Numerosity processing is context driven even in the subitizing range: an fMRI study

Dear Author, Please, note that changes made to the HTML content will be added to the article before publication, but are not reflected in this PDF. Note also that this file should not be used for submitting corrections.

Numerosity processing is context driven even in the subitizing range:An fMRI study

Tali Leibovich a,b,n, Avishai Henik b,c, Moti Salti a,bQ1

a Department of Cognitive and Brain Sciences, Ben-Gurion University of the Negev, Beer-Sheva, Israelb The Zlotowski Center for Neuroscience, Ben-Gurion University of the Negev, Beer-Sheva, Israelc Department of Psychology, Ben-Gurion University of the Negev, Beer-Sheva, Israel

a r t i c l e i n f o

Article history:Received 6 April 2015Received in revised form30 July 2015Accepted 17 August 2015

KeyQ3 words:Numerical cognitionNumerosity processingContinuous magnitudesCongruity effectSubitizingfMRI

a b s t r a c t

Numerical judgments are involved in almost every aspect of our daily life. They are carried out so effi-ciently that they are often considered to be automatic and innate. However, numerosity of non-symbolicstimuli is highly correlated with its continuous properties (e.g., density, area), and so it is hard to de-termine whether numerosity and continuous properties rely on the same mechanism. Here we examinedthe behavioral and neuronal mechanisms underlying such judgments. We scanned subjects' hemody-namic responses to a numerosity comparison task and to a surface area comparison task. In these tasks,numerical and continuous magnitudes could be either congruent or incongruent. Behaviorally, an in-teraction between the order of the tasks and the relevant dimension modulated the congruency effects.Continuous magnitudes always interfered with numerosity comparison. Numerosity, on the other hand,interfered with the surface area comparison only when participants began with the numerosity task.Hemodynamic activity showed that context (induced by task order) determined the neuronal pathwaysin which the dimensions were processed. Starting with the numerosity task led to enhanced activity inthe right hemisphere, while starting with the continuous task led to enhanced left hemisphere activity.Continuous magnitudes processing relied on activation of the frontal eye field and the post-central gyrus.Processing of numerosities, on the other hand, relied on deactivation of these areas, suggesting activesuppression of the continuous dimension. Accordingly, we suggest that numerosities, even in the sub-itizing range, are not always processed automatically; their processing depends on context and taskdemands.

& 2015 Published by Elsevier Ltd.

1. Introduction

In daily life we rely profoundly on numerical judgments, forexample, to choose the shortest checkout line in a store, orchoosing the larger pile of candy. To study the cognitive me-chanisms underlying numerical cognition, many studies used non-symbolic representations of numbers—arrays of items—as stimuli.In a typical non-symbolic numerosity comparison task, partici-pants are presented with two arrays of items and asked to choosethe array containing more dots (e.g., Cantlon et al., 2006; Piazza,2010). These studies have suggested that processing of numer-osities is innate and automatic (e.g., Cantlon et al., 2009; Coubartet al., 2014; Feigenson et al., 2004). Moreover, It was proposed thatdue to their importance, numerosities are processed by a dedi-cated brain circuitry (e.g., Dehaene et al., 2003; Harvey et al.,

2013).However, recent studies have demonstrated that it is very dif-

ficult to study the mechanisms underlying non-symbolic numer-osity in isolation from continuous magnitudes. Indeed, numer-osities and continuous magnitudes usually correlate. For example,more apples will fill more of a bag than fewer ones; when placedon the ground they would either occupy more area, or be morecrowded. Therefore, whenever two to-be-compared arrays differin numerosity, they also differ in their continuous magnitudes.These continuous magnitudes can potentially influence perfor-mance (for reviews see Leibovich and Henik (2013) and Mix et al.(2002)). Studies investigating the impact of continuous magni-tudes in non-symbolic numerical comparison tasks have sug-gested that numerical abilities are not primary but depend onother cognitive abilities and do not rely on designated brain re-gions (e.g., Cantrell and Smith, 2013; Gebuis and Reynvoet, 2012a,2013; Leibovich and Henik, 2014). Accordingly, it is unclear whe-ther numerosity and continuous magnitudes processing are se-parable and what behavioral and neuronal mechanisms theyshare.

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

67686970717273747576777879808182838485868788899091

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/neuropsychologia

Neuropsychologia

http://dx.doi.org/10.1016/j.neuropsychologia.2015.08.0160028-3932/& 2015 Published by Elsevier Ltd.

n Corresponding author at: Ben-Gurion University of the Negev, Department ofCognitive and Brain Sciences, P.O.B. 653, Beer-Sheva 84105, Israel.

E-mail address: labovich@post.bgu.ac.il (T. Leibovich).

Please cite this article as: Leibovich, T., et al., Numerosity processing is context driven even in the subitizing range: An fMRI study.Neuropsychologia (2015), http://dx.doi.org/10.1016/j.neuropsychologia.2015.08.016i

Neuropsychologia ∎ (∎∎∎∎) ∎∎∎–∎∎∎

1.1. Numerosity and continuous magnitudes—a mere correlation orinseparable processing?

The notion of numerosity as an innate and primary ability isbased on studies with non-human animals (McComb et al., 1994;Nieder and Dehaene, 2009; Pisa and Agrillo, 2008), young babiesand newborns (e.g., Cantlon et al., 2009; Coubart et al., 2014; Xuand Spelke, 2000) who exhibited a spontaneous ability to dis-criminate numerosities. These studies, however, suffered from aninherent confound. The numerosity in the presented arrays wascorrelated with continuous magnitudes (Mix et al., 2002). It istherefore possible that animals and newborns do not rely on nu-merosities to discriminate between arrays but instead they rely oncontinuous magnitudes (for an extended discussion see Leibovichand Henik (2013)). In 2002, Mix et al. showed that results ofseminal developmental studies that used non-symbolic numer-osities could be explained by discrimination of continuous mag-nitudes. For example, Xu and Spelke (2000) habituated 6-month-old babies to arrays containing dots of different sizes. The testdisplay was different either in the number of the dots or in theoverall area of the dots. Since the babies looked longer when thenumber of dots changed (compared to when the area changed),the authors concluded that babies were sensitive to changes innumerosities. However, Mix et al. noted that the contour lengthwas positively correlated with numerosity. Hence, babies mighthave responded to changes in contour length rather thannumerosity.

With the problem of correlation between numerosity andcontinuous magnitudes in mind, different studies employed dif-ferent methods to control for a possible influence of continuousmagnitudes in the experimental designs. Some manipulated onlyone continuous magnitude at a time (Mussolin et al., 2010), othersassigned a random dot size to each array (Piazza and Izard, 2009),some used a single array containing two different colors of dotswhere participants had to indicate the color of the more numerousdots (Halberda et al., 2008), etc.

All these methods shared an inherent assumption that con-tinuous magnitudes are not processed automatically and wouldnot affect performance unless they reliably predicted numerosity.Recent studies, however, suggested that continuous magnitudesare processed even when they are irrelevant and are an unreliablecue of numerosity. Leibovich and Henik (2014) presented partici-pants with pairs of dot arrays (5–25 dots per array) and askedthem to choose the array containing more dots. Using stimuligenerated by code provided by Gebuis and Reynvoet (2011), nu-merosity was minimally correlated with continuous magnitudes,so the magnitudes were neither a reliable cue of numerosity norrelevant to the task. Regression analysis with the numerosity ratioand the ratio of five continuous magnitudes as predictors (i.e.,average diameter, total surface area (the sum of all dots area in thearray), area extended (the smallest contour that included all of thedots, as if an elastic band was wrapped around the dots), densityand total circumference) revealed that half of the explained var-iance in response times (RTs) was explained by continuousmagnitudes.

The importance and utility of the continuous magnitudes innumerosity judgments was demonstrated recently in an eventrelated potentials (ERP) experiment (Gebuis and Reynvoet, 2013).ERPs were recorded while participants passively viewed arrays ofdots in which continuous magnitudes were carefully controlled.Results showed that ERP fluctuations were correlated with con-tinuous magnitude changes but not with numerosity change, evenwhen participants were told that the number of dots wouldchange. The authors concluded that extraction of continuousmagnitudes and not the extraction of numerosities from a visualscene is automatic. In a functional magnetic resonance imaging

(fMRI) study, Leroux et al. (2009) presented a “Piaget-like” task foradults. In the experiment, same-size horizontal bars were pre-sented in two rows. The spaces between the bars were manipu-lated and participants were asked to decide if the two lines con-tained the same number of bars or not. In congruent trials, in therow containing more bars, the bars were further apart from eachother than in the row with the fewer bars (i.e., convex-hull posi-tively correlated with numerosity), and vice-versa in incongruenttrials (i.e., convex-hull negatively correlated with numerosity).Although success rates were high in both conditions, there wasmore activation of brain areas related to conflict monitoring andcognitive control in incongruent trials, suggesting that adultsprocess continuous magnitudes, even when they are not con-sistently correlated with numerosity. Taken together, these studiesshow that continuous magnitudes are processed during numericaljudgments, even when they are completely irrelevant to the taskand do not consistently correlate with numerosity.

1.2. Brain areas related to processing of numerosity and continuousmagnitudes

Behavioral evidence suggesting numerosity processing is basicand innate led to the notion that numerosity processing is exe-cuted by a dedicated brain circuitry (Dehaene, 1997; Dehaene andChangeux, 1993). Efforts to find such a brain region led to con-flicting results. While some studies reported the existence of adedicated neuronal substrate (e.g., Castelli et al., 2006; Chassy andGrodd, 2012; Dehaene et al., 2003; Harvey et al., 2013; Piazza et al.,2007), others proposed numerosities were processed by the samebrain areas as other magnitudes (e.g., Fias et al., 2003; Piazza et al.,2007; Pinel et al., 2004). Naturally, isolating the neuronal me-chanism underlying numerical cognition requires dissociatingnumerosity from continuous magnitudes; but, is that at all pos-sible? In an fMRI study, Chassy and Grodd (2012) had participantscompare either disk size (i.e., continuous task) or numerosity ofdots (i.e., numerosity task). While some brain areas were active tothe same extent for both tasks (e.g., the left middle occipital gyrusand the right supramarginal gyrus), some areas were more activein the numerosity task (e.g., bilateral primary visual cortices(BA17), right superior parietal lobule, and the bilateral middleoccipital gyri). In this task, all the dots were presented in the samesize. As a result, continuous magnitude and numerosity werecorrelated and the same strategy could have been employed forboth tasks. This can explain the overlapping areas. Importantly, inthe numerosity comparison task, one has to integrate items, thatis, to sum the numerosity (or the total surface area) of each arraybefore comparing. Such a step does not exist in the continuouscomparison task. Hence, the areas that were more active duringthe numerosity comparison task might reflect item integration andnot processing of numerosities.

In a more recent study aiming to find brain circuitry specific tonumerosity processing, Harvey et al. (2013) asked whether pro-cessing of numerosity is organized topographically in the brain.Participants were presented with 1–7 dots. To control for the in-fluence of continuous magnitudes, five different groups of 3 arraysof dots were used; in each different group, a different continuousmagnitude was kept constant across all numerosities. For example,in the “constant area” group of stimuli, the total surface area of the3 arrays (i.e., with 1, 4 and 7 dots) was identical; that is, the 1 dotin the 1-dot array was 4 times larger than the 4 dots in the 4-dotarray, and at the same time the 1 dot was 7 times larger than the7 dots in the 7-dot array (see Fig. 1 in Harvey et al. (2013)). Theauthors reported that areas in the right posterior parietal lobeshowed a topographical representation, with small numerositiesprocessed in more medial areas and larger numerosities in morelateral areas. However, as suggested by Gebuis et al. (2014), while

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

T. Leibovich et al. / Neuropsychologia ∎ (∎∎∎∎) ∎∎∎–∎∎∎2

one continuous magnitude was controlled for, all the other con-tinuous magnitudes correlated with numerosity. Hence, the to-pographic map found “reflects a weighted response of neuronsthat encode different sensory cues rather than a pure numerosityestimate” (p. 1). Up until this point, it is unclear whether the au-tomaticity associated with numerical processing is a consequenceof continuous magnitudes processing, or whether it is a primaryfunction. As such, it is unclear whether brain areas dedicated tonumerosity processing exist.

1.3. The current study

The current fMRI study has two major aims: first, to investigatethe reciprocal relations between numerosity and continuousmagnitudes; and second, to investigate the corresponding brainmechanisms underlying these relations.

We chose to use numerosities between 2 and 4. This range ofnumerosities is known as the subitizing range (Revkin et al., 2008;Trick and Pylyshyn, 1994). The main reason for choosing this rangeis that estimation of numerosities in this range has been con-sistently found to be very fast and accurate. Some researchers evenassert that subitizing is a perceptual rather than a numericalprocess (Hyde, 2011; but see also Burr et al. (2010)). Thus, we usedthe easiest numerical task available relative to continuous mag-nitudes to decrease salience differences between the tasks. In eachtrial, participants were presented with two arrays of gray dots andwere asked to indicate where there were more dots (i.e., numericaltask) or more gray area (i.e., continuous task), while their neuralactivity was measured using fMRI. Half of the trials were con-gruent; namely, the array containing more dots also containedmore surface area, was denser, had larger dots, etc., and vice-versafor incongruent trials. The same stimuli were used for both tasks.Importantly, half of the participants started with the numericaltask (i.e., the NC group) and half with the continuous task (i.e., theCN group), and the effect of order was tested at both the beha-vioral and functional levels.

To investigate the reciprocal relations at the behavioral level,we analyzed the difference (in RTs and error rates) between con-gruent and incongruent trials (i.e., the congruity effect) for eachtask and each group. A congruity effect in the numerical taskwould indicate that the irrelevant continuous magnitudes wereprocessed automatically. A congruity effect in the continuous taskwould indicate that the irrelevant numerosity was processed au-tomatically. An asymmetric congruity effect, namely a congruityeffect in the numerical task but not in the continuous task (i.e.,

task� congruity interaction), would support the notion that con-tinuous magnitudes underlie numerosity processing. If the size ofthe congruity effect is modulated not only by task, but also bytasks order (i.e., task� order interaction), it would suggest that thereciprocal relations between the dimensions are not constant, butdepend on the context in which the task is being performed.

In order to understand the underlying neural mechanisms ofthe reciprocal relationship between numerosity and continuousmagnitudes, we conducted a whole brain analysis correspondingto the behavioral ones. We contrasted the different groups,namely, the group that started with the numerical task and thegroup that started with the continuous task, in order to revealbrain areas where the activity was modulated by the context inwhich the task was performed.

We contrasted the brain activity during incongruent trials ofthe numerosity and continuous tasks, separately for the differentgroups, to test whether the brain resolves conflict differently whenthe conflict arises from different types of magnitudes. To testwhether activity during conflicting trials is further modulated bycontext, we also contrasted incongruent trials from both groups(NC group: incongruent continuous task4 incongruent numericaltask)4(CN group: incongruent continuous task4 incongruentnumerical task).

To explore the brain areas activated when top-down attentionis directed towards numerosity and continuous magnitudes, wecontrasted the brain activity during congruent trials of the nu-merosity and continuous tasks separately for the different groups.If there are brain areas that are more active when attention isdirected to numerosities, they would be more active in the con-gruent trials of the numerical task (for a similar idea see O’Cravenet al., 1999). To test whether activity during congruent trials isfurther modulated by context, we also contrasted congruent trialsfrom both groups (NC group: congruent continuoustask4congruent numerical task)4(CN group: congruent con-tinuous task4congruent numerical task).

2. Methods

2.1. Participants

Forty-eight students from Ben-Gurion University of the Negev participated inthe experiment. The experimental procedures were approved by the local Helsinkicommittee. All participants were right-handed, monolingual native Hebrewspeakers, with intact or corrected vision, and no reported learning disabilities orattention deficits. Participants were compensated for their participation in the

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 1. Example for congruent and incongruent trials.

T. Leibovich et al. / Neuropsychologia ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 3

experiment with a monetary reimbursement (100 NIS). Eight participants wereexcluded from the analysis; three subjects were removed due to excessive motionduring scanning (we allowed no more than a 3 mm deviation from the first imagecollected and no more than a 1 mm deviation between one functional image to thenext functional image); three others were excluded because they did not complywith the tasks instructions (e.g., responding to the numerical and not the con-tinuous magnitudes or vice versa) and two were excluded due to technical pro-blems during the experiment. For the remaining 40 participants, 20 (seven females)started with the numerosity task, and 20 (nine females) started with the con-tinuous task. The average age of the participants was 24 years and 5 months (SD¼2years and 1 month). There were no significant age differences between the groups(to1, ns).

2.2. Stimuli

Stimuli were composed of pairs of light gray dot arrays presented on a blackbackground and separated by a vertical gray line. Stimuli were presented in thecenter of a 1024�768 pixel screen consisting of an area of 300�175 pixels. The dotarray pairs were created with Matlab code detailed in Gebuis and Reynvoet (2011).This code records five different continuous magnitudes separately for each array ina pair. It recorded average diameter, total surface area (the sum of all dots area inthe array), area extended (the smallest contour that included all of the dots, as if anelastic band was wrapped around the dots), density and total circumference. Eacharray contained 2, 3 or 4 dots. Pairs were always composed of two different-sizeddot arrays. One dot was not used since terms such as density and convex-hull donot apply for one dot.

2.3. Congruity manipulation

Numerical magnitude was either congruent or incongruent with all five con-tinuous properties. In congruent stimuli, in the array containing more dots, theaverage size of the dots, their total circumference, total surface area and the areaoccupied by the dots was larger and the dots were denser, compared with the arraycontaining fewer dots. In incongruent stimuli, all five continuous properties weresmaller compared to the array that contained the more numerous dots (see Fig. 1).

2.4. Controlling for the continuous magnitudes of the dot arrays

In order to assure that in our set of stimuli the influence of continuous mag-nitudes was consistent across different numerical ratios, the following analyseswere conducted. The current set of stimuli had a numerical ratio of either 0.5 (i.e.,2 versus 4 dots), 0.67 (i.e., 2 versus 3 dots) or 0.75 (i.e., 3 versus 4 dots). To check ifthe ratio between the 5 continuous magnitudes mentioned above was different inthe three numerical ratios, we performed a two-way analysis of variance (ANOVA)with each ratio as a group (i.e., 3 groups of numerical ratio; 0.5, 0.67 and 0.75),continuous magnitudes as the independent measure (5 levels for 5 different con-tinuous magnitudes), and the ratio between the continuous magnitudes of eachpair of dot arrays as the dependent measure. This analysis revealed no interactionbetween continuous properties and numerical ratio groups, F (8, 468)¼1.12,p¼0.35, ƞ2p¼0.02, suggesting that the continuous ratios did not differ among thedifferent groups of dots. Therefore, in the continuous task, the level of difficultyshould have been similar across different numerical ratios. Moreover, in the nu-merical task, the level of interference from continuous magnitudes should havebeen similar across different numerical ratios.

Numerical ratio might affect performance even in the subitizing range (Hyde,2011). Namely, the larger the ratio, the more difficult it is to decide which arraycontains more dots (Moyer and Landauer, 1967). This ratio effect might also explainsome of the congruency effect. Barth (2008) showed that the ratio between thecumulative area in incongruent trials tends to be higher (i.e., the magnitudes aremore similar and the ratio between them is closer to 1) than in congruent trials. Insuch cases, one might suggest that the effect is driven not by congruency of nu-merical and continuous magnitudes but instead it is driven by the higher ratio ofincongruent compared to congruent trials. To avoid this confound we kept the ratiobetween average continuous properties in all our stimuli (both congruent and in-congruent) within a constant range. The average continuous ratio for a pair of dotswas calculated by averaging the ratio of all five continuous magnitudes for everypair of arrays. For example, for pair “x”, where the ratio between the densities is0.5, between the total surface areas is 0.7, between average dot sizes is 0.3, betweentotal circumference is 0.6 and between the area occupied by the arrays is 0.5, theaverage continuous ratio would be 0.52. The range of average continuous ratios forall our stimuli was between 0.34 and 0.41 (average¼0.38, SD¼0.01).

2.5. Tasks

Participants performed two tasks in two separate runs. The only differencebetween the two runs was the task instructions given at the beginning of each run.In the numerosity task, participants were asked to choose the array containing

more dots. In the continuous task, participants were asked to choose the arraycontaining more gray area. Each run contained 120 stimuli: 2(congruity)�6(pairs)�10(different variations for each pair). The side of the larger dot array wascounterbalanced. All stimuli were repeated only once in each run. Thus, partici-pants encountered the same stimuli twice: once in the numerical run and once inthe continuous run.

2.6. Procedure

Before starting the scan, participants signed a consent form and were givengeneral instructions. An event-related fMRI design was used to acquire functionalimaging data. Each scan included an anatomical scan and two functional runs.Before each functional run, participants read instructions specific to the task. Eachtrial started with a black screen with a vertical gray line in its center. In the middleof the gray line a red fixation cross appeared. To achieve deconvolution of the bloodoxygen level dependent (BOLD) response, a jitter interval between 4000 and6000 ms (120 different intervals in total) was used before the fixation crosschanged to green (for 250 ms) in order to alert the participant that a dot stimuluswas soon to appear and thereby allowing participants to get ready to respond. Afterthe green fixation cross disappeared, a black screen with a vertical gray line waspresented for 700 ms. Thereafter, two dot arrays were presented for 700 ms andthen were replaced by a black screen with a gray vertical line for 1100 ms. Parti-cipants were able to respond to the stimuli from the time the dot arrays weredisplayed on the screen until a new trial with a red cross started. The participantswere instructed to respond with a button press with the hand corresponding to theside of the presentation of the array containing the larger area/numerosity. Be-tween the first and second run, there was a four-minute break where participantswatched a short video of nature scenes. This was done in an effort to preventgeneral habituation to the dot stimuli that could potentially have reduced brainactivity during the second run. The procedure of a typical trial is depicted in Fig. 2.

2.7. fMRI data acquisition

A 3-Tesla Philips ingenia whole-body MRI scanner was used to collect thefunctional and structural data of this study. The brain anatomy of each participantwas collected with high-resolution T1 weighted sequence (1�1x1 mm3, TR:8200 ms, TE: 3.8 ms, flip angle: 8°). An echo planar (EPI SE) sequence was used tomeasure the BOLD brain signal of the functional run with a 32-channel Siemenshead coil. The order of imaging acquisition was ascending – interleaved, coveringthe whole brain of participants. The acquisition resulted in 353 whole-brain imagesper functional run, with a total length of 10 min and 30 s per run. For each func-tional volume, 38 slices were collected resulting in a 3 mm isovoxel resolution overa 64�64 voxel matrix. The time of repetition (TR) was 2700 ms, the echo time (TE)was 52 ms and the flip angle was 78°.

2.8. Imaging analysis

Brain Voyager QX 2.8 (Brain Innovation, Maastricht, The Netherlands) was usedto analyze the functional and structural data sets. Each individual data set waspreprocessed according to the following steps. Functional imaging data were firstcorrected for slice scan time acquisition (ascending – interleaved; using a cubic-spline interpolation algorithm). A high-pass (GLM – Fourier) frequency filter with acut off value of 2 sines/cosines cycles was applied to remove low frequency signals.Finally, a Trilinear/sinc interpolation approach was used to remove and to adjusthead motions. In order to be included into the study, a participant's movementparameters had to stay within 3 mm of overall movement (maximum deviation

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 2. An example of a typical trial. Q4(For interpretation of the references to color inthis figure, the reader is referred to the web version of this article.)

from the first volume) and within 1 mm volume-to-volume movement (maximumdeviation from one collected functional image to the next collected functionalimage).

An automatic alignment procedure (implemented in Brain Voyager) was usedin order to spatially align the functional runs of each participant onto the corre-sponding anatomical scan. The quality of the alignment was checked visually andcorrected manually if the automatic procedure did not reveal a sufficient align-ment. Subsequently, the co-aligned images were transformed into Talairach space(Talairach and Tournoux, 1988). This was achieved in two consecutive steps: first,using the landmarks of the anterior commissure (AC) and the posterior commissure(PC), the anatomical image of each participant was transformed into the ACPC-plane position. Then the boundaries of the brain tissue were manually selected andtransformed into the Talairach grid using a trilinear interpolation algorithm (Ta-lairach and Tournoux, 1988).

Individual data sets were entered into a general linear model (GLM) for group-based analysis. All functional events of the two conditions (i.e., numerical andcontinuous) were convolved with a two-gamma hemodynamic response function(HRF) in order to predict the BOLD function (Friston et al., 1998). Congruent andincongruent trials across the two functional runs were modeled separately in orderto investigate brain activation differences related to congruency. The statisticalmaps derived from brain activation contrasts (see below) were applied a thresholdwith an uncorrected p Value of 0.005 and subsequently cluster corrected in order tocorrect for multiple comparisons and to adjust the Type I error to a level of po0.05.This was achieved by an iterative “Monte Carlo Simulation” (1000 iterations), whichestimates the minimum size of a functional cluster to be significant on the basis offunctional data from the present study (Forman et al., 1995).

3. Results

3.1. Behavioral analysis

An ANOVA with task and congruity as within-subject variables,and the order of task administration as a between-subjects vari-able, was performed twice: once with error rates and once with RTas dependent variables. Averages and standard errors (SEs) aredepicted in Fig. 3. A main effect of order was marginally significantfor error rates, F (1, 38)¼3.06, p¼0.09, ƞ2p¼0.07, and only showeda pattern for RTs, F (1, 38)¼1.6, p¼0.2, ƞ2p¼0.04. This patternsuggests that the group that started with the continuous task (i.e.,the CN group) was slightly more accurate and faster than the NCgroup.

Both RTs and error rates were affected by task demands, con-gruity and order. Participants were generally more accurate andfaster in congruent than incongruent trials across tasks; F (1, 38)¼24.36, po0.001, ƞ2p¼0.39 and F (1, 38)¼72.85, po0.001,ƞ2p¼0.66, for accuracy and RT, respectively. Responses to thecontinuous task were generally more accurate and faster than tothe numerosity task; F (1, 38)¼33.65, po0.001, ƞ2p¼0.47 and F (1,38)¼131.91, po0.001, ƞ2p¼0.77, for accuracy and RT, respectively.

The interaction of task and order revealed that error rates inthe numerosity task were modulated by task order; in the nu-merosity task, participants in the NC group were less accurate thanparticipants in the CN group. This modulation was specific to thenumerosity task; error rates in the continuous task were notmodulated by task order, F (1, 38)¼4.63, po0.05, ƞ2p¼0.11. ForRTs, the interaction of task and order was marginally significant; F(1, 38)¼3.69, p¼0.06, ƞ2p¼0.09. Further investigation of this in-teraction revealed that while order did not affect RTs in the nu-merosity task (Fo1, ns), performance in the continuous task wasslightly faster in the CN group, F (1, 38)¼3.02, p¼0.09, ƞ2p¼0.07.

The interaction of task and congruity revealed that the differ-ence in error rates and in RT between congruent and incongruenttrials (i.e., the congruity effect) was greater in the numerosity taskthan in the continuous task, F (1, 38)¼18.88, po0.001, ƞ2p¼0.33and F (1, 38)¼10.63, po0.05, ƞ2p¼0.22, for error rates and RT,respectively. For RT, there was also a triple interaction of task,congruity and order, F (1, 38)¼6.09, po0.05, ƞ2p¼0.14; in the NCgroup the congruity effect was similar for both tasks (Fo1, ns). Incontrast, in the CN group there was no congruity effect in the

continuous task, F (1, 38)¼1.22, p¼0.27, ƞ2p¼0.03, whereas thesize of the congruity effect in the numerosity task remained si-milar to the congruity effect in the numerosity task of the NCgroup (Fo1, ns).

3.2. Functional analysis

To test our hypothesis that task order might affect the waymagnitudes are processed and also in light of the effect of order onRT and accuracy, we conducted a whole-brain analysis for eachgroup separately, in addition to the contrasts between the groups.

3.2.1. Main effect of orderIn order to investigate how task order affected brain activation,

we calculated a whole-brain t-test statistic, contrasting the brainsignals of both groups (NC and CN) across all conditions. The re-sults of this analysis (Table 1) revealed two regions of activation –

the posterior medial aspect of the cingulate gyrus, and the rightsuperior frontal gyrus (Fig. 4). These areas were more strongly

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 3. Behavioral analysis. (A) Error rates in the different groups and conditions.(B) RTs in the different groups and conditions. All participants conducted bothtasks. The NC group started with the numerosity task (i.e., where are there moredots?) and the CN group started with the continuous task (i.e., where is there moregray area?). C¼congruent, IC¼ incongruent. * po0.05.

Table 1Order of main effect.

Brain region Coordinates

x y z t Cluster size

Order main effect: CN4NCR.SFG/MiFG (FEF) 11 37 36 4.9 1141PoG (cingulate gyrus) �1 �53 45 3.68 799

Note: po0.005 (cluster corrected for multiple comparisons: p¼0.05). Coordinatesare in Talairach space. SFG¼superior frontal gyrus; MiFG¼middle frontal gyrus;FEF¼frontal eye field; PoG¼postcentral gyrus.

activated in the CN group than in the NC group.

3.2.2. Brain activation related to task effect for congruent trialsIn order to investigate how the different tasks affected brain

activation, we performed a whole-brain t-test, contrasting brainsignals related to the different relevant dimensions according totask instructions (i.e., either numerosity or continuous magni-tudes). This was done only for congruent trials; in these trials,participants in both tasks viewed the same stimuli and respondedwith the same hand. The only difference was task instructions,directing attention to different dimensions. This analysis wasmade for each group separately because of possible effects oforder.

3.2.2.1. Brain activation for the different tasks in the CN group. Theresults of the CN group analysis revealed fronto-parietal areas inthe left hemisphere that were more strongly activated during thecontinuous task; the superior frontal gyrus, precentral gyrus andthe middle frontal gyrus. No areas were found to be more stronglyactivated during the numerosity task in this group (see Table 2 andFig. 5 – yellow clusters).

3.2.2.2. Brain activation for the different tasks in the NC group. Theresults of the NC group analysis revealed fronto-parietal areas thatwere more strongly activated during the numerosity task; theright middle frontal gyrus and the right precentral gyrus in thefrontal lobe, and the left postcentral gyrus (around the insula) andthe left supramarginal gyrus (around the postcentral gyrus) in theparietal lobe. No areas were found to be more strongly activatedduring the continuous task in this group (see Table 2 and Fig. 5 –

red clusters).

3.2.2.3. Differences in brain activation related to task and order. Toinvestigate if the impact of task was further modulated by taskorder, we performed a whole-brain t-test, pitting the brain signalsrelated to task for the congruent trials of the NC group to those ofthe CN group. The results of this analysis revealed activation infrontal, parietal and occipital areas; the right cingulate gyrus (inthe area of the superior frontal gyrus), right middle frontal gyrus,left insula (in the area of the precentral gyrus), and the inferiortemporal gyrus, on the border of the occipital gyrus. In the CNgroup, these areas were more strongly activated during the con-tinuous task; in the NC group the pattern was reversed—the sameareas were more strongly activated during the numerosity task. Inthe left inferior frontal gyrus on the boarder of the orbitofrontalgyrus, the activation was negative compared to baseline. Namely,

in the CN group this area was more strongly deactivated in thecontinuous task and in the NC group, this area was more stronglydeactivated during the numerosity task (see Table 2 and Fig. 5 –

blue clusters).

3.2.3. Brain activation related to conflict in the different tasksTo investigate the impact of conflict from different sources (e.g.,

when the to-be-ignored dimension was different), we performed awhole-brain t-test, pitting the brain signals related to incongruenttrials in the numerosity task with those of the continuous task.This was done separately for each group.

3.2.3.1. Brain activation related to conflict in the different tasks forthe CN group. This analysis revealed parietal and occipital areas inthe right hemisphere that were modulated by the different sourcesof conflict. Specifically, the right postcentral gyrus in the area ofthe supramarginal gyrus was more active during incongruent trialsof the continuous task (where the irrelevant dimension was nu-merosity), and the right posterior cingulate and the right inferioroccipital gyrus (the striate area) were more strongly activatedduring incongruent trials of the numerosity task (where the irre-levant dimension was continuous) (see Table 3 and Fig. 6 – yellowclusters).

3.2.3.2. Brain activation related to conflict in the different tasks forthe NC group. This analysis revealed activation in frontal andparietal areas in the left and right hemispheres—along the rightsuperior frontal gyrus, middle frontal gyrus and cingulate gyrus,and the left pre and postcentral gyri. These areas were morestrongly activated in the NC group during the incongruent trials ofthe numerosity task—where the irrelevant dimension was con-tinuous (see Table 3 and Fig. 6 – red clusters). There were no areasthat were more active during the continuous task.

3.2.3.3. Differences in brain activation related to conflict and taskorder differences. To investigate if the impact of conflict from dif-ferent sources was further modulated by task order, we performed

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 4. Main effect for task. Transversal view; x¼0, y¼�51. The z coordinates areindicated in gray in the figure. R.SFG: right superior frontal gyrus; MiFG: middlefrontal gyrus. PoG: postcentral gyrus.

Table 2Task effect for congruent trials.

CN; congruent; Cont4NumL.SFG �4 55 33 5.05 642L.PrG �7 �38 66 5.5 1612L.MiFG �22 13 42 4.19 679

NC; congruent; Num4ContR.SMG/IPL 44 �32 48 5.27 1252R.PrG 29 �23 63 4.6 864R.SFG 5 �11 60 4.18 1140L.PoG (Insula) �52 �17 15 5.16 633

Congruent; [NC (Num4Cont)]4[CN(Cont4Num)]R.ITG/OcG 50 �65 3 5.25 1396R.SFG/MiFG 23 �17 63 4.13 802R.SFG 5 �11 60 4.38 845L.SFG (cingulate gyrus) �7 �17 45 4.34 2304L.PrG (Insula) �43 �26 18 5.18 2774L.IFG/OFG �43 31 �9 �3.81 603

Note: po0.005 (cluster corrected for multiple comparisons: p¼0.05). Coordinatesare in Talairach space. SFG¼superior frontal gyrus; MiFG¼middle frontal gyrus;PoG¼postcentral gyrus; SMG¼supramarginal gyrus; PrG¼precentral gyrus;OcG¼occipital gyrus; ITG¼ inferior temporal gyrus; IFG¼ inferior frontal gyrus;OFG¼orbitofrontal gyrus; IPL¼ inferior parietal lobule.

a whole-brain t-test, pitting the brain signals related to task for theincongruent trials of the NC group to those of the CN group. Theresults of this analysis revealed activation in frontal, parietal andoccipital areas in the left and right hemispheres; the right and left

superior frontal gyri, the left insula, the right supramarginal gyrusin the area of the inferior parietal lobe and the intraparietal sulcus(IPS), and the right inferior temporal gyrus on the boarder of theoccipital gyrus. In the CN group, the activity in these areas wassimilar. However, in the NC group these areas were more stronglyactivated in the numerosity task (see Table 3 and Fig. 6 – blueclusters).

4. Discussion

In the current work, we investigated the neuronal mechanismsunderlying magnitude comparisons. Namely, we were interestedin examining the effect of processing continuous magnitudes onnumerosities in the subitizing range and vice versa. We chose thesubitizing range because estimating numerosities in this range isfast and accurate (e.g., Knops et al., 2014; Piazza et al., 2011; Re-vkin et al., 2008; Trick and Pylyshyn, 1994). Furthermore, we chosethis range in order to try to make the processing of numerosities aseasy as possible. Despite these efforts, the behavioral and func-tional results suggest that numerosity comparison was more dif-ficult than continuous magnitudes comparison. In addition, con-tinuous magnitudes always influenced numerosity comparisons,whereas numerosities affected continuous magnitude compar-isons only if the concept of numerosities was primed. Our resultssuggest that performance during a comparative judgment task isaffected by context; namely, the relevance of the dimension andthe order of the tasks.

4.1. Performance in a comparative judgment task is affected by orderand relevance of the dimension even in the subitizing range

The behavioral results exhibited an asymmetrical pattern;continuous magnitudes interfered with numerical magnitudejudgment regardless of task order, whereas numerosities affectedcontinuous magnitude judgment only if the numerosity task

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 5. Brain activation related to the task effect for congruent trials. Transversal view; x¼0, y¼�40. z coordinates are indicated in gray in the figure.; MiFG¼middle frontalgyrus; PoG¼postcentral gyrus; SMG¼supramarginal gyrus; PrG¼precentral gyrus; OcG¼occipital gyrus; ITG¼ inferior temporal gyrus; IFG¼ inferior frontal gyrus;OFG¼orbitofrontal gyrus; IPL¼ inferior parietal lobule. AnG¼angular gyrus. (For interpretation of the references to color in this figure, the reader is referred to the webversion of this article.)

Table 3Task effect for incongruent trials.

CN; incongruent; Cont4NumR.PoG/SMG (IPS) 60 �29 42 �4.99 690

CN; incongruent; Num4ContR.PoG 17 �62 12 4.92 1088R.IOG �10 �89 �12 4.45 860

NC; incongruent; Num4ContR.PoG/SMG (IPS) 56 �26 45 4.8 964R.SFG/MiFG 26 �17 66 4.79 1731SFG (cingulate gyrus) �1 �14 45 4.73 2735L.PrG �16 �11 66 4.34 915L.PoG �43 �26 27 4.62 981

Incongruent; [NC (Num4Cont)]4[CN (Cont4Num)]R.SMG/IPS 59 �29 42 4.21 1297R.ITG/OcG 47 �65 3 5.27 810R.SFG 20 �14 60 3.89 868L.PrG �16 �14 63 4.42 755L.IPS (AnG) �37 �50 48 4.05 797L.Insula �43 �29 21 4.71 949

Note: po0.005 (cluster corrected for multiple comparisons: p¼0.05). Coordinatesare in Talairach space. PoG¼postcentral gyrus; SMG¼supramarginal gyrus;IOG¼ inferior orbital gyrus; SFG¼superior frontal gyrus; PrG¼precentral gyrus;MiFG¼middle frontal gyrus; IPL¼ inferior parietal lobe; ITG¼ inferior temporalgyrus; OcG¼occipital gyrus; IPS¼ intraparietal sulcus; AnG¼angular gyrus.

preceded the continuous task. This pattern of results is in line withprevious suggestions that continuous properties affect perfor-mance even when irrelevant to the task and not correlated withnumerosities (Gebuis and Reynvoet, 2012a; Leibovich and Henik,2014).

Similar context-dependency has been shown with a basic vi-sual feature like color (Yee et al., 2012). Yee and colleagues' studyincluded two tasks: a color-word Stroop task, where participantshad to name the font color and avoid reading the word; and apriming task, in which participants were presented with a primeword and a target word, and asked to decide whether the targetword was the name of an animal. The order of the tasks wascounterbalanced. The results revealed that responses were fasterwhen the prime and the target word represented the same color(e.g., emerald and cucumber), compared with target words inwhich there was no such connection. This, however, occurred onlywhen the color-word Stroop task preceded the priming task,namely, only when the notion of color was primed. Accordingly,the authors suggested that the extent to which color informationis activated depends not only on long-term factors but also onshort-term, context-dependent factors.

Following the logic of Yee et al. (2012), our result suggests thatthe extent to which numerical information is processed is modu-lated by short-term, context-related factors as well; it is not au-tomatically processed all the time. In contrast to numerosities, theprocessing of continuous magnitudes was not modulated by con-text; the same congruity effect appeared in the numerical task,regardless of tasks order.

The novel finding here is that continuous magnitudes affectperformance even when the to-be-compared numerosities are inthe subitizing range—where it is the easiest. Nonetheless, someresearchers assume that in the subitizing range a different systemis applied than for larger numerosities (i.e., the two-system view;Feigenson et al., 2004). Therefore, it should be tested whether the

same pattern of results would generalize to larger numerosities aswell.

Some studies suggest that subitizing is made possible by aparallel individuation process; each item is assigned an index to-ken in parallel; the tokens are then mapped into number names.Since indexing and parallel individuation are limited by the ca-pacity of visual working memory, this process is possible only withsmall numerosities (Knops et al., 2014; Trick and Pylyshyn, 1994).Dehaene and Changeux (1993) suggested that individuation doesnot take into account different continuous magnitudes. If such anindividuation process that ignores continuous magnitudes (i.e.,size, density) occurred during the numerosity comparison task, wewould not expect continuous magnitudes to be processed whenirrelevant. Therefore, we would not expect our resultant congruityeffect. We argue that this congruity effect, in the numerosity task,suggests that even in the subitizing range, continuous magnitudesinfluence performance. This notion may support the one-systemview, suggesting that all numerosities, within and outside of thesubitizing range, are represented by the approximate numbersystem (e.g., ANS).

There are, however, findings in the literature supporting therole of individuation in the subitizing range. Knops et al. (2014),for example, demonstrated that object enumeration and visualworking memory share a neural mechanism. Namely, that thenumerosity of objects in the subitizing range is represented viasaliency maps; maps that topographically represent the saliency ofitems in a specific location (for a different view see Pagano et al.,2014). The authors also provide evidence for the flexibility of thismechanism and its ability to deal with changing environmentaldemands. Importantly, the findings of Knops et al. support thepossibility that during individuation, the continuous magnitudesof the items (e.g., the individual item size) or of the array (e.g., thedensity of the items) are also encoded. Thus, it is possible thatduring the comparison of two numerosities in the subitizing

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 6. Brain activation related to conflict in the different tasks. Transversal view; x¼0, y¼�40. z coordinates are indicated in gray in the figure. PoG¼postcentral gyrus;SMG¼supramarginal gyrus; IOG¼ inferior orbital gyrus; SFG¼superior frontal gyrus; PrG¼precentral gyrus; MiFG¼middle frontal gyrus; IPL¼ inferior parietal lobe;ITG¼ inferior temporal gyrus; OcG¼occipital gyrus; IPS¼ intraparietal sulcus; AnG¼angular gyrus. (For interpretation of the references to color in this figure, the reader isreferred to the web version of this article.)

range, an individuation process occurs that takes into account thecontinuous magnitudes of the stimuli. Moreover, because thissystem is flexible, the magnitude that would be encoded can becontext-dependent, as demonstrated by the con-gruity� task� order interaction in our results. Further workshould empirically test this notion.

4.2. The effect of task order on brain activity

The posterior cingulate was more active in the group thatstarted with the continuous task. The posterior cingulate has a keyrole in the default mode network (Fransson and Marrelec, 2008).Activations of areas in the default mode network are inverselyrelated to task difficulty; the more difficult a task, the less activethe area. According to the behavioral results, the group that startedwith the continuous task was generally faster than the group thatstarted with the numerosity task. This suggests that the tasks wereeasier and required fewer resources in the CN group, since thenotion of numerosity was not prompted and hence, did not in-terfere with the continuous task.

The posterior cingulate also plays a role in shifts of visual at-tention (Small et al., 2003); it monitors eye movements and re-sponds to sensory stimuli (Vogt et al., 1992). The differential ac-tivation in this area according to task order suggests that the twogroups used different scanning strategies. Namely, the group thatstarted with the numerosity task could have used less eye move-ments, since numerosities in the subitizing range may be graspedin parallel (e.g., Trick and Pylyshyn, 1994). Watson et al. (2007)have shown that there are significantly more eye movementsduring enumeration of numerosity beyond the subitizing rangecompared with numerosities within the subitizing range. The au-thors also reported that enumeration of up to four items was ac-curate and fast even when participants were not allowed to movetheir eyes. This strategy of individuation with less eye movementscould have been carried over to the continuous task. In contrast,the group that started with the continuous task might have usedmore eye movements in order to assess and sum all surfacesacross dots. This could have been manifested in higher activationof the posterior cingulate. It might appear counter-intuitive thatthe strategy of less eye movements would result in longer RTs,however, it should be noted that the initial visual processing isonly one step in the comparison process. As suggested by Durgin(2008), processing numerosities requires more calculations thanprocessing continuous properties. Thus, even though the visualprocessing might be faster for subitizing, the RT patterns showthat the rest of the comparison process is probably more lengthyand complicated than that for continuous magnitudes. To ourknowledge, there is no work comparing eye movement duringestimation of numerosity and continuous properties. Subse-quently, this suggestion should be examined empirically.

Different scanning strategies might also be the cause for theright middle frontal gyrus (MiFG) activation. The MiFG (BA8) thatincludes the frontal eye field (FEF) was activated in the group thatstarted with the continuous task, and deactivated (compared torest) in the group that started with the numerosity task. Tradi-tionally, the FEF is known to have a role in controlling and mon-itoring eye movements. Importantly, single-cell recordings andfMRI studies converge to suggest that the human FEF region, al-though located in the prefrontal cortex, is part of a cortical net-work of low-level sensory areas. As such, the FEF is able to dif-ferentiate simple stimuli such as 2D shapes during active or pas-sive viewing (Peng et al., 2008) and colors (Münch et al., 2014).ERP studies have revealed that the FEF response to stimuli canoccur very fast (as soon as 50 ms after the onset of the visualstimuli). The time of FEF activation is modulated by the complexitylevel of the stimuli (Kirchner et al., 2009). Kirchner et al. also

raised the possibility that the human brain may have kept, throughspecies evolution, a specialized system to process some aspects ofsensory information very quickly. In light of these findings, weargue that due to their evolutionary importance (Henik et al.,2012; Leibovich and Henik, 2013), the FEF supports the automaticprocessing of continuous properties. In order to voluntarily attendto numerosities, attention to continuous properties should be in-hibited via deactivation of the FEF (for the role of the FEF in in-hibiting reflexive saccades to exogenous signal see Henik et al.,1994). This pattern fits with the behavioral interaction of task,congruity and order (i.e., while the continuous dimension alwaysinterferes with performance, numerosity interferes only whenbeing prompted), and with our suggestion that continuous prop-erties are processed more automatically than numerosities, evenin the subitizing range.

4.3. The effect of the relevant dimension on brain activity

We contrasted congruent trials in the different tasks for the CNand the NC groups separately. This contrast allowed us to trackbrain activity that was modulated by the allocation of attention toa specific dimension. Both contrasts revealed activity in the pre-central and the superior frontal gyri, albeit with a different pat-tern. In the CN group, the left precentral and superior frontal gyriwere more active in the continuous task than in the numerositytask. In contrast, in the NC group, the right precentral and superiorfrontal gyri were more active in the numerosity task. These areas,as well as other fronto-parietal areas that were found here (seeTable 2), were previously found to be involved in numerical cog-nition tasks (for review see Ansari (2012)). The right precentralgyrus was found to be modulated by the distance between two to-be-compared symbolic numbers in adults (Ansari et al., 2005). Theleft precentral gyrus and the left superior frontal gyrus were foundto be more active in 4-year-old children when they were adaptedto numerosities compared with shapes. The right superior frontalgyrus was more active in response to shape dishabituation, whichis also a basic visual feature (Cantlon et al., 2006).

Is there a lateralization of numerical abilities? In a variety ofmethods (lesion studies, transcranial magnetic stimuli (TMS), po-sitron-emission tomography (PET), fMRI), it has been shown thatfor non-symbolic numerosities, both the right and left fronto-parietal cortices play a role. However, the left parietal lobe is moreassociated with symbolic number processing (Andres et al., 2005;Ashkenazi et al., 2008; Cantlon et al., 2006; Cohen Kadosh et al.,2005). In almost all studies that investigated non-symbolic nu-merical processing, the relevant dimension was numerosity, andparticipants were able to use continuous magnitudes to solve thetask (for more details on such studies see Gebuis and Reynvoet(2012b, 2013); Leibovich and Henik, 2013). This brings to mind thepossibility that while both lobes are engaged when processingnon-symbolic numerosities, the left fronto-parietal region mightprocess the continuous aspect and the right fronto-parietal regionmight process the numerosity of the non-symbolic stimuli. Sincethis is the first functional study that presented the same non-symbolic stimuli and asked subjects to voluntarily attend to nu-merosity or continuous properties, and considered the order of thetasks, more empirical evidence is required to confirm thishypothesis.

In addition to these areas, the right inferior parietal lobe (r.IPL)was fond only in the NC group. The r.IPL was found to be anato-mically different in a 22q11.2 deletion syndrome ― a syndrome inwhich patients have lower abilities in mathematics (Simon et al.,2005). It was also found to be involved in more complex ar-ithmetical calculations (Fehr et al., 2007). Dehaene and Cohen(1997) described a patient with a right inferior parietal lesion whowas able to read symbolic numerals and write them but could not

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

do calculations. From such studies, it is clear that the r.IPL plays animportant role in processing numerosities (maybe in order tomanipulate them to be able to perform calculations). However, thisregion was only active when the notion of numerosity wasprompted by being the first task. This finding is in line with ourhypothesis that although some brain regions might be more spe-cific to numerosities than to continuous properties, they are con-text-dependent and not automatic.

4.4. The effect of congruity on brain activity

Contrasting only incongruent trials was aimed at revealingareas that were involved in attending to and processing of therelevant dimension. Because we hypothesized that task order mayaffect performance and brain activity, we conducted this analysisfor each group separately and also directly contrasted task andorder.

Some areas that were found in the current contrast were alsofound in previous contrasts; the r.MiFG and the PoG were found tobe related to the main effect of order (see Section 4.2), and in-volved in the general direction of attention according to the firsttask. The activity in the PrG and SFG (bilaterally) was modulatedby the relevant dimension (see Section 4.3).

The left angular gyrus (l.AnG) activity was modulated by bothtask order and the relevant dimension: namely, in the NC groupthe activity was higher in the numerosity task, and in the CNgroup the activity was higher in the continuous task. The activityof this area was not modulated by task within each group. Instead,interaction stemmed from an opposite activity pattern in the twogroups.

The anterior-lateral aspect of the right intraparietal sulcus (IPS)was modulated by the context determined according to the firsttask; in the NC group, the r.IPS was more active in the numerositytask, and in the CN group, the r.IPS was more active in the con-tinuous task. Importantly, this area was not active in the maineffect of order, or when comparing only congruent trials in thedifferent tasks. This suggests that the activity of the r.IPS is notselective to a specific aspect of magnitude; instead, it has a role indirecting attention according to the relevant dimension, and it isaffected by task context (i.e., by the first task). Similar findings,where IPS activity was found to be modulated by cues that de-termined a task’s relevant dimension, were found in several stu-dies (e.g., Ruge et al., 2009; Schultz and Lennert, 2009; Waskomet al., 2014).

Ventral areas ― the right inferior orbital gyrus (r.IOG), the rightinferior temporal gyrus (r.ITG) and the right occipital gyrus (r.OCG)― were also more active in the first task, regardless of the nature ofthe task. These areas were not significant in the previous contrasts,suggesting that they might be specific to the incongruent nature ofthe trials.

5. Conclusion

Our results indicated that the context of the task determinesboth behavioral performance and the neuronal routes in which thestimuli are processed. The pattern of brain activity was de-termined by the first relevant dimension and carried over to thenext task. Continuous magnitudes processing relied on activationof the FEF and POG, associated with eye movements, and proces-sing of basic visual features. Processing of numerosities, on theother hand, relied on deactivation of these areas, suggesting activesuppression of the continuous dimension. Our experimental de-sign allowed us to dissociate continuous magnitudes and numer-osity processing; whereas previous studies usually reported bi-lateral activity associated with non-symbolic numerosity

processing, we demonstrated lateralized activity that was asso-ciated with task context and the relevant dimension. Starting withthe numerosity task led to enhanced activity in the right hemi-sphere, while starting with the continuous task led to enhancedleft hemisphere activity. Further research is required to examinethe mechanisms underlying this lateralization. Moreover, our ex-perimental design allowed us to demonstrate that parietal acti-vations that were previously associated with magnitude proces-sing are in fact related to task context. This result does not supportthe existence of a dedicated brain circuitry for the processing ofnumerosities (e.g., Burr and Ross, 2008; Dehaene and Changeux,1993; Harvey et al., 2013; Piazza, 2010; Piazza et al., 2013).

Acknowledgments

This work was supported by Q5the European Research Council(ERC) under the European Union's Seventh Framework Programme(FP7/2007–2013)/ERC Grant Agreement 295644 to AH.

References

Andres, M., Seron, X., Olivier, E., 2005. Hemispheric lateralization of numbercomparison. Cogn. Brain Res. 25 (1), 283–290. http://dx.doi.org/10.1016/j.cogbrainres.2005.06.002.

Ansari, D., 2012. Culture and education: new frontiers in brain plasticity. TrendsCogn. Sci. 16 (2), 93–95. http://dx.doi.org/10.1016/j.tics.2011.11.016.

Ansari, D., Garcia, N., Lucas, E., Hamon, K., Dhital, B., 2005. Neural correlates ofsymbolic number processing in children and adults. Neuroreport 16 (16),1769–1773.

Ashkenazi, S., Henik, A., Ifergane, G., Shelef, I., 2008. Basic numerical processing inleft intraparietal sulcus (IPS) acalculia. Cortex 44 (4), 439–448. http://dx.doi.org/10.1016/j.cortex.2007.08.008.

Barth, H.C., 2008. Judgments of discrete and continuous quantity: an illusory Stroopeffect. Cognition 109 (2), 251–266. http://dx.doi.org/10.1016/j.cognition.2008.09.002.

Burr, D., Ross, J., 2008. A visual sense of number. Curr. Biol. 18 (6), 425–428. http://dx.doi.org/10.1016/j.cub.2008.02.052.

Burr, D.C., Turi, M., Anobile, G., 2010. Subitizing but not estimation of numerosityrequires attentional resources. J. Vis. 10 (6), 20. http://dx.doi.org/10.1167/10.6.20.

Cantlon, J.F., Brannon, E.M., Carter, E.J., Pelphrey, K.A., 2006. Functional imaging ofnumerical processing in adults and 4-y-old children. PLoS Biol. 4 (5), e125. http://dx.doi.org/10.1371/journal.pbio.0040125.

Cantlon, J.F., Libertus, M.E., Pinel, P., Dehaene, S., Brannon, E.M., Pelphrey, K.A.,2009. The neural development of an abstract concept of number. J. Cogn.Neurosci. 21 (11), 2217–2229. http://dx.doi.org/10.1162/jocn.2008.21159.

Cantlon, J.F., Platt, M.L., Brannon, E.M., 2009. Beyond the number domain. TrendsCogn. Sci. 13 (2), 83–91. http://dx.doi.org/10.1016/j.tics.2008.11.007.

Cantrell, L., Smith, L.B., 2013. Open questions and a proposal: a critical review of theevidence on infant numerical abilities. Cognition 128 (3), 331–352. http://dx.doi.org/10.1016/j.cognition.2013.04.008.

Castelli, F., Glaser, D.E., Butterworth, B., 2006. Discrete and analogue quantityprocessing in the parietal lobe: a functional MRI study. Proc. Natl. Acad. Sci. USA103 (12), 4693–4698. http://dx.doi.org/10.1073/pnas.0600444103.

Chassy, P., Grodd, W., 2012. Comparison of quantities: core and format-dependentregions as revealed by fMRI. Cerebr. Cortex 22 (6), 1420–1430. http://dx.doi.org/10.1093/cercor/bhr219.

Cohen Kadosh, R., Henik, A., Rubinsten, O., Mohr, H., Dori, H., van de Ven, V., et al.,2005. Are numbers special? The comparison systems of the human brain in-vestigated by fMRI. Neuropsychologia 43 (9), 1238–1248. http://dx.doi.org/10.1016/j.neuropsychologia.2004.12.017.

Coubart, A., Izard, V., Spelke, E.S., Marie, J., Streri, A., 2014. Dissociation betweensmall and large numerosities in newborn infants. Dev. Sci. 17 (1), 11–22. http://dx.doi.org/10.1111/desc.12108.

Dehaene, S., 1997. The Number Sense: How the Mind Creates Mathematics. OxfordUniversity Press, New York.

Dehaene, S., Changeux, J.P., 1993. Development of elementary numerical abilities: aneuronal model. J. Cogn. Neurosci. 5 (4), 390–407. http://dx.doi.org/10.1162/jocn.1993.5.4.390.

Dehaene, S., Cohen, L., 1997. Cerebral pathways for calculation: double dissociationbetween rote verbal and quantitative knowledge of arithmetic. Cortex 33 (2),219–250. http://dx.doi.org/10.1016/S0010-9452(08)70002-9.

Dehaene, S., Piazza, M., Pinel, P., Cohen, L., 2003. Three parietal circuits for numberprocessing. Cogn. Neuropsychol. 20 (3), 487–506. http://dx.doi.org/10.1080/02643290244000239.

Durgin, F.H., 2008. Texture density adaptation and visual number revisited. Curr.

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Biol. 18 (18), R855–R856. http://dx.doi.org/10.1016/j.cub.2008.07.053.Fehr, T., Code, C., Herrmann, M., 2007. Common brain regions underlying different

arithmetic operations as revealed by conjunct fMRI-BOLD activation. Brain Res.1172, 93–102. http://dx.doi.org/10.1016/j.brainres.2007.07.043.

Feigenson, L., Dehaene, S., Spelke, E., 2004. Core systems of number. Trends Cogn.Sci. 8 (7), 307–314. http://dx.doi.org/10.1016/j.tics.2004.05.002.

Fias, W., Lammertyn, J., Reynvoet, B., Dupont, P., Orban, G.A., 2003. Parietal re-presentation of symbolic and nonsymbolic magnitude. J. Cogn. Neurosci. 15 (1),47–56. http://dx.doi.org/10.1162/089892903321107819.

Forman, S.D., Cohen, J.D., Fitzgerald, M., Eddy, W.F., Mintun, M.A., Noll, D.C., 1995.Improved assessment of significant activation in functional magnetic resonanceimaging (fMRI): Use of a cluster-size threshold. Magn. Reson. Med. 33 (5),636–647. http://dx.doi.org/10.1002/mrm.1910330508.

Fransson, P., Marrelec, G., 2008. The precuneus/posterior cingulate cortex plays apivotal role in the default mode network: evidence from a partial correlationnetwork analysis. NeuroImage 42 (3), 1178–1184. http://dx.doi.org/10.1016/j.neuroimage.2008.05.059.

Friston, K.J., Josephs, O., Rees, G., Turner, R., 1998. Nonlinear event-related re-sponses in fMRI. Magn. Reson. Med. 39 (1), 41–52. http://dx.doi.org/10.1002/mrm.1910390109.

Gebuis, T., Gevers, W., Cohen Kadosh, R., 2014. Topographic representation of high-level cognition: numerosity or sensory processing? Trends Cogn. Sci. 18 (1),1–3. http://dx.doi.org/10.1016/j.tics.2013.10.002.

Gebuis, T., Reynvoet, B., 2011. Generating nonsymbolic number stimuli. Behav. Res.Methods 43 (4), 981–986. http://dx.doi.org/10.3758/s13428-011-0097-5.

Gebuis, T., Reynvoet, B., 2012a. he interplay between nonsymbolic number and itscontinuous visual properties. J. Exp. Psychol. Gen. 141 (4), 642–648. http://dx.doi.org/10.1037/a0026218.

Gebuis, T., Reynvoet, B., 2012b. The role of visual information in numerosity esti-mation. PLoS One 7 (5), e37426. http://dx.doi.org/10.1371/journal.pone.0037426.

Gebuis, T., Reynvoet, B., 2013. The neural mechanisms underlying passive and activeprocessing of numerosity. NeuroImage 70, 301–307. http://dx.doi.org/10.1016/j.neuroimage.2012.12.048.

Halberda, J., Mazzocco, M.M., Feigenson, L., 2008. Individual differences in non-verbal number acuity correlate with maths achievement. Nature 455 (7213),665–668. http://dx.doi.org/10.1038/nature07246.

Harvey, B.M., Klein, B.P., Petridou, N., Dumoulin, S.O., 2013. Topographic re-presentation of numerosity in the human parietal cortex. Science 341 (6150),1123–1126. http://dx.doi.org/10.1126/science.1239052.

Henik, A., Leibovich, T., Naparstek, S., Diesendruck, L., Rubinsten, O., 2012. Quan-tities, amounts, and the numerical core system. Front. Hum. Neurosci. 5 (186).http://dx.doi.org/10.3389/fnhum.2011.00186.

Henik, A., Rafal, R., Rhodes, D., 1994. Endogenously generated and visually guidedsaccades after lesions of the human frontal eye fields. J. Cogn. Neurosci. 6 (4),400–411. http://dx.doi.org/10.1162/jocn.1994.6.4.400.

Hyde, D.C., 2011. Two systems of non-symbolic numerical cognition. Front. Hum.Neurosci. 5, 150. http://dx.doi.org/10.3389/fnhum.2011.00150.

Kirchner, H., Barbeau, E.J., Thorpe, S.J., Régis, J., Liégeois-Chauvel, C., 2009. Ultra-rapid sensory responses in the human frontal eye field region. J. Neurosci. 29(23), 7599–7606. http://dx.doi.org/10.1523/JNEUROSCI.1233-09.2009.

Knops, A., Piazza, M., Sengupta, R., Eger, E., Melcher, D., 2014. A shared, flexibleneural map architecture reflects capacity limits in both visual short-termmemory and enumeration. J. Neurosci. 34 (30), 9857–9866. http://dx.doi.org/10.1523/JNEUROSCI.2758-13.2014.

Leibovich, T., Henik, A., 2013. Magnitude processing in non-symbolic stimuli. Front.Psychol. 4, 375. http://dx.doi.org/10.3389/fpsyg.2013.00375.

Leibovich, T., Henik, A., 2014. Comparing performance in discrete and continuouscomparison tasks. Q. J. Exp. Psychol. 67 (5), 899–917. http://dx.doi.org/10.1080/17470218.2013.837940.

Leroux, G., Spiess, J., Zago, L., Rossi, S., Lubin, A., Turbelin, M.-R., Joliot, M., 2009.Adult brains don't fully overcome biases that lead to incorrect performanceduring cognitive development: an fMRI study in young adults completing aPiaget-like task. Dev. Sci. 12 (2), 326–638. http://dx.doi.org/10.1111/j.1467-7687.2008.00785.x.

McComb, K., Packer, C., Pusey, A., 1994. Roaring and numerical assessment incontests between groups of female lions, Panthera leo. Anim. Behav. 47 (2),379–387. http://dx.doi.org/10.1006/anbe.1994.1052.

Mix, K.S., Huttenlocher, J., Levine, S.C., 2002. Multiple cues for quantification ininfancy: is number one of them? Psychol. Bull. 128 (2), 278–294. http://dx.doi.org/10.1037/0033-2909.128.2.278.

Moyer, R.S., Landauer, T.K., 1967. Time required for judgements of numerical in-equality. Nature 215 (5109), 1519–1520. http://dx.doi.org/10.1038/2151519a0.

Münch, M., Plomp, G., Thunell, E., Kawasaki, A., Scartezzini, J.L., Herzog, M.H., 2014.Different colors of light lead to different adaptation and activation as

determined by high-density EEG. NeuroImage 101, 547–554. http://dx.doi.org/10.1016/j.neuroimage.2014.06.071.

Mussolin, C., Mejias, S., Noël, M.-P., 2010. Symbolic and nonsymbolic numbercomparison in children with and without dyscalculia. Cognition 115 (1), 10–25.http://dx.doi.org/10.1016/j.cognition.2009.10.006.

Nieder, A., Dehaene, S., 2009. Representation of number in the brain. Ann. Rev.Neurosci. 32, 185–208. http://dx.doi.org/10.1146/annurev.neuro.051508.135550.

O’Craven, K.M., Downing, P.E., Kanwisher, N., 1999. fMRI evidence for objects as theunits of attentional selection. Nature 401 (6753), 584–587. http://dx.doi.org/10.1038/44134.

Pagano, S., Lombardi, L., Mazza, V., 2014. Brain dynamics of attention and workingmemory engagement in subitizing. Brain Res. 1543, 244–252. http://dx.doi.org/10.1016/j.brainres.2013.11.025.

Peng, X., Sereno, M.E., Silva, A.K., Lehky, S.R., Sereno, A.B., 2008. Shape selectivity inprimate frontal eye field. J. Neurophysiol. 100 (2), 796–814. http://dx.doi.org/10.1152/jn.01188.2007.

Piazza, M., 2010. Neurocognitive start-up tools for symbolic number representa-tions. Trends Cogn. Sci. 14 (12), 542–551.

Piazza, M., Fumarola, A., Chinello, A., Melcher, D., 2011. Subitizing reflects visuo-spatial object individuation capacity. Cognition 121 (1), 147–153. http://dx.doi.org/10.1016/j.cognition.2011.05.007.

Piazza, M., Izard, V., 2009. How humans count: numerosity and the parietal cortex.Neurosci. 15 (3), 261–273. http://dx.doi.org/10.1177/1073858409333073.

Piazza, M., Pica, P., Izard, V., Spelke, E.S., Dehaene, S., 2013. Education enhances theacuity of the nonverbal approximate number system. Psychol. Sci. 24 (6),1037–1043. http://dx.doi.org/10.1177/0956797612464057.

Piazza, M., Pinel, P., Le Bihan, D., Dehaene, S., 2007. A magnitude code common tonumerosities and number symbols in human intraparietal cortex. Neuron 53(2), 293–305. http://dx.doi.org/10.1016/j.neuron.2006.11.022.

Pinel, P., Piazza, M., Le Bihan, D., Dehaene, S., 2004. Distributed and overlappingcerebral representations of number, size, and luminance during comparativejudgments. Neuron 41 (6), 983–993.

Pisa, P.E., Agrillo, C., 2008. Quantity discrimination in felines: a preliminary in-vestigation of the domestic cat (Felis silvestris catus). J. Ethol. 27 (2), 289–293.http://dx.doi.org/10.1007/s10164-008-0121-0.

Revkin, S.K., Piazza, M., Izard, V., Cohen, L., Dehaene, S., 2008. Does subitizing reflectnumerical estimation? Psychol. Sci. 19 (6), 607–614. http://dx.doi.org/10.1111/j.1467-9280.2008.02130.x.

Ruge, H., Braver, T., Meiran, N., 2009. Attention, intention, and strategy in pre-paratory control. Neuropsychologia 47 (7), 1670–1685. http://dx.doi.org/10.1016/j.neuropsychologia.2009.02.004.

Schultz, J., Lennert, T., 2009. BOLD signal in intraparietal sulcus covaries withmagnitude of implicitly driven attention shifts. NeuroImage 45 (4), 1314–1328.http://dx.doi.org/10.1016/j.neuroimage.2009.01.012.

Simon, T.J., Ding, L., Bish, J.P., McDonald-McGinn, D.M., Zackai, E.H., Gee, J., 2005.Volumetric, connective, and morphologic changes in the brains of children withchromosome 22q11.2 deletion syndrome: an integrative study. NeuroImage 25(1), 169–180. http://dx.doi.org/10.1016/j.neuroimage.2004.11.018.

Small, D., Gitelman, D., Gregory, M., Nobre, A., Parrish, T., Mesulam, M.M., 2003. Theposterior cingulate and medial prefrontal cortex mediate the anticipatory al-location of spatial attention. NeuroImage 18 (3), 633–641. http://dx.doi.org/10.1016/S1053-8119(02)00012-5.

Talairach, J., Tournoux, P., 1988. Co-planar Stereotaxic Atlas of the Human Brain. 3-Dimensional Proportional System: an Approach to Cerebral Imaging. ThiemeMedical Publishers, New York, Retrieved fromhttp://www.citeulike.org/group/96/article/745727.

Trick, L.M., Pylyshyn, Z.W., 1994. Why are small and large numbers enumerateddifferently? A limited-capacity preattentive stage in vision. Psychol. Rev. 101(1), 80–102. http://dx.doi.org/10.1037/0033-295X.101.1.80.

Vogt, B.A., Finch, D.M., Olson, C.R., 1992. Functional heterogeneity in cingulatecortex: the anterior executive and posterior evaluative regions. Cereb. Cortex 2(6), 435–443. http://dx.doi.org/10.1093/cercor/2.6.435-a.

Waskom, M.L., Kumaran, D., Gordon, A.M., Rissman, J., Wagner, A.D., 2014. Fron-toparietal representations of task context support the flexible control of goal-directed cognition. J. Neurosci. 34 (32), 10743–10755. http://dx.doi.org/10.1523/JNEUROSCI.5282-13.2014.

Watson, D.G., Maylor, E.A., Bruce, L.A.M., 2007. The role of eye movements insubitizing and counting. J. Exp. Psychol.: Hum. Percept. Perform. 33 (6),1389–1399. http://dx.doi.org/10.1037/0096-1523.33.6.1389.

Xu, F., Spelke, E.S., 2000. Large number discrimination in 6-month-old infants.Cognition 74 (1), B1–B11. http://dx.doi.org/10.1016/S0010-0277(99)00066-9.

Yee, E., Ahmed, S.Z., Thompson-Schill, S.L., 2012. Colorless green ideas (can) primefuriously. Psychol. Sci. 23 (4), 364–369. http://dx.doi.org/10.1177/0956797611430691.

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Numerosity processing is context driven even in the subitizing range: an fMRI study

Documents

Transcript of Numerosity processing is context driven even in the subitizing range: an fMRI study

Lattice Independent Component Analysis for fMRI Analysis

analisis break even point sebagai alat perencanaan laba

Triangulated neighborhoods in even-hole-free graphs

Dissociations and interactions between time, numerosity and space processing

FMRI Study of Emotional Speech Comprehension - CiteSeerX

Dynamical Cluster Analysis of Cortical fMRI Activation

Dissociation between numerosity and duration processing in aging and early Parkinson's disease

AAG DIGITAL EVEN Oral Presentations - Australian ...

Fingersomatotopy in area 3b: an fMRI-study

New mixing plant for even better architectural precast concrete ...

31 – 05 - 2012 Practicing English language and…even more

ANALISIS BREAK EVEN POINT USAHA PEMBIBITAN KAKAO ...

2022 Catalogue | Even Spray

Duchamp : love and death, even - Monoskop

A Model-Based fMRI Analysis With Hierarchical Bayesian ...

V112-3.0 MW Offshore Making offshore wind even more ...

Network modelling methods for FMRI

Torching State Crowdfunding for Startups Without Even Trying

Q3 G1&2 Mother Tongue (Ilokano) Multigrade Outline (Even ...

Language lateralization in schizophrenia, an fMRI study