This article appeared in a journal published by Elsevier. The... copy is furnished to the author for internal non-commercial research

This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Brain and Cognition 69 (2009) 180–187
Contents lists available at ScienceDirect
Brain and Cognition
journal homepage: www.elsevier.com/locate/b&c
Mental rotation of mirrored letters: Evidence from event-related brain potentials q
M. Isabel Núñez-Peña a,b,*, J. Antonio Aznar-Casanova b,c
a
Department of Behavioral Science Methods, Faculty of Psychology, University of Barcelona, Passeig Vall d’Hebron, 171, 08035 Barcelona, Spain
Cognitive Neuroscience Research Group, Department of Psychiatry and Clinical Psychobiology, Faculty of Psychology, University of Barcelona,
Passeig Vall d’Hebron, 171, 08035 Barcelona, Spain
c
Department of Basic Psychology, Faculty of Psychology, University of Barcelona, Passeig Vall d’Hebron, 171, 08035 Barcelona, Spain
b
a r t i c l e
i n f o
Article history:
Accepted 10 July 2008
Available online 17 August 2008
Keywords:
Event-related brain potentials
Rotation-related negativity
Mirrored letters
Parity-judgment task
a b s t r a c t
Event-related brain potentials (ERPs) were recorded while participants (n = 13) were presented with mirrored and normal letters at different orientations and were asked to make mirror-normal letter discriminations. As it has been suggested that a mental rotation out of the plane might be necessary to decide on
mirrored letters, we wanted to determine whether this rotation occurs after the plane rotation in mirror
rotated letters. The results showed that mirrored letters in the upright position elicited a negative-going
waveform over the right hemisphere in the 400–500 ms window. A similar negativity was also present in
mirrored letters at 30°, 60°, and 90°, but in these cases it was delayed. Moreover, the well-known orientation effect on the amplitude of the rotation-related negativity was also found, although it was more evident for normal than for mirrored letters. These results indicate that the processing of mirrored letters
differs from that of normal letters, and suggest that a rotation out of the plane after the plane rotation
may be involved in the processing of mirror rotated letters.
Ó 2008 Elsevier Inc. All rights reserved.
1. Introduction
Mental rotation is a classical psychological process. It was first
reported by Shepard and Metzler (1971) in an experiment where
participants were presented with pairs of three-dimensional block
figures at different orientations, and were required to determine
whether both figures were the same or one was a mirror reflection
of the other. Results showed that reaction time (RT) was longer for
larger angles of misorientation. It was proposed that this increased
RT was due to the fact that in order to perform the parity-judgment
task the image had to be mentally rotated to put it in the upright
position. Since then, the mental rotation effect has been reported
in studies with alphanumeric characters (Cooper & Shepard,
1973; Koriat & Norman, 1985a), letter-like characters (Tarr & Pinker, 1989), left-right hands (Cooper & Shepard, 1975), and even in a
naming task with natural objects (Jolicoeur, 1985, 1988, 1990).
Although the mental rotation effect has been reported with different types of stimuli, it has been suggested that the form of these RT
functions depends on the familiarity of the stimuli. When the stimulus is unfamiliar the RT function is linear (Shepard & Metzler,
1971), whereas when the stimulus is familiar—i.e., alphanumeric
characters—the RT function departs from linearity and shows a
q
This research was supported by Grants SEJ2006-000496/PSIC, SEJ2006-15095/
PSIC, and Consolider-Ingenio 2010-CSD2007-00012 from the Spanish Ministry of
Science and Technology, and SGR2005-00953 from the Generalitat de Catalunya.
* Corresponding author. Fax: +34 93 402 13 59.
E-mail address: [email protected] (M.I. Núñez-Peña).
0278-2626/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.bandc.2008.07.003
quadratic trend (Cooper & Shepard, 1973). The suggested explanation for this nonlinearity effect is that familiar stimuli are overlearned visual stimuli that achieve a certain degree of
indifference to small misorientations from their normal position
(Cooper & Shepard, 1973; Koriat & Norman, 1985b). However,
unfamiliar stimuli require rotation even for small deviations from
upright.
Similar mathematical functions relate the angle of misorientation and the amplitude of an event-related brain potential (ERP)
component in mental rotation tasks. This component, known as
‘rotation-related negativity’, was first reported by Stuss, Sarazin,
Leech, and Picton (1983), Peronnet and Farah (1989), and Wijers,
Otten, Feenstra, Mulder, and Mulder (1989). It consists of a negative-going waveform, maximum over parietal regions, whose
amplitude is modulated by the angle of misorientation: the greater
the angle of misorientation, the larger the rotation-related negativity. The rotation-related negativity has been reported in studies
with alphanumeric characters (Heil, Rauch, & Hennighausen,
1998; Heil & Rolke, 2002; Milivojevic, Johnson, Hamm & Corbalis,
2003), letter-like shapes (Núñez-Peña, Aznar, Linares, Corral & Escera, 2005), paper-folding stimuli (Milivojevic, et al., 2003), leftright hands (Thayer & Johnson, 2006), and geometric objects
(Muthukumaraswamy, Johnson, & Hamm, 2003; Rösler, Heil, Bajric, Pauls, & Hennighausen, 1995). It has been suggested that this
component is a neurophysiological correlate of the mental rotation
process (Heil, 2002), because its amplitude is modulated by the
amount of mental rotation needed to make a parity decision.
Moreover, there are other evidences. First, Heil, Bajric, Rösler,
Author's personal copy
M.I. Núñez-Peña, J.A. Aznar-Casanova / Brain and Cognition 69 (2009) 180–187
and Hennighausen (1996a) and Heil et al. (1998) found that the
rotation-related negativity was evoked by a misoriented stimulus
only if mental rotation is required to solve the task. Second, Heil
and Rolke (2002) provided evidence that the onset of this negative
component is delayed by delaying the mental rotation process.
As regards the spatial distribution of the mental rotation process, neuroimaging and electrophysiological studies have reported
inconsistent results. Although positron emission tomography
(PET) and functional magnetic resonance imaging (fMRI) studies
have reported clear evidence of the involvement of parietal regions in mental rotation, considerable debate remains as to
whether mental rotation is a right parietal function or whether
neither hemisphere is dominant (Alivisatos & Petrides, 1997;
Cohen et al., 1996; Harris et al, 2000; Jordan, Heinze, Lutz,
Kanowski, & Jancke, 2001; Richter, Ugurbil, Georgopoulos, &
Kim, 1997; Yoshino, Inoue, & Suzuki, 2000). Milivojevic et al.
(2003) suggested that the type of task might account for some
of these contradictory results: ‘the right hemisphere may be preferentially engaged when the task is simple and involves a single
transformation, but the left hemisphere is also engaged as the
task becomes more complex, as when a coordinated sequence
of transformations are required’ (Milivojevic et al., p. 1355). This
explanation agrees with that proposed by Corballis (1997), who
differentiates between holistic and analytic mental rotation processes. According to his view, the right hemisphere is preferentially engaged in holistic mental rotation processes—when the
entire image is mentally rotated in a unitary process—and the left
hemisphere is preferentially engaged in analytic processes—when
the image is parsed into units, which are then rotated individually. However, there is also some evidence that the right hemisphere contribution to spatial performance increases with the
complexity of the task. Roberts and Bell (2003) reported greater
activation of the right parietal region in a three-dimensional mental rotation task than in a two-dimensional one. Alivisatos and
Petrides (1997) provided evidence that activity in the left parietal
cortex was more intense in a task that required active mental
rotation in the picture plane than in one that requires making a
mirror-normal decision regarding upright letters.
Whereas the change in orientation has been extensively investigated with both behavioral and psychophysiological measures,
the mirror-normal difference has attracted less interest among scientists. Behavioral studies have systematically shown that a mirrored stimulus decision takes longer than a normal stimulus
decision (see for example, Bajric, Rösler, Heil, & Hennighaugen,
1999; Hamm, Johnson, & Corballis, 2004; Milivojevic et al.,
2003). A suggested explanation for this difference in RT is that
the mirrored stimulus is rotated both in the picture plane, in order
to put it in the vertical upright position, and out of the plane, in order to put it in the normal upright position. This explanation is
supported by several psychophysiological studies. First, Alivisatos
and Petrides (1997), in a PET study, found that mirror-normal
judgment of upright letters activated similar brain areas to those
activated in a classical mental rotation task. This study suggests
that both experimental tasks—the mirror-normal judgment task
with upright letters and the mirror-normal judgment task with
letters presented at different orientations—require visuo-spatial
processing to identify misoriented stimuli. Second, in a recent
ERP study, Hamm et al. (2004) concluded that mirrored stimuli
are not only rotated in the picture plane but are subsequently rotated out of the plane, involving a ‘flip’ to fully normalize the mirrored stimuli.
While it seems clear that mirrored letters in the upright condition need to be rotated out of the picture plane in order to make a
mirror-normal judgment, the case of a mirrored letter presented at
an orientation different from upright has been less studied. Hamm
et al. (2004) suggested that the flipping of mirrored stimuli will
181
occur at whatever orientation and that it will occur after the plane
rotation. They stated that this flipping of mirrored stimuli ‘‘will occur after the plane rotation because of the theoretical complications that arise if one postulates it as occurring prior to the plane
rotation” (p. 819). If information about the mirror-normal status
of the stimuli is available before the plane rotation, then rotating
the mirrored stimuli in the picture plane will be unnecessary.
However, no evidence to support this sequential processing has
been brought forward so far.
The purpose of the present study was to add evidence in support of the ideas that (1) mental rotation out of the picture plane
is necessary to make a mirror-normal judgment in mirrored letters
and (2) this mental rotation occurs after the plane rotation in orientations different from the upright. Participants were presented
with mirrored and normal letters at eleven different orientations—the 0° orientation and the 30°, 60°, 90°, 120°, or 150° clockwise or counterclockwise orientations—and were asked to perform
a parity-judgment task. The decision on the normal versions of letters requires only the mental rotation of the stimulus in the picture
plane, whereas the decision on the mirrored versions of letters requires mental rotation of the stimulus in the picture plane and an
extra rotation out of the plane. Mirrored upright letters, where the
decision requires only rotation out of the picture plane, served as a
control condition to isolate the flip effect. Once this effect had been
isolated, we performed a detailed analysis of the ERPs at different
50-ms windows in order to study the mirror-normal difference in
other rotated stimuli. We hypothesized that if a mirror rotated letter is rotated out of the picture plane after the plane rotation, then
the flip effect (the difference mirror-normal) would be delayed in
the ERP pattern. Moreover, it was predicted that the typical modulation of the amplitude of the rotation-related negativity would
be present in normal letters and that this ERP pattern would be different for mirrored letters, where rotation in and out of the picture
plane would be needed.
2. Methods
2.1. Participants
Fifteen healthy volunteers were tested in this study (12 women;
age 19–28 years, mean = 21.8, standard deviation = 2.5). All were
university students and had normal or corrected-to-normal visual
acuity. Because of a large number of artifacts, data from two participants were excluded from the ERP data analysis; this analysis was
thus performed with data from thirteen subjects (10 women; age
19–28 years, mean = 21.9, standard deviation = 2.7). Subjects had
no history of neurological or psychiatric disorder, and gave written
informed consent to participate after the nature of the study had
been explained to them.
2.2. Stimuli and procedure
The characters were the uppercase letters F, L, P, and R, which
were presented at an orientation of 0° or at 30°, 60°, 90°, 120°, or
150° clockwise or counterclockwise orientations. Fig. 1 shows
some of the stimuli used in the experiment.
Stimuli were shown in green on a white background (luminance
110 cd/m2), and at an orientation of 0° subtended a vertical visual
angle of 2.3° and a horizontal visual angle of 1.37°. The program
used to manage the experiment was developed by the authors
using C++/Open GL (glut library).
Participants were seated in an electrically shielded, soundattenuating room at a distance of 150 cm from the display screen,
whose center was at eye level. They were monitored continuously
with a closed circuit video camera. The experiment started with a
training period to familiarize participants with the procedure and
Author's personal copy
182
M.I. Núñez-Peña, J.A. Aznar-Casanova / Brain and Cognition 69 (2009) 180–187
Fig. 1. Normal and mirrored versions of the F letter in each orientation (from 0° to 330°).
the equipment. All subjects achieved a minimum of 90% correct answers in the practice trials.
During the recording period, subjects were instructed to relax
and to keep their eyes on the screen. They were encouraged to
make any eye-blinks during the presentation of a fixation point
or during the pauses. The sequence of events began with a red
fixation point presentation at the center of the screen that remained in view for 500 ms. The character was then presented
for 500 ms, after which the screen remained black until the subject pressed one of the two response buttons. The task was to decide whether each character displayed was presented in a normal
or a mirrored version, as quickly as possible, while keeping errors
to a minimum. Response fingers were counterbalanced across
subjects.
Each participant was given ten blocks of 96 trials. A message
indicating a 1-min pause appeared on the screen after each block,
and a 5-min pause was provided to participants halfway through
the experimental trials. The type of trials was controlled within
each block in such a way that a block included trials resulting from
the factorial combination of the following variables: orientation (0°,
30°, 60°, 90°, 120°, and 150°),1 direction of the angular disparity
(clockwise and counterclockwise), version (normal and mirror-reversed), and type of stimulus (four letters). The sequence of presentation in each block was randomized per participant. All
participants were tested on 960 trials, 40 for each experimental
condition resulting from the combination of orientation, direction
of the angular disparity and version.
eye movements an electrode placed at the external canthi of the
right eye was used.
2.4. Data analysis
2.4.1. Behavioral data
Response times for correctly solved trials and error rate were
analyzed with repeated-measures ANOVAs, taking version (normal
and mirror-reversed) and orientation (0°, 30°, 60°, 90°, 120°, and
150°) as within-subjects factors.2 The repeated-measures ANOVA
was performed with the Greenhouse–Geisser correction for sphericity departures, which was applied when appropriate. The F value,
the uncorrected degrees of freedom, the probability level following
correction, the e value and the g2 effect size index (Kirk, 1996) are reported. Whenever a main effect reached significance, pairwise comparisons were conducted using t tests, and the Hochberg approach
was used to control for the increase in Type I error (Keselman,
1998). Tests of simple effects were calculated in the presence of a significant interaction. Finally, trend analyses were also performed.
EEG was recorded with the SynAmps/SCAN 4.3 hardware and
software (NeuroScan, Inc., Herndon, VA) from 31 tin electrodes
mounted in a commercial electro-cap (Electro-Cap International,
Eaton, OH). Nineteen electrodes were positioned according to the
10–20 International System: three electrodes were placed over
midline sites at Fz, Cz, and Pz locations, along with 8 lateral pairs
of electrodes over standard sites on frontal (FP1/FP2, F7/F8, F3/
F4), central (C3/C4), temporal (T3/T4, T5/T6), parietal (P3/P4),
and occipital (O1/O2) positions. Two electrodes were placed at
Fpz and Oz, and ten electrodes were placed halfway between the
following additional locations: fronto-central (FC1/FC2), frontotemporal (FT3/FT4), centro-parietal (CP1/CP2), temporo-parietal
(TP3/TP4), and mastoids (M1/M2). The common reference electrode for EEG and EOG measurements was placed on the tip of
the nose. EEG channels were continuously digitized at a rate of
500 Hz by a SynAmpTM amplifier (5083 model, NeuroScan, Inc.,
Herndon, VA). A band pass filter was set from 0.16 to 30 Hz, and
electrode impedance was always kept below 5 kX. For monitoring
2.4.2. EEG analysis
Only trials on which the subjects responded correctly were included in the ERP analysis. First, epochs for every subject in each
experimental condition were averaged relative to a pre-stimulus
baseline consisting of the 100 ms of activity preceding the epoch
of interest. Second, trials with artifacts (voltage exceeding ±50 lV
in FP1, FP2, FPz, or HEOG) and those with response errors were excluded from the ERP average. The mean number of epochs included
in each ERP average varied between 45.9 and 58.5 for the various
types of stimuli used.
The orientation effect was studied by analyzing mean amplitude measures in the 400–500 ms window. This latency window
was selected because according to visual inspection of ERP waveforms it was representative of the orientation effect. A
2 6 3 5 repeated-measures ANOVA was performed on the
ERP amplitudes at 15 electrodes (F7, F3, Fz, F4, F8, T3, C3, Cz, C4,
T4, T5, P3, Pz, P4, and T6), taking as factors version (normal and
mirror-reversed), orientation (0°, 30°, 60°, 90°, 120°, and 150°),
frontality (frontal, central, and parietal), and laterality (five levels
from left to right). Statistical analyses were performed as described
for behavioral data. Topographic maps were plotted using the EEProbe 3.1 program (ANT Software BV, Enschede, The Netherlands).
The version effect was studied in a more detailed way in order
to detect whether the mirror-normal difference was delayed across
orientations. Mean amplitude measures in 50-ms windows from
400 to 700 ms at nine electrodes (F3, Fz, F4, C3, Cz, C4, P3, Pz,
and P4,) were analyzed. Repeated-measures ANOVAs were performed at each orientation, taking as factors version (normal and
1
The letters presented at 0° were presented twice as often as letters at the other
orientations because the 30°, 60°, 90°, 120°, and 150° clockwise and counterclockwise
are usually treated as equivalent.
2
Response times were subjected to an initial analysis to test for symmetry about
0°. No asymmetries were detected, so data were collapsed into six orientations (0°,
30°, 60°, 90°, 120°, and 150°) for all analyses.
2.3. Electrophysiological recording
Author's personal copy
M.I. Núñez-Peña, J.A. Aznar-Casanova / Brain and Cognition 69 (2009) 180–187
mirror-reversed), frontality (frontal, central, and parietal), and laterality (left, middle, and right).
3. Results
183
interaction Orientation Version, tests of simple effects showed
that accuracy was worse in normal than in mirrored letters at
150° angular disparity from upright (F(1, 14) = 18.10, p < .001,
g2 = .56). However, there were no differences in accuracy between
the two versions of letters for the other orientations.
3.1. Behavioral data
3.2. Event-related potentials
Mean response times for normal and mirrored letters as a function of stimulus orientation are plotted in Fig. 2. Overall response
time increased with angular deviation from upright
(F(5, 70) = 44.47, p < .001, e = .23, g2 = .76), and responses to normal
letters were faster than responses to mirrored letters
(F(1, 14) = 25.20, p < .001, g2 = .64). However, the ANOVA also
yielded a significant Orientation Version interaction (F(5, 70) =
4.44, p = .012, e = .51, g2 = .24), so the orientation effect varied
according to the letter version. A more detailed analysis of the
interaction showed that the orientation effect in normal letters
was described by a linear (F(1, 14) = 73.31, p < .001, g2 = .84) and
a quadratic (F(1, 14) = 34.97, p < .001, g2 = .71) trend, whereas the
same effect in mirror-reversed letters was described by a linear
trend (F(1, 14) = 33.08, p < .001, g2 = .70). Moreover, tests of simple
effects demonstrated that responses to mirrored stimuli were
slower than to normal stimuli at all the orientations (all p-values
< .001) except for 150 degrees (p = .07). Therefore the advantage
of normal over mirrored letters decreases gradually with increasing angular deviation from upright.
Error rate ANOVA showed a significant effect for orientation
(F(5, 70) = 24.54, p < .001, e = .29, g2 = .64), version (F(1, 14) = 9.26,
p = .009, g2 = .40) and the interaction Orientation Version
(F(5, 70) = 15.17, p < .001, e = .29, g2 = .52). Overall, error rate increased with greater angular disparity from the upright. The orientation effect was described by a linear and a quadratic trend both
in mirrored (F(1, 14) = 8.10, p = .013, g2 = .37 for linear trend, and
F(1, 14) = 17.75, p = .001, g2 = .56 for quadratic trend) and normal
letters (F(1, 14) = 26.42, p < .001, g2 = .65 for linear trend, and
F(1, 14) = 31.22, p < .001, g2 = .69 for quadratic trend). As for the
Fig. 2. Response time means (in milliseconds) for normal and mirrored letters as a
function of stimulus orientation (in degrees).
Fig. 3A and B show the grand-average ERPs for each orientation of normal and mirror reversed letters at P3, Pz, and P4. The
orientation effect is evident for normal letters. As can be seen
in Fig. 3A, the rotation-related negativity becomes more negative with increasing angular disparity from upright, the effect
being more evident for larger deviations. However, the orientation effect is not so clear for mirror-reversed letters (see Fig.
3B): although the voltage tends to be more negative the greater
the degree to be rotated, these differences seem not to be as
large as those for normal letters. Fig. 4 shows the amplitude
means in the 400–500 ms window for normal and mirror reversed letters as a function of stimulus orientation at P3, Pz
and P4. These plots show again that the orientation effect is
different for normal and for mirror-reversed letters: the orientation effect over the ERP amplitude is more evident for normal
letters than for mirrored letters. Voltage maps in Fig. 5A and
B showed the spatial distribution of the orientation effect in
normal and mirror-reversed letters over all electrodes at the
scalp surface in the 400–500 ms window. These voltage maps
show that the orientation effect has a centro-parietal scalp distribution in normal letters and is not so clear for mirrored
letters.
The statistical analysis performed on the 400–500 ms window
supports these observations. The overall ANOVA showed significant effects of orientation (F(5, 60) = 11.40, p < .001, e = .54,
g2 = .49), Orientation Version (F(5, 60) = 2.68, p = .03, e = .77,
g2 = .18), Orientation Version Frontality (F(10, 120) = 4.37,
p = .003, e = .42, g2 = .27) and Orientation Version Laterality
(F(20, 240) = 2.46, p = .035, e = .29, g2 = .17). A more detailed analysis of the Orientation Version effect was carried out by performing ANOVAs at frontal, central and parietal sites. The
Orientation Version effect reached statistical significance at central (F(5, 60) = 2.79, p = .025, e = .76, g2 = .18) and parietal sites
(F(5, 60) = 3.98, p = .003, e = .71, g2 = .25). The Orientation Version Laterality interaction reached statistical significance at central sites (F(20, 240) = 2.48, p = .032, e = .21, g2 = .17).
The analysis performed at parietal sites revealed that the orientation effect was significant for mirrored (F(5, 60) = 3.56, p = .024,
e = .60, g2 = .23) and normal letters (F(5, 60) = 17.14, p < .001,
e = .65, g2 = .59). A linear trend could be fitted for both mirrored
(F(1, 12) = 7.87, p = .016, g2 = .40) and normal letters (F(1, 12) =
48.20, p < .001, g2 = .80): the more the letter was rotated, the more
negative the potential. However, when paired contrasts were performed in order to determine whether there were specific differences between the different orientations the results were as
follows: while there were no differences between orientations for
mirrored letters (all adjusted p-values > .05), differences were
found between orientations for normal letters, specifically, between the orientations 0–120, 0–150, 30–120, 30–150, 60–120,
60–150, and 90–150 degrees (all adjusted p-values < .05).
The analysis performed at central sites yielded significant effects for orientation (F(5, 60) = 9.87, p < .001, e = .61, g2 = .45) and
Orientation x Laterality (F(20, 240) = 5.07, p = .001, e = .24,
g2 = .30) only for normal letters. There were no significant effects
for mirrored letters. The Orientation Laterality interaction was
analyzed by performing separate ANOVAs for each central electrode and taking orientation as a factor. With normal letters the
orientation effect reached statistical significance at all the central
Author's personal copy
184
M.I. Núñez-Peña, J.A. Aznar-Casanova / Brain and Cognition 69 (2009) 180–187
Fig. 3. (A) Grand-average ERPs (n = 13) elicited by normal letters in each orientation at the P3, Pz, and P4 electrodes. (B) Grand-average ERPs (n = 13) elicited by mirrored
letters in each orientation at the P3, Pz, and P4 electrodes.
Fig. 4. Amplitude mean (in microvolts) in the 400–500 ms window for normal (solid line) and mirrored letters (dotted line) as a function of stimulus orientation at the P3, Pz,
and P4 electrodes.
Fig. 5. (A) Spatial distribution of the orientation effect in normal letters over all electrodes at the scalp surface (the voltage difference between 400 and 500 ms). From left to
right, voltage differences between 30°, 60°, 90°, 120°, 150°, and the 0° normal upright. (B) Spatial distribution of the orientation effect in mirrored letters over all electrodes at
the scalp surface (the voltage difference between 400 and 500 ms). From left to right, voltage differences between 30°, 60°, 90°, 120°, 150°, and the 0° mirrored upright.
electrodes (all p-values < .002), and a linear trend could be fitted for
all of them (all p-values < .004). Again, the voltage became more
negative the more the letter was rotated.
Fig. 6 shows the spatial distribution of the version effect at each
orientation in 50-ms windows from 400 to 700 ms. It can be seen
that at 0°, where no rotation in the picture plane is required at all,
Author's personal copy
M.I. Núñez-Peña, J.A. Aznar-Casanova / Brain and Cognition 69 (2009) 180–187
185
Fig. 6. Spatial distribution of the version effect over all electrodes at the scalp surface. Voltage differences mirrored minus normal letters at 0°, 30°, 60°, 90°, 120°, and 150° in
50-ms windows from 400 to 700 ms post-stimulus.
the voltage is more negative in mirrored than in normal letters in
the 400–450 and the 450–500 ms windows and that this effect is
right lateralized. The version effect is also present at 30°, 60°,
and 90° but in those cases the effect seems to be delayed. This effect is not present at large orientations.
A detailed analysis of the version effect confirmed these observations. First, the ANOVAs in the upright condition showed that the
interaction Version Laterality reached statistical significance in
the 400–450 ms window (F(2, 24) = 7.43, p = .003, g2 = .38) and
the 450–500 ms window (F(2, 24) = 7.67, p = .003, g2 = .39). Tests
of simple effects showed that the amplitude was more negative
in mirrored than in normal letters over the right sites in both windows (F(1, 12) = 6.10, p = .03, g2 = .34 in the 400–450 ms window
and F(1, 12) = 4.64, p = .05, g2 = .28 in the 450–500 ms window).
Second, when the analysis was performed for misoriented stimuli
the results were as follows: (1) the analysis in the 30° orientation
showed that the version effect reached statistical significance in
the 400–450 ms windows over right sites (F(1, 12) = 4.91, p = .04,
g2 = .29) and in the 450–500 ms windows over the middle and
right sites (F(1, 12) = 5.35, p = .039, g2 = .31 and F(1, 12) = 7.27,
p = .019, g2 = .38, respectively): again the amplitude was more negative in mirrored than in normal letters; (2) the analysis in the 60°
orientation showed that mirror-normal difference was delayed
comparing to the upright and the 30° conditions and, moreover,
has a different scalp distribution: amplitude was more negative
for mirrored than for normal letters at parietal sites in the 450–
500 ms window (F(1, 12) = 5.23, p = .013, g2 = .30); (3) the analysis
in the 90° orientation again showed a delay in the mirror-normal
difference over the scalp: the same pattern of differences as previously described was found in the 500–550 ms window at central
and parietal sites (F(1, 12) = 6.1, p = .03, g2 = .34 and F(1,12) =
4.79, p = .04, g2 = .29, respectively); (4) no mirror-normal differences were found for the 120° and 150° orientations in any
window.
4. Discussion
Previous studies have shown that RT is longer for mirrored than
for normal letters in mental rotation tasks (Bajric et al. 1999;
Hamm et al., 2004; Milivojevic et al., 2003). This increase in RT
has been attributed to the fact that only rotation in the picture
plane is involved in misoriented normal letters, whereas rotation
in and out of the picture plane is involved in misoriented mirrored
letters. Hamm et al. (2004) and Alivisatos and Petrides (1997) provided psychophysiological evidence that an extra rotation out of
the plane is involved in mirrored letters in the upright position.
However, to our knowledge, this extra rotation in mirrored letters
has not been studied to date in orientations other than the upright.
The present study aimed (1) to examine whether mental rotation
of normal and mirrored letters differs in a letter discrimination
task, and (2) to study the extent to which these differences can
be explained by the fact that an extra rotation after the plane rotation is involved in parity judgment on mirror rotated letters. Therefore, we focused our attention on (1) the orientation effect in
Author's personal copy
186
M.I. Núñez-Peña, J.A. Aznar-Casanova / Brain and Cognition 69 (2009) 180–187
normal and mirrored letters and (2) the version effect in different
orientations.
First, the orientation effect was studied. Our results replicated
the classical orientation effect both in RT and for the amplitude
of the rotation-related negativity in normal letters (Hamm et al.
2004; Heil & Rolke, 2002): RT increases and the rotation-related
negativity becomes more negative with angular deviation from upright. The rotation-related negativity showed a centro-parietal
scalp distribution without hemispheric asymmetry. These results
suggest that mental rotation in the picture plane is used in misoriented normal letters. Moreover, the typical indifference of normal
letters to small deviations from upright was also found in RT,
where a quadratic function between orientation and RT could be
fitted. For rotations of 60 degrees or less there was no difference
from the upright condition, which suggests that mental rotation
is not necessary to decide on the parity of most of these stimuli.
Concerning mirrored letters, the classical orientation effect was
again found both in RT and for the amplitude of the rotation-related negativity. However, the effect over the rotation-related negativity was not as evident as that found in normal letters. Although
a linear trend between orientation and amplitude could be fitted,
differences between orientations were not found when paired
comparisons were performed. ERP differences between normal
and mirrored letters suggest that the mental rotation process in
the two types of stimuli is different. As suggested by Hamm et al.
(2004), mirrored letters might require the mental rotation in the
picture plane and out of the picture plane in order to fully normalize the stimuli. This fact could explain why the orientation effect is
less evident in mirrored than in normal letters, because ‘‘the flip
rotation ERP effect would cancel the planar rotation ERP effect
for mirror rotated stimuli” (Hamm et al., p. 816).
Second, we performed a detailed analysis of the version effect.
Our analysis of the ERP data comparing mirrored and normal letters in the upright condition suggests that the rotation out of the
picture plane has an impact on the electrophysiological activity.
A negative-going waveform, right lateralized, and with a latency
between 400 and 500 ms was elicited by mirrored letters in the upright condition. This negative waveform was nearly identical to the
well-known rotation-related negativity. Both were similar in latency and polarity but showed a different scalp distribution. The
negativity elicited by mirrored upright letters was right lateralized
whereas the negativity elicited by misoriented normal letters did
not show hemispheric asymmetry, suggesting that the neural response to the two types of rotations is different. These results agree
with those obtained by Alivisatos and Petrides (1997) who reported the role of the right parietal cortex in a mirror-normal discrimination task of upright letters and the role of right and left
parietal cortex in a task that required active rotation in the picture
plane. Our results are also consistent with those of Roberts and Bell
(2003), who suggests that rotation of simple two-dimensional
stimuli, can lead to greater activation of the left parietal area than
of the right parietal area. Although mirrored letters in the present
study were two-dimensional stimuli, they involve a three-dimensional mental rotation because the flipping strategy needs a mental
transformation of the stimuli out of plane. In contrast, normal letters involve a two-dimensional mental rotation because they are
believed to be rotated only in the picture plane. These differences
in the hemispheric lateralization of mirror and normal letters may
be explained in terms of different mental rotation processes.
Whereas the rotation out of the picture plane may need a holistic
mental rotation process, which preferentially engaged the right
hemisphere, the rotation in the picture plane may also need an
analytic, ‘‘piecemeal” mental rotation process, which preferentially
engaged the left hemisphere (Corballis, 1997).
Differences between mirrored and normal letters were also
found at 30°, 60°, and 90°. At 30°, the ERP pattern was similar to
that observed at the upright position: a negative-going waveform,
right lateralized, and with a latency between 400 and 500 ms was
elicited by mirrored letters. This result was predictable because it
is generally agreed that the planar rotation is not needed to make
a mirror-normal decision for small departures from upright. Thus,
mirrored letters at 30° only have to be rotated out of the picture
plane in order to make a mirror-normal decision and, therefore,
differences between 0° and 30° were not expected. As we have previously mentioned, our data confirmed that there was no difference either in reaction time or in the rotation-related negativity
between these two orientations.
In contrast to the findings at 0° and 30°, when the stimuli were
presented at orientations of 60° and 90°, a delay in the mirror-normal differences was observed. At 60° the difference was found in
the 450–500 ms window and at 90° it was found in the 500–
550 ms window. This pattern of results is consistent with the idea
that the rotation out of the picture plane involved in mirrored rotated letters might occur after the plane rotation, because the mirror-normal difference is delayed across the orientations. However,
scalp distribution of the negative-going waveform at 0° and 30°
differed from that at 60° and 90°. Whereas the first was right-lateralized, the second was centro-parietally distributed and without
hemispheric differences.
As for large misorientations, the mirror-normal difference was
not found. There are two possible explanations for this result. First,
large misorientations may place heavy visuo-spatial demands on
normal letters. The reaction time and error rate results support this
interpretation. Response time analysis showed that the advantage
of normal over mirrored letters decreases with an increment in
misorientation; moreover, participants were less accurate in normal than in mirrored letters at large misorientations. A similar pattern of results has been reported by Heil, Bajric, Rösler, and
Hennighausen (1996b), who found no RT differences between mirrored and normal letters at large misorientations. The second
explanation for the absence of mirror-normal ERP difference for
large misorientations is that the mental rotation in and out of the
picture plane for mirrored letters may occur in parallel. If both
mental rotation processes were assumed to occur sequentially, differences between mirrored and normal letters should be found
even at large misorientations.
In summary, two main conclusions can be drawn from the present study. First, we found evidence that the processing of normal
and mirrored letters in a letter discrimination task has a different
impact on brain activity. The presence of a mental rotation out of
the plane might account for this difference, because this extra rotation might cancel the plane rotation in mirrored letters. Future
mental rotation research should take this difference into account
and study mirror and normal letters separately. Second, our data
provide evidence that the rotation out of the plane in rotated mirrored letters may occur after the plane rotation, because mirrored
letters elicit a negative-going component whose amplitude was
delayed across orientations.
References
Alivisatos, B., & Petrides, M. (1997). Functional activation of the human brain during
mental rotation. Neuropsychologia, 35, 111–118.
Bajric, J., Rösler, F., Heil, M., & Hennighaugen, E. (1999). On separating processes of
event categorization, task preparation, and mental rotation proper in a
handedness recognition task. Psychophysiology, 36, 399–408.
Cohen, M. S., Kosslyn, S. M., Breiter, H. C., DiGirolamo, G. J., Thompson, W. L.,
Anderson, A. K., et al. (1996). Changes in cortical activity during mental rotation.
A mapping study using functional MRI. Brain, 119, 89–100.
Cooper, L. A., & Shepard, R. N. (1973). Chronometric studies of the rotation of mental
images. In W. G. Chase (Ed.), Visual information processing (pp. 75–176). New
York: Academic Press.
Cooper, L. A., & Shepard, R. N. (1975). Mental transformations in the identification of
left and right hands. Journal of Experimental Psychology: Human, Perception and
Performance, 104(1), 48–56.
Author's personal copy
M.I. Núñez-Peña, J.A. Aznar-Casanova / Brain and Cognition 69 (2009) 180–187
Corballis, M. C. (1997). Mental rotation and the right hemisphere. Brain and
Language, 57, 100–121.
Hamm, J. P., Johnson, B. W., & Corballis, M. C. (2004). One good turn deserves
another: An event-related brain potential study of rotated mirror-normal letter
discriminations. Neuropsychologia, 42, 810–820.
Harris, I. M., Egan, G. F., Sonkkila, C., Tochon-Danguy, M. J., Paxinos, G., & Watson, J.
D. G. (2000). Selective right parietal lobe activation during mental rotation: A
parametric PET study. Brain, 123, 65–73.
Heil, M. (2002). The functional significance of ERP effects during mental rotation.
Psychophysiology, 39, 535–545.
Heil, M., Bajric, J., Rösler, F., & Hennighausen, E. (1996a). Event-related potentials
during mental rotation: Disentangling the contributions of character
classification and image transformation. Journal of Psychophysiology, 10, 326–335.
Heil, M., Bajric, J., Rösler, F., & Hennighausen, E. (1996b). A rotation aftereffect
changes both the speed and the preferred direction of mental rotation. Journal of
Experimental Psychology: Human Perception and Performance, 23(3), 681–692.
Heil, M., Rauch, M., & Hennighausen, E. (1998). Response preparation begins before
mental rotation is finished: Evidence from event-related brain potentials. Acta
Psychologica, 99, 217–232.
Heil, M., & Rolke, B. (2002). Towards a chronopsychophysiology of mental rotation.
Psychophysiology, 39, 414–422.
Jolicoeur, P. (1985). The time to name disoriented natural objects. Memory &
Cognition, 13, 289–303.
Jolicoeur, P. (1988). Mental rotation and the identification of disoriented objects.
Canadian Journal of Psychology, 42, 461–478.
Jolicoeur, P. (1990). Identification of disoriented objects: A dual-system theory.
Mind & Language, 5, 387–410.
Jordan, K., Heinze, H. J., Lutz, K., Kanowski, M., & Jancke, L. (2001). Cortical
activations during the mental rotation of different visual objects. Neuroimage,
13, 143–152.
Keselman, H. J. (1998). Testing treatment effects in repeated measures designs: An
update for psychophysiological researchers. Psychophysiology, 35, 470–478.
Kirk, R. E. (1996). Practical significance: A concept whose time has come.
Educational Psychology Measurement, 56, 746–759.
Koriat, A., & Norman, J. (1985a). Reading rotated words. Journal of Experimental
Psychology: Human Perception and Performance, 11, 490–508.
187
Koriat, A., & Norman, J. (1985b). Mental rotation and visual familiarity. Perception &
Psychophysics, 37, 429–439.
Milivojevic, B., Johnson, B. W., Hamm, J. P., & Corballis, M. C. (2003). Non-identical
neural mechanisms for two types of mental transformation: Event-related
potentials during mental rotation and mental paper folding. Neuropsychologia,
41, 1345–1356.
Muthukumaraswamy, S. D., Johnson, B. W., & Hamm, J. P. (2003). A high density ERP
comparison of mental rotation and mental size transformation. Brain and
Cognition, 52, 271–280.
Núñez-Peña, M. I., Aznar, J., Linares, D., Corral, M. J., & Escera, C. (2005). Effects of
dynamic rotation on event-related brain potentials. Cognitive Brain Research, 24,
307–316.
Peronnet, F., & Farah, M. J. (1989). Mental rotation: an event-related potential study
with a validated mental rotation task. Brain and Cognition, 9, 279–288.
Richter, W., Ugurbil, K., Georgopoulos, A., & Kim, S. G. (1997). Time-resolved fMRI of
mental rotation. Neuroreport, 89, 3697–3702.
Roberts, J. E., & Bell, M. A. (2003). Two- and three-dimensional mental rotation tasks
lead to different parietal laterality for men and women. International Journal of
Psychophysiology, 50, 235–246.
Rösler, F., Heil, M., Bajric, J., Pauls, A. C., & Hennighausen, E. (1995). Patterns of
cerebral activation while mental images are rotated and changed in size.
Psychophysiology, 32, 135–149.
Shepard, R. N., & Metzler, J. (1971). Mental rotation of three-dimensional objects.
Science, 191, 701–703.
Stuss, D. T., Sarazin, F. F., Leech, E. E., & Picton, T. W. (1983). Event-related potentials
during naming and mental rotation. Electroencephalography and Clinical
Neurophysiology, 56, 133–146.
Tarr, M. J., & Pinker, S. (1989). Mental rotation and orientation-dependence in shape
recognition. Cognitive Psychology, 21, 233–282.
Thayer, Z. C., & Johnson, B. W. (2006). Cerebral processes during visuo-motor
imagery of hands. Psychophysiology, 43, 401–412.
Wijers, A. A., Otten, L. J., Feenstra, S., Mulder, G., & Mulder, L. J. M. (1989). Brain
potentials during selective attention, memory search, and mental rotation.
Psychophysiology, 26, 452–467.
Yoshino, A., Inoue, M., & Suzuki, A. (2000). A topographical electrophysiological
study of mental rotation. Cognitive Brain Research, 9, 121–124.
`