This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

Research report

How time modulates spatial responses

Antonino Vallesi a,b,*, Anthony R. McIntosh b,c and Donald T. Stuss b,c

a International School for Advanced Studies, SISSA, Trieste, Italyb Rotman Research Institute at Baycrest, Toronto, Canadac University of Toronto, Canada

a r t i c l e i n f o

Article history:

Received 23 July 2009

Reviewed 2 September 2009

Revised 8 September 2009

Accepted 14 September 2009

Action editor Paolo Bartolomeo

Published online 19 September 2009

Keywords:

Stearc effect

Stimulus–response

compatibility effect

Event-related potentials

Spatial attention

Time processing

Motor preparation

a b s t r a c t

Behavioural evidence suggests a left-to-right directionality in the representation of

elapsing time. We tested whether this representation produces a spatial attentional shift

that activates a corresponding left-to-right spatial response code. Fourteen participants

judged whether a cross lasted for a short (1 sec) or a long (2 sec) duration with left and right

responses, respectively, or vice versa, while event-related potentials (ERPs) were measured.

Responses were faster when participants judged short and long durations with their left

and right hand, respectively, than vice versa. In these compatible conditions only (short-

left; long-right), ERP negativity developed over the right motor scalp region around the

short duration, a finding that is compatible with an early pre-activation of left-hand

responses, and over the left motor region around the long duration, suggesting a later pre-

activation of right hand responses. These findings confirm that in this task elapsing time is

represented from left to right, and that this representation generates corresponding

response codes that influence performance.

ª 2009 Elsevier Srl. All rights reserved.

It is still debated whether we perceive the passage of time

through a dedicated neural system or through intrinsic repre-

sentations arising from other systems (Ivry and Schlerf, 2008).

Time is a very abstract and yet ubiquitous concept. When

dealing with abstract concepts, our cognitive system benefits

from the use of concrete motor and perceptual representations

(Gallese and Lakoff, 2005; Lakoff and Johnson, 1999; Rohrer,

2006; Scaife and Rogers, 1996; Tversky, 1995). Spatial repre-

sentations in particular are an effective way to convey cogni-

tive meaning (Lakoff and Johnson, 1980; Zacks and Tversky,

1999). For instance, spatial graphical representations improve

performance in syllogistic reasoning (Stenning and

Oberlander, 1995), analytical problem solving (Cox and Brna,

1995), and working memory tasks (Larkin and Simon, 1987).

Cognitive representations of ordinal information, such

as numerals, letters, and months, may also have spatial

features. These spatial features can influence the speed of

manual responses (Dehaene et al., 1993; Fischer et al., 2003;

Franklin et al., 2009; Gevers et al., 2003, 2004). With respect

to numerals, for instance, left-side responses are faster

for relatively smaller numbers compared to right-side

responses, whereas right-side responses are faster with

* Corresponding author. Cognitive Neuroscience Sector, International School for Advanced Studies (SISSA-ISAS), Via Beirut 2-4, 34014Trieste, Italy.

E-mail address: vallesi@sissa.it (A. Vallesi).

ava i lab le a t www.sc iencedi rec t .com

journa l homepage : www.e lsev ie r . com/ loca te /cor tex

0010-9452/$ – see front matter ª 2009 Elsevier Srl. All rights reserved.doi:10.1016/j.cortex.2009.09.005

c o r t e x 4 7 ( 2 0 1 1 ) 1 4 8 – 1 5 6

Author's personal copy

relatively larger numbers (Dehaene et al., 1993). This suggests

the existence of a left-to-right ‘mental number line’, which

interacts with motor preparation in the horizontal space. This

left-to-right representation of numbers is probably related to

the writing direction (Dehaene et al., 1993; Zebian, 2005).

Analogously, the representation of time can also be influ-

enced by processing non-temporal perceptual features. Subjects

judge the duration of an array of dotsas longer if it contains more,

brighter or bigger dots (Xuan et al., 2007). Moreover, subjects are

more accurate in judging short and long durations of a low and

high value digit, respectively, than the other way around (Xuan

et al., 2009). These interactions suggest that the mechanisms for

the representation of perceptual magnitude information are also

used in the representation of the more abstract magnitude

information concerning time durations.

Moreover, as for numbers, spatial attention and spatial

representations can also influence time processing (e.g.,

Santangelo and Spence, 2009; Santiago et al., 2007; Torralbo

et al., 2006; see Oliveri et al., 2009, for a review), while the

opposite is not necessarily true (Casasanto and Boroditsky,

2008). In a recent study (Weger and Pratt, 2008; see also Weger

and Pratt, 2009), the task was to indicate the location of the

target through a bimanual response, or to detect (with one

button only) left- or right-side stimuli following prospective or

retrospective time words. Only when a bimanual response

choice was required with the spatial discrimination task,

compatibility effects between the dimensions of space (left–

right) and time (earlier–later) were found, with performance

being better when left spatial position was associated with

time words referring to the past and right spatial position was

associated with time words referring to the future than vice

versa. These findings suggest that the compatibility effect

between temporal and spatial dimensions affects more

response-related than stimulus-related processes (cf. Stoia-

nov et al., 2008). However, results of a recent study show that

the discrimination of lateralized stimuli (leftward or right-

ward pointing arrows) can also benefit from past-future

temporal cue words independently of the specific motor

responses required (Ouellet et al., 2010, Experiment 2). These

data suggest a more central mechanism of spatial attention

orienting driven by the processing of temporal information,

which can flexibly boost the processing of spatially compat-

ible stimuli or/and responses, depending on the specific

experimental settings and task demands.

Compatibility effects between time and space emerge even

when subjects have to judge the duration of stimuli that do

not have any lateralized spatial feature. In a previous study of

ours (Vallesi et al., 2008), the task was to judge the duration of

a centrally presented cross (1 and 3 sec, for short and long

durations, respectively) using a bimanual spatial response

(left vs right). The duration-hand correspondence was coun-

terbalanced across blocks. Participants were faster to judge

short durations with their left hand and long durations with

their right hand than vice versa. Critically, no spatial infor-

mation was presented along with the stimuli, since the

temporal interval was operationalized as the duration of

a cross, which appeared foveally. The same results were

obtained by using two hands or two fingers of the same hand

to respond. These results suggest that elapsing time can be

mentally represented in terms of spatial coordinates

developing from left to right and that this representation

affects left–right responses (see Ishihara et al., 2008, for an

analogous effect in the auditory domain).

This effect, known as Spatio-Temporal Association

between Response Codes (Stearc) effect (Ishihara et al., 2008;

Vallesi et al., 2008), may be analogous to those observed in

other spatial compatibility tasks where stimulus spatial

position is real and not imagined. For instance, in the Simon

task (Simon and Small, 1969), although subjects have to judge

the stimuli on a non-spatial dimension and stimulus location

is task-irrelevant, it still affects performance: responses are

faster and more accurate if the stimulus position corresponds

to the response position than vice versa (see Hommel and

Prinz, 1997; Umilta and Nicoletti, 1990, for reviews).

These effects have been observed not only with static

spatial stimuli but also with moving ones (Proctor et al., 1993). A

critical factor in the Simon effect seems to be the attentional

shift towards the stimulus spatial location, which in turn

generates a response code on the same side of the attentional

shift (e.g., Umilta and Nicoletti, 1992; Vallesi et al., 2005). It is

possible that a similar spatial attentional mechanism produces

the Stearc effect and related time–space compatibility

phenomena, as suggested by recent evidence (Ouellet et al., in

press). Specifically, we hypothesize that the mental represen-

tation of elapsing time develops from left to right, orienting

spatial attention and priming compatible responses accord-

ingly. If this is the case, it should be possible to continuously

detect pre-activations of the compatible responses (i.e., left

responses for short durations, and right responses for long

durations) during the response preparation stage.

The aim of this study was to test this hypothesis. So far, only

behavioural studies of left–right spatial–temporal compatibility

effects have been carried out, suggesting the testable hypothesis

that the locus of this effect involves spatial attention (Ouellet

et al., in press) and can have consequences at the response pro-

cessing stages (Weger and Pratt, 2008). However, to fully under-

stand the nature of this family of effects, it is necessary to use

a technique, such as event-related potentials (ERPs), that can

continuously track cognitive processes, such as response prep-

aration, as they unfold in time, unlike response times (RTs) that

only mark the final stage of the processing flow (i.e., response

execution). To this aim, we recorded ERPs while participants

performed a similar time judgment task as in Vallesi et al. (2008).

It is known that preparing a hand movement produces

relatively more negativity over the hand-motor area contra-

lateral to the response side (Brunia and Van Boxtel, 2000;

Toma et al., 2002), as it can be appreciated by subtracting

electrical activity in the ipsilateral scalp motor region from

that in the contralateral one (Shibasaki et al., 1981). Given the

hypothesis that a temporal timeline evolving from left to

right, if it exists, would produce a response code that also

continuously evolves from left to right, we focused our anal-

yses to the two scalp electrodes that pick up most of the

electrical activity from the hand-motor cortices of both

hemispheres (i.e., C3 and C4). By subtracting activity in one of

those electrodes from that in the other, we obtained

a continuous measure of which response side is more acti-

vated at a given time-point during a trial.

It is worth mentioning that, in the electrophysiological

literature, a common measure of lateralized motor

c o r t e x 4 7 ( 2 0 1 1 ) 1 4 8 – 1 5 6 149

Author's personal copy

preparation is the Lateralized Readiness Potential (LRP; De

Jong et al., 1988; Gratton et al., 1988), which implies the further

step of averaging the hemispheric differences for left and right

responses in order to minimize the effect of possible lateral-

ized hemispheric processes that do not co-vary with motor

responses per se. However, this measure by definition does

not allow to continuously tracking which of the two motor

cortices (left vs right) is more activated at each time-point.

Therefore, we could not compute the full LRP here, since this

measure would not be suitable to test our hypothesis that

partial response activation should evolve along a specific

direction, namely from left to right, and not vice versa.

1. Method

1.1. Participants

Fourteen young healthy volunteers (7 females; mean age: 26

years, range: 19–34) took part in the study after providing

informed consent. The participants had normal or corrected-to-

normal vision and no history of neurological or psychiatric

disorders. All were right-handed, with the average Edinburgh

Handedness Inventory score (Oldfield, 1971) being 84 (range: 40–

100). They received 10 dollars per hour for participating in the

experimental session, which lasted 1–2 h, including Electroen-

cephalographic (EEG) cap preparation and removal. The study

was previously approved by the Baycrest Research Ethic Board.

1.2. Material and task

Participants were tested individually in a sound-attenuated

semi-dark room. A 64-channel cap was mounted at the

beginning of the session (see ERP procedure). Visual stimuli

were presented through a computer display at a distance of

~60 cm. A trial started with a central fixation cross (2 yellow

crossed bars, 1.0� .5 cm) lasting 1 or 2 sec (50% each). A

downward pointing white arrow (a 1.5� 1 cm bar attached to

a .5 cm arrowhead with a maximum width of 2 cm) was then

presented as a response-signal.

When judging two temporal intervals that are very

different from each other with two responses, participants

might prepare their responses in a bimodal fashion, for

instance activating a response side close to the end of the

short duration and then the opposite response side only when

they get close to the end of the long duration. As an attempt to

discourage this strategy, and therefore to stress the contin-

uous nature of elapsing time, we used two durations that were

closer than in the previous experiments (1 and 2 sec here vs 1

and 3 sec in Vallesi et al., 2008; but see footnote 1).

The task consisted of pressing the left key ‘C’ of the

computer keyboard (with the left index finger) for a short cross

duration (1 sec), and the right key ‘M’ (with the right index

finger) for a long cross duration (2 sec), respectively. The

response key assignment to each cross duration was inverted

for half of the four runs. The order of presentation of each run

with each of the 2 possible Stimulus duration–Response posi-

tion (S–R) mappings was randomized across participants. A

familiarization run, consisting of 10 trials, was presented

before each experimental run, which consisted of 80 trials

each. The total number of test trials was 320. On each trial,

a cross was presented for 1 or 2 sec. After this variable interval

the response-signal (arrow) replaced the cross for 200 msec,

followed by a blank screen for another 1800 msec. Therefore,

the response deadline was 2 sec (from the arrow onset to the

end of the trial). A short rest was provided after each run.

Participants were discouraged from using any strategy (e.g.,

counting) in judging the cross duration, and asked to minimize

eye blinks, eye movements or contractions of facial muscles.

1.3. Behavioral data analysis

Practice trials, the first trial of each test run and trials with

responses beyond 100–1500 msec after the arrow onset were

discarded from further analyses. RT analyses only included

trials with correct responses. Both RTs and accuracy data were

submitted to a 2� 2 repeated-measures ANOVA with cross

duration (1 vs 2 sec) and response side (left vs right hand) as

the repeated-measures factors.

1.4. Electrophysiological recording and pre-processing

Scalp voltages were recorded using ElectroCaps with 64 Ag/AgCl

electrodes (10/20 system), which included electrodes on the

external canthus and infra-orbital ridge. The EEG recording

system was NeuroScan SynAmps (El Paso, TX). The online

reference electrode was the vertex, with the mid-frontal elec-

trode AFz used as the ground. Electrode impedances were

maintained below 5 kU. The electrical signals were filtered with

a .1–100 Hz bandwidth filter and digitized with a 250 Hz

sampling rate.

The ERP data were pre-processed with Brain Electrical

Source Analysis (BESA 5.2; Germany). Data were re-referenced

to an average reference. Eye artifacts (i.e., eye blinks, lateral

and vertical movements) were compensated from the ERP

waveforms using source components derived from the

recordings obtained before and after the performance of the

task (Picton et al., 2000). Stimulus-locked ERP data from

correct trials were averaged as a function of the 4 conditions

obtained by crossing 2 durations (1 vs 2 sec) by 2 response

hand (right vs left) and digitally filtered (.1–10 Hz). ERP epochs

for short durations were averaged over a 2700 msec period

beginning 200 msec before the cross onset and ending

1500 msec after arrow onset. ERP epochs for the long dura-

tions were averaged over a 3700 msec period beginning

200 msec before the cross onset and ending 2500 msec after

the arrow onset. All ERPs were baseline-corrected according to

the 200 msec preceding the cross onset. The final ERPs (aver-

aged across runs) were based on a total of 57–79 trials per

condition (average: 72).

1 In the present task, we asked participants to judge twodiscrete durations (1 and 2 sec). A task requiring a comparisonbetween a standard duration and different target durations,perhaps, would have been a better way to encourage participantsto monitor elapsing time continuously. However, that time wasrepresented continuously and sequentially from left to right alsoin the present task was suggested by the measures provided bythe C3–C4 ERP difference, which support the prediction that thisrepresentation produces a continuous spatial attentional shiftfrom left to right, which in turn primes left–right responses.

c o r t e x 4 7 ( 2 0 1 1 ) 1 4 8 – 1 5 6150

Author's personal copy

1.5. C3–C4 differential waveform analysis

To calculate the differential activation of the scalp regions

over the left versus right hand-motor cortex, we subtracted

amplitude in the right electrode C4 (over the left-hand-motor

area) from amplitude in the left electrode C3 (over the right

hand-motor area) for each of the 4 conditions. To evaluate

when one of the two response sides was consistently more

activated than the other, a bootstrap procedure was imple-

mented (1000 bootstrapped samples) to calculate 95% confi-

dence intervals for each time point from 100 msec after the

cross onset to 1000 msec after the arrow onset. When both

inferior and superior confidence intervals were above the zero

line, a left-more-than-right motor activation was inferred. On

the other hand, when both inferior and superior confidence

intervals were below the zero line, a right-more-than-left

motor activation was inferred.

To test whether the phenomenon reflected an effect of

spatial attention on the response activation and not other

lateralized processes unrelated to the response, the response-

locked differential waveform was also extracted. By triggering

the ERP averaging on the response rather than on the stim-

ulus, response-related processes are further enhanced while

stimulus-locked processes will be much less evident (cf.

Osman et al., 1995). This measure was calculated from

600 msec before to 400 msec after the RT. Response-locked

waveforms were baseline-corrected by subtracting the

average voltage during the 600–400 msec before RTs from the

whole segment (cf. Osman et al., 1995).

2. Results

2.1. Behavioral results

Trials with RTs shorter than 100 msec (.6%) and longer than

1500 msec (.6%), and the first trial of each run (1.25%) were

discarded from the analyses. The error percentage was almost

equally distributed across conditions (4.5–5.2%) and no effect

was significant in the ANOVA concerning accuracy. Fig. 1

represents mean RTs according to the 4 experimental condi-

tions. Responses were overall faster for long durations as

indicated by the duration main effect [F(1,13)¼ 16.3, p< .01],

and the critical duration by response side interaction was also

significant [F(1,13)¼ 11.3, p< .01]. This interaction demon-

strates that participants were faster in judging short durations

with their left hand and long durations with their right hand

than vice versa (i.e., the classical Stearc effect). More detailed

planned comparisons showed that the RT difference between

right and left responses was significant for the long duration

(40 msec, p¼ 01) and was a strong tendency for the short

duration (�26 msec, p¼ .07).

2.2. Stimulus-locked C3–C4 differential waveformanalysis

Fig. 2 shows the activity in the critical electrodes C3 and C4

according to the 4 different conditions. Fig. 3 shows the results

of the bootstrap procedure used to evaluate the differential

activation of left versus right motor scalp regions for the

stimulus-locked ERPs. The circles above each differential

waveform indicate time-points that were reliably positive (left

motor activation), whereas the circles below mark time-points

that were reliably negative (right motor activation). The

intervals in which sample-points were reliably different from

0 are also reported in the Table 1.

A left-hand response was pre-activated for short durations

soon after the cross onset in all conditions (range from 160 to

510 msec post-cross onset). Moreover, in the compatible

conditions (short duration: left response; long duration: right

response), there was an increase of negativity in the scalp

motor site contralateral (vs ipsilateral) to the left response

around the end of a short cross duration (from 824 msec post-

cross onset), which then slowly shifted towards the opposite

side if time passed by towards the long duration. The mirror

phenomenon was not observed during the incompatible long

duration condition; that is, a right motor response was not pre-

activated around the possible occurrence of the stimulus for

short durations. On the contrary, when a short duration was

associated with a right response (incompatible condition),

after the arrow onset there was a brief pre-activation of the

wrong (left) response side (56–216 msec) that was then inverted

towards the correct (right) response side (most time-points

from 348 to 996 msec). Finally, there was a rebound effect with

negativity developing in the opposite (left) hemisphere for left

responses only (both with short and with long durations).

2.3. Response-locked C3–C4 differential waveformanalysis

Fig. 4 and Table 2 show the results of the bootstrap procedure

used to evaluate the differential activation of left versus right

motor scalp regions (C3 vs C4) for the response-locked ERPs.

Like in Fig. 3, the circles above each differential waveform

indicate time-points that were reliably positive (left motor

activation), whereas the circles below mark time-points that

were reliably negative (right motor activation).

Fig. 1 – Mean response times (and standard errors of the

mean) as a function of cross duration (x-axis) and

responding hand (histograms).

c o r t e x 4 7 ( 2 0 1 1 ) 1 4 8 – 1 5 6 151

Author's personal copy

The task-relevant response side was activated earlier in

the two compatible conditions than in the two incompatible

ones. In particular, when a short cross duration should be

responded to with the left hand (compatible conditions), the

differential waveform becomes reliably positive (suggesting

a right motor cortex pre-activation: from 508 msec before to

40 msec after the RT) much earlier than it becomes reliably

negative when the same short duration should be responded

to with the right hand (suggesting a left motor cortex pre-

activation: from 164 msec before to 64 msec after the RT).

Fig. 2 – Grand-average amplitudes of the stimulus-locked ERPs in the two electrodes located on hand-motor scalp regions of

the left and right hemisphere: C3 (continuous line) and C4 (dashed line), respectively. Panels (A) and (B) show waveforms

associated to the short duration to be responded to with left and right hand, respectively (as shown by the drawing of the

two hands). Panels (C) and (D) show waveforms associated to the long duration to be responded to with right and left hand,

respectively. The two conditions with compatible spatio-temporal mapping (A and C) are surrounded by a bold rectangle,

while the two conditions with incompatible spatio-temporal mapping (B and D) are surrounded by a thin rectangle. The two

critical stimuli, that is the cross (whose duration had to be judged) and the arrow (i.e., the response-signal), are shown on

the top two panels and their onset is marked on each panel with a continuous and a dashed vertical line, respectively.

Fig. 3 – Grand-average amplitude of the stimulus-locked differential waveforms between electrodes C3 and C4 (black lines)

and inferior and superior 95% confidence intervals (grey lines). Panels (A) and (B) show waveforms associated to the short

duration to be responded to with left and right hand, respectively (as shown by the drawing of the two hands). Panels (C)

and (D) show waveforms associated to the long duration to be responded to with right and left hand, respectively. Black

circles at the top of each panel denote time-points where both inferior and superior confidence intervals are above the 0 line,

indicating that activity in C4 is reliably more negative than that in C3. Grey circles at the bottom of each panel mark time-

points in which both inferior and superior confidence intervals are below the 0 line, indicating that activity in C3 is reliably

more negative than that in C4. See Fig. 2 for an explanation of the other symbols.

c o r t e x 4 7 ( 2 0 1 1 ) 1 4 8 – 1 5 6152

Author's personal copy

Moreover, when a long cross duration should be responded to

with the right hand (compatible condition), the differential

waveform becomes reliably negative (suggesting a left motor

cortex pre-activation: most of the time-points from 480 msec

before to 52 msec after the RT) earlier than it becomes positive

when the same long duration should be responded to with the

left hand (suggesting a right motor cortex pre-activation: most

time-points from 360 msec before to 16 msec after the RT).

Finally, for all the 4 conditions, the differential wave is reliably

negative for a variable interval after RT, suggesting a post-

response activation of the left motor cortex (right hand).

3. Discussion

The present results show that when a horizontal bimanual

response is given according to the temporal duration of

a stimulus, the response side affects performance. Specifi-

cally, participants are faster if they respond ‘short’ with the

left hand and ‘long’ with the right hand one than vice versa.

This pattern of behavioral results suggests that elapsing time

is internally mapped onto left-to-right spatial coordinates and

replicate previous findings (Ishihara et al., 2008; Vallesi et al.,

2008).

The main effect of duration indicates the presence of the

variable foreperiod effect, that is shorter RTs for longer

durations (Niemi and Naatanen, 1981; Vallesi and Shallice,

2007; Vallesi et al., 2007; Woodrow, 1914). However, the

duration by hand interaction cannot be explained in terms of

a greater foreperiod effect with either hand, because when

only a single hand is used in a block, no interaction between

duration and response side is observed (Vallesi et al., 2008,

Experiment 2).

The ERP data provide further insights towards the under-

standing of this phenomenon. Stimulus-locked ERPs, and in

particular the differential C3–C4 waveforms, show that a left-

hand response is pre-activated soon after the cross onset in all

conditions (Fig. 3). Moreover, when a right response is

Table 1 – Intervals (in msec, with respect to the cross onset) in which the differential ERP waveform (C3–C4) was reliablymore positive or more negative than 0 for each task condition.

Cross duration Short Long

Response side Left Right Left Right

Polarity Negative Positive Negative Positive Negative Positive Negative Positive

1712–1960 172–212,

292–368,

824–1532

1348–1612,

1676–1796,

1876–1912,

1948–1996

204–220,

492–510,

932–988,

1056–1216

2808–3316 172–216,

304–376,

2236–2252

1848–1880,

2100–2224,

2248–2360,

2404–2528

160–272,

312–456,

824–844,

904–952,

1076–1152,

1240–1320

Fig. 4 – Grand-average amplitude of the response-locked differential waveforms between electrodes C3 and C4 (black lines)

and inferior and superior 95% confidence intervals (grey lines). See Fig. 3 for an explanation of the symbols used.

c o r t e x 4 7 ( 2 0 1 1 ) 1 4 8 – 1 5 6 153

Author's personal copy

assigned to short durations (incompatible condition), there

is a brief pre-activation of the wrong (left) response side that is

then corrected towards the correct (right) response side after

the go-stimulus onset (Fig. 3, panel B), an effect that resembles

the inappropriate response pre-activation (Gratton dip) in the

LRP for the incompatible conditions of the horizontal Simon

effect (Gratton et al., 1988; Vallesi et al., 2005). Finally, when

subjects have to respond to short durations with the left hand

and to long durations with the right one (compatible condi-

tions), there is an increase of activation of the scalp motor site

contralateral (vs ipsilateral) to the left response near the end

of a short duration, which then gradually shifts towards the

opposite side if the duration becomes longer (Fig. 3, panel C).

The mirror phenomenon is not observed during the

incompatible condition in which a long duration should be

responded to with a left response: a right motor response is not

pre-activated around the possible occurrence of the go-stim-

ulus for short durations (Fig. 3, panel D). Hence, in this condi-

tion the relevant right response is not pre-activated around the

short duration, but it is activated only in a stimulus-driven

fashion if the response-signal (arrow) appears at the end of the

short duration (1 sec), and only after a brief pre-activation of

the wrong but compatible left response (Fig. 3, panel B).

Converging and complementary information can be gained

from the response-locked differential waveforms. The task-

relevant response side was activated earlier in the two

compatible conditions than in the two incompatible ones. In

particular, with a short cross duration, the moment in which

the C3–C4 differential waveform becomes reliably positive

(suggesting a right motor cortex pre-activation) for a left

response is earlier than the moment in which the differential

waveform becomes reliably negative (suggesting a left motor

cortex pre-activation) for a right response. On the other hand,

with a long cross duration, the moment in which the differ-

ential waveform becomes reliably negative (suggesting a left

motor cortex pre-activation) for a right response is earlier

than the moment in which the differential waveform becomes

reliably positive (suggesting a right motor cortex pre-activa-

tion) for a left response.1

Finally, a rebound effect (more negativity) occurred in the

opposite hemisphere, which in the stimulus-locked differen-

tial ERPs was significant only for left responses (both with

short and long durations). A somewhat similar phenomenon

could be observed in the response-locked differential ERPs,

since C3 (i.e., the electrode picking up most of the right hand

motor activation) was more negative than C4 for some time

after the RT, independently of the condition. This pattern

could be related to the fact that, in right-handed subjects, the

left motor cortex keeps more in check the right homologous

area through inhibitory control than the other way around

(Ziemann and Hallett, 2001). However, whatever are the

physiological mechanisms underlying this hemispheric

asymmetry, this effect is not relevant for the main hypotheses

under study here, since it occurs after the RT and indepen-

dently of the task condition.

Handedness could be a factor in determining the left–right

direction of the Stearc effect. No study up to now has tested,

together with right-handed subjects, a complementary

sample of left-handed subjects. However, there is no apparent

reason why right-handed subjects should show a left-hand

advantage for short intervals.

Changing the experimental settings, such as the spatial

presentation (back–front vs left–right) of time words to be

judged as belonging to the past or to the future, has been

shown to flexibly change the nature of the association

between time codes (past-future) and the spatial codes arising

from the experimental context (Torralbo et al., 2006). Simi-

larly, the experimental settings adopted in the present and

related previous studies, such as the fact that the required

responses to judge short and long time durations were located

along the left–right horizontal plan, are likely to play a role.

This however does not entirely explain why left and right

responses are associated with short and long durations, respec-

tively, and not the other way around. The direction of writing is

a likely factor in determining the direction of the association

(Santiago et al., 2007; Tversky et al., 1991; Zwaan, 1965). Direc-

tionality of writing affects also other spatial compatibility effects,

such as the SNARC effect (Dehaene et al., 1993; Zebian, 2005). The

prolonged exposure to timelines that run from left to right, such

as in the x-axis of the Cartesian coordinate system, may also play

a role. A follow up study is desirable to test whether the

phenomenon disappears or is reversed in people adopting

writing systems with a directionality different from the left-to-

right one (e.g., Mandarin, Hebrew, Arabic), provided that they

have not been also exposed to left-to-right writing systems or

time representations.

In summary, the present findings further confirm the

hypothesis that, when responses should be given in the

horizontal dimension, elapsing time is also represented as

continuously evolving from left to right. This dynamic spatial

representation of elapsing time produces a response code that

pre-activates the corresponding motor cortex and speeds

up the response when the response mapping is compatible

with the short/left and long/right internal representation of

time. The same is not true for incompatible mappings, which

actually produces a cost in terms of speed due to the brief pre-

activation of the compatible but task-irrelevant response side

for short durations, and to the absence of pre-activation of the

Table 2 – Intervals (in msec, with respect to the RT) in which the differential ERP waveform (C3–C4) was reliably morepositive or more negative than 0 for each task condition.

Cross duration Short Long

Response side Left Right Left Right

Polarity Negative Positive Negative Positive Negative Positive Negative Positive

196–396 �508 to 40 �164 to 64,

164–396

�420, �352 80–364 �360 to �212,

�184 to 16

�480 to �404,

�352 to 52,

168–212

–

c o r t e x 4 7 ( 2 0 1 1 ) 1 4 8 – 1 5 6154

Author's personal copy

relevant response side for long durations. This study suggests

that time mentally flies from left to right, and this spatial

representation of time spreads, probably mediated by an

attentional shift, to the response preparation stage affecting

motor performance.

Acknowledgements

This research was supported by: postdoctoral fellowship

funding from Canadian Institute of Health Research (CIHR,

MFE-87658) and a research Award from the Jack & Rita Cath-

erall Fund to AV; from Canadian Foundation for Innovation

(CFI) and Ontario Innovation Trust to University of Toronto

Functional Imaging Network (#1226) to DTS. We thank Juan

Lupianez for useful suggestions.

r e f e r e n c e s

Brunia CHM and Van Boxtel GJM. Motor preparation. InCacioppo JT, Tassinary LG, and Berntson GG (Eds), Handbook ofPsychophysiology. Cambridge, UK: Cambridge University Press,2000: 507–532.

Casasanto D and Boroditsky L. Time in the mind: Using space tothink about time. Cognition, 106: 579–593, 2008.

Cox R and Brna P. Supporting the use of external representationsin problem solving: The need for flexible learningenvironments. Journal of Artificial Intelligence in Education, 6:239–302, 1995.

Dehaene S, Bossini S, and Giraux P. The mental representationof parity and number magnitude. Journal of ExperimentalPsychology: General, 122: 371–396, 1993.

De Jong R, Wierda M, Mulder G, and Mulder LJM. Use of partialinformation in responding. Journal of Experimental Psychology:Human Perception and Performance, 14: 682–692, 1988.

Fischer MH, Castel AD, Dodd MD, and Pratt J. Perceiving numberscauses spatial shifts of attention. Nature Neuroscience, 6:555–556, 2003.

Franklin MS, Jonides J, and Smith EE. Processing of orderinformation for numbers and months. Memory & Cognition,37: 644–654, 2009.

Gallese V and Lakoff G. The brain’s concepts: The role of thesensory-motor system in conceptual knowledge. CognitiveNeuropsychology, 22: 455–479, 2005.

Gevers W, Reynvoet B, and Fias W. The mental representationof ordinal sequences is spatially organized. Cognition, 87:B87–B95, 2003.

Gevers W, Reynvoet B, and Fias W. The mental representation ofordinal sequences is spatially organized: Evidence from daysof the week. Cortex, 40: 171–172, 2004.

Gratton G, Coles MGH, Sirevaag EJ, Eriksen CW, and Donchin E.Pre- and post-stimulus activation of response channels: Apsychophysiological analysis. Journal of ExperimentalPsychology: Human Perception and Performance, 14: 331–344, 1988.

Hommel B and Prinz W. Theoretical Issues in Stimulus–ResponseCompatibility. Amsterdam: North-Holland, 1997.

Ishihara M, Keller PE, Rossetti Y, and Prinz W. Horizontal spatialrepresentations of time: Evidence for the STEARC effect.Cortex, 44: 454–461, 2008.

Ivry RB and Schlerf JE. Dedicated and intrinsic models of timeperception. Trends in Cognitive Sciences, 12: 273–280, 2008.

Lakoff G and Johnson M. Metaphors We Live By. Chicago: Universityof Chicago Press, 1980.

Lakoff G and Johnson M. Philosophy in the Flesh: The Embodied Mindand its Challenge to Western Thought. Chicago: University ofChicago Press, 1999.

Larkin JH and Simon HA. Why a diagram is (sometimes) worth tenthousand words. Cognitive Science, 11: 65–99, 1987.

Niemi P and Naatanen R. Foreperiod and simple reaction time.Psychological Bulletin, 89: 133–162, 1981.

Oldfield RC. The assessment and analysis of handedness: TheEdinburgh inventory. Neuropsychologia, 9: 97–113, 1971.

Oliveri M, Koch G, and Caltagirone C. Spatial–temporalinteractions in the human brain. Experimental Brain Research,195: 489–497, 2009.

Osman A, Moore CM, and Ulrich R. Bisecting RT with lateralisedreadiness Potentials: Precue effects after LRP onset. ActaPsychologica, 90: 111–127, 1995.

Ouellet M, Santiago J, Funes MJ, and Lupianez J. Thinking aboutthe future moves attention to the right. Journal ofExperimental Psychology: Human Perception and Performance,36(1): 17–24, 2010.

Picton TW, Bentin S, Berg P, Donchin E, Hillyard SA, Johnson R,et al. Guidelines for using human event-related potentials tostudy cognition: Recording standards and publication criteria.Psychophysiology, 37: 127–152, 2000.

Proctor RW, Van Zandt T, Lu CH, and Weeks DJ. Stimulus–response compatability for moving stimuli: Perception ofaffordances or directional coding? Journal of ExperimentalPsychology: Human Perception and Performance, 19: 81–91, 1993.

Rohrer T. The body in space: Dimensions of embodiment. InZlatev J, Ziemke T, Frank R, and Dirven R (Eds), Body, Languageand Mind. Mouton de Gruyter: Berlin, Germany, 2006.

Santangelo V and Spence C. Crossmodal exogenous orientingimproves the accuracy of temporal order judgments.Experimental Brain Research, 194: 577–586, 2009.

Santiago J, Lupianez J, Perez E, and Funes MJ. Time (also) fliesfrom left to right. Psychonomic Bulletin & Review, 14:512–516, 2007.

Scaife M and Rogers Y. External cognition: How do graphicalrepresentations work? International Journal of Human ComputerStudies, 45: 185–213, 1996.

Shibasaki H, Barrett G, Halliday E, and Halliday AM. Corticalpotentials associated with voluntary foot movement in man.Electroencephalography and Clinical Neurophysiology, 52:507–516, 1981.

Simon JR and Small AM. Processing auditory information:Interference from an irrelevant cue. The Journal of AppliedPsychology, 53: 433–435, 1969.

Stenning K and Oberlander J. A cognitive theory of graphical andlinguistic reasoning: Logic and implementation. CognitiveScience, 19: 97–140, 1995.

Stoianov I, Kramer P, Umilta C, and Zorzi M. Visuospatial primingof the mental number line. Cognition, 106: 770–779, 2008.

Toma K, Matsuoka T, Immisch I, Mima T, Waldvogel D, Koshy B,et al. Generators of movement-related cortical potentials:fMRI-constrained EEG dipole source analysis. NeuroImage,17: 161–173, 2002.

Torralbo A, Santiago J, and Lupianez J. Flexible conceptualprojection of time onto spatial frames of reference. CognitiveScience, 30: 745–757, 2006.

Tversky B. Cognitive origins of graphic conventions. InMarchese FT (Ed), Understanding Images. New York: Springer,1995: 29–53.

Tversky B, Kugelmass S, and Winter A. Cross-cultural anddevelopmental trends in graphic productions. CognitivePsychology, 23: 515–557, 1991.

Umilta C and Nicoletti R. Spatial stimulus–responsecompatibility. In Proctor RW and Reeve TG (Eds), Stimulus–Response Compatibility: An Integrated Perspective. Amsterdam:North Holland, 1990.

c o r t e x 4 7 ( 2 0 1 1 ) 1 4 8 – 1 5 6 155

Author's personal copy

Umilta C and Nicoletti R. An integrated model of the Simon effect.In Alegria J, Holander D, Junca de Morais J, and Redener M(Eds), Analytic Approach to Human Cognition. Amsterdam:Elsevier, 1992: 331–350.

Vallesi A, Binns MA, and Shallice T. An effect of spatial-temporalassociation of response codes: Understanding the cognitiverepresentations of time. Cognition, 107: 501–527, 2008.

Vallesi A, Mapelli D, Schiff S, Amodio P, and Umilta C. Horizontaland vertical Simon effect: Different underlying mechanisms?Cognition, 96: B33–B43, 2005.

Vallesi A and Shallice T. Developmental dissociations ofpreparation over time: Deconstructing the variable foreperiodphenomena. Journal of Experimental Psychology: HumanPerception and Performance, 33: 1377–1388, 2007.

Vallesi A, Shallice T, and Walsh V. Role of the prefrontal cortex in theforeperiod effect: TMS evidence for dual mechanisms in temporalpreparation. Cerebral Cortex, 17: 466–474, 2007.

Weger UW and Pratt J. Time-words guide spatial attention[Abstract]. Journal of Vision, 7: 447a, 2009.

Weger UW and Pratt J. Time flies like an arrow: Space-timecompatibility effects suggest the use of a mental timeline.Psychonomic Bulletin & Review, 15: 426–430, 2008.

Woodrow. The measurement of attention. PsychologicalMonographs, 5: 1–158, 1914.

Xuan B, Chen XC, He S, and Zhang DR. Numerical magnitudemodulates temporal comparison: En ERP study. Brain Research,1269: 135–142, 2009.

Xuan B, Zhang D, He S, and Chen X. Larger stimuli are judgedto last longer. Journal of Vision, 7: 2–5, 2007.

Zacks J and Tversky B. Bars and lines: A study of graphiccommunication. Memory & Cognition, 27: 1073–1079, 1999.

Zebian S. Linkages between number concepts, spatial thinking,and directionality of writing: The SNARC effect and theREVERSE SNARC effect in English and Arabic monoliterates,biliterates, and illiterate Arabic speakers. Journal of Cognitionand Culture, 5: 165–190, 2005.

Ziemann U and Hallett M. Hemispheric asymmetry of ipsilateralmotor cortex activation during unimanual motor tasks:Further evidence for motor dominance. ClinicalNeurophysiology, 112: 107–113, 2001.

Zwaan EWJ. Links en rechts in waarneming en beleving. [Left andright in visual perception as a function of the direction ofwriting]. Ph.D. dissertation, Rijksuniversiteit Utrecht, TheNetherlands (quoted by Winn, 1994), 1965.

c o r t e x 4 7 ( 2 0 1 1 ) 1 4 8 – 1 5 6156