of Neuroscience Methods,164(1),
177–190. https://doi.org/10.1016/j.jneumeth.2007.03.024
47. Miyashita, Y. (1993). Inferior temporal cortex: Where visual percep-
tion meets memory. Annual Review of Neuroscience,16(1), 245–263.
https://doi.org/10.1146/annurev.ne.16.030193.001333
48. Trimmel, K., Caciagli, L., Xiao, F., Van Graan, L. A., Koepp, M. J.,
Thompson, P. J., & Duncan, J. S. (2021). Impaired naming performance
in temporal lobe epilepsy: Language fMRI responses are modulated by
disease characteristics. Journal of Neurology,268(1), 147–160. https://
doi.org/10.1007/s00415-020- 10116-x
49. Tsapkini, K., Frangakis, C. E., & Hillis, A. E. (2011). The function of the
left anterior temporal pole: Evidence from acute stroke and infarct
volume. Brain,134(10), 3094–3105. https://doi.org/10.1093/brain/
awr050
50. Ralph, M. A. L., Jefferies, E., Patterson, K., & Rogers, T. T. (2017). The
neural and computational bases of semantic cognition. Nature Reviews
Neuroscience,18(1), 42–55. https://doi.org/10.1038/nrn.2016.150
51. Taevs, M., Dahmani, L., Zatorre, R. J., & Bohbot, V. D. (2010). Seman-
tic elaboration in auditory and visual spatial memory. Frontier s in
Psychology,1, 228. https://doi.org/10.3389/fpsyg.2010.00228
52. Chaumon, M., Schwartz, D., & Tallon-Baudry, C. (2009). Unconscious
learning versus visual perception: Dissociable roles for gamma oscil-
lations revealed in MEG. Journal of Cognitive Neuroscience,21(12),
2287–2299. https://doi.org/10.1162/jocn.2008.21155
53. Donner, T. H., & Siegel, M. (2011). A framework for local cortical oscil-
lation patterns. Trendsin Cognitive Sciences,15(5), 191–199. https://doi.
org/10.1016/j.tics.2011.03.007
54. Fries, P.(2015). Rhythms for cognition: Communication through coher-
ence. Neuron,88(1), 220–235. https://doi.org/10.1016/j.neuron.2015.
09.034
55. Helfrich, R. F.,& Knight, R. T. (2016). Oscillatory dynamics of prefrontal
cognitive control. Trends in Cognitive Sciences,20(12), 916–930. https://
doi.org/10.1016/j.tics.2016.09.007
56. Van Schouwenburg, M. R., Zanto, T. P., & Gazzaley, A. (2017). Spa-
tial attention and the effects of frontoparietal alpha band stimulation.
Frontiers in Human Neuroscience,10, 658. https://doi.org/10.3389/
fnhum.2016.00658
57. Cabeza, R. (2008). Role of parietal regions in episodic memory
retrieval: The dual attentional processes hypothesis. Neuropsychologia,
46(7), 1813–1827. https://doi.org/10.1016/j.neuropsychologia.2008.
03.019
58. Gilmore, A. W., Nelson, S. M., & McDermott, K. B. (2015). A parietal
memory network revealed by multiple MRI methods. Trends in Cog-
nitive Sciences,19(9), 534–543. https://doi.org/10.1016/j.tics.2015.07.
004
59. Wagner, A. D., Shannon, B. J., Kahn, I., & Buckner, R. L. (2005). Pari-
etal lobe contributions to episodic memory retrieval. Trends in Cognitive
Sciences,9(9), 445–453. https://doi.org/10.1016/j.tics.2005.07.001
60. Cabeza, R., Mazuz, Y. S., Stokes, J., Kragel, J. E., Woldorff, M. G.,
Ciaramelli, E., Olson, I. R., & Moscovitch, M. (2011). Overlapping pari-
etal activity in memory and perception: Evidence for the attention to
memory model. Journal of Cognitive Neuroscience,23(11), 3209–3217.
https://doi.org/10.1162/jocn_a_00065
61. Ciaramelli, E., Grady, C. L., & Moscovitch, M. (2008). Top-down and
bottom-up attention to memory: A hypothesis (AtoM) on the role of
the posterior parietal cortex in memory retrieval. Neuropsychologia,
46(7), 1828–1851. https://doi.org/10.1016/j.neuropsychologia.2008.
03.022

ANNALS OF THE NEW YORK ACADEMY OF SCIENCES 171
62. Kelly, S. P., Lalor, E. C., Reilly, R. B., & Foxe, J. J. (2006). Increases in
alpha oscillatory power reflect an active retinotopic mechanism for
distracter suppression during sustained visuospatial attention. Jour-
nal of Neurophysiology,95(6), 3844–3851. https://doi.org/10.1152/jn.
01234.2005
63. Rihs, T.,Michel, C., & Thut, G. (2009). A bias for posterior α-band power
suppression versus enhancement during shifting versus maintenance
of spatial attention. Neuroimage,44(1), 190–199. https://doi.org/10.
1016/j.neuroimage.2008.08.022
64. Grafton, S. T., Fagg, A. H., & Arbib, M. A. (1998). Dorsal premotor cor-
tex and conditional movement selection: A PET functional mapping
study. Journal of Neurophysiology,79(2), 1092–1097. https://doi.org/
10.1152/jn.1998.79.2.1092
65. Pesaran, B., Nelson, M. J., & Andersen, R. A. (2006). Dorsal premotor
neurons encode the relative position of the hand, eye, and goal dur-
ing reach planning. Neuron,51(1), 125–134. https://doi.org/10.1016/j.
neuron.2006.05.025
66. Gremel, C. M., & Costa, R. M. (2013). Premotor cortex is critical for
goal-directed actions. Frontiers in Computational Neuroscience,7, 110.
https://doi.org/10.3389/fncom.2013.00110
67. Awh, E., Belopolsky, A. V., & Theeuwes, J. (2012). Top-down versus
bottom-up attentional control: A failed theoretical dichotomy. Trends
in Cognitive Sciences,16(8), 437–443. https://doi.org/10.1016/j.tics.
2012.06.010
68. Jiang, Y. V. (2018). Habitual versus goal-driven attention. Cortex,102,
107–120. https://doi.org/10.1016/j.cortex.2017.06.018
69. Friston, K. (2005). A theory of cortical responses. Philosophical Trans-
actions of the Royal Society B: Biological Sciences,360(1456), 815–836.
https://doi.org/10.1098/rstb.2005.1622
70. Slotnick, S. D. (2004). Source localization of ERP generators. In T. C.
Todd (Ed.), Event-related potentials: A methods handbook (pp. 149–166).
MIT Press.
71. Cohen, D., Cuffin, B. N., Yunokuchi, K., Maniewski, R., Purcell, C.,
Cosgrove, G. R., Ives, J., Kennedy, J. G., & Schomer, D. L. (1990). MEG
versus EEG localization test using implanted sources in the human
brain. Annals of Neurology,28(6), 811–817. https://doi.org/10.1002/
ana.410280613
72. Cuffin, B. N., Cohen, D., Yunokuchi, K., Maniewski, R., Purcell, C.,
Cosgrove, G. R., Ives, J., Kennedy, J., & Schomer, D. (1991). Tests
of EEG localization accuracy using implanted sources in the human
brain. Annals of Neurology,29(2), 132–138. https://doi.org/10.1002/
ana.410290204
How to cite this article: Fischer, M., Moscovitch, M., & Alain,
C. (2025). Memory-guided perception is shaped by dynamic
two-stage theta- and alpha-mediated retrieval. Ann NY Acad
Sci.,1544, 159–171. https://doi.org/10.1111/nyas.15287