Brief articleAttentional SNARC: There’s something special about numbers (let us count the ways)☆
Introduction
As we navigate our visual world, we are continually confronted with more information than we can process simultaneously. Consequently, a critical function of our visual system is to efficiently direct attention to features of our environment to determine which items are to be processed and which are to be ignored. Attention is often said to shift throughout the environment in one of two ways: endogenously or exogenously. Endogenous shifts of attention are those that occur in a volitional (top-down) manner, such as when one scans a crowd looking for a friend. In the laboratory, endogenous attention is often studied by presenting a central cue, such as an arrow, that indicates where a target is likely to appear. That target detection is speeded at cued relative to uncued locations is taken as evidence that participants shifted attention voluntarily based on the cue’s meaning (Posner, 1980, Ristic and Kingstone, 2006). Exogenous shifts of attention are reflexive (bottom-up), and in the laboratory they are often studied by presenting a sudden spatially nonpredictive visual cue, such as a light pulse, in the periphery (e.g., Posner, 1980, Yantis and Hillstrom, 1994). Target detection is speeded at cued relative to uncued locations when the target appears immediately after the cue, suggesting that spatial attention was pulled automatically to the cued location.
While the aforementioned paradigms are commonly used to study endogenous and exogenous visual spatial attention, it has also been established that attention can be influenced by the presentation of overlearned, spatially meaningful, symbols at fixation, even when these symbols do not predict the upcoming target location. For example, Hommel, Pratt, Colzato, and Godijn (2001; see also Eimer, 1997, Pratt and Hommel, 2003) demonstrated that the presentation of a spatially nonpredictive arrow or directional word (e.g., “left”) results in targets being detected more quickly at the location consistent with the cue’s directional meaning. That a cuing effect occurred for these spatially nonpredictive cues suggests that attention was shifted reflexively in the direction of these cues. Of course, given that the presentation of an arrow in the real world is almost always spatially predictive and meaningful makes findings such as the above relatively intuitive (see also Gibson and Kingstone, 2006, Ristic et al., 2002, Tipples, 2002).
Less intuitive, however, is the finding that the presentation of numbers at fixation also seems to influence the allocation of attention in the visual field. Dehaene, Bossini, and Giraux (1993) had participants indicate whether a number was odd or even with a left- or right-hand keypress. Participants were faster to respond to low odd digits (e.g., 1) relative to high odd digits (e.g., 9) with their left hand and were faster to respond to high even digits (e.g., 8) relative to low even digits (e.g., 2) with their right hand. Dehaene et al. posited that this finding was attributable to the mental organization of numbers, which may be stored in a mental number line running from left to right—with low digits occupying left space and high digits occupying right space. Accordingly this was coined the SNARC effect (Spatial Numerical Association of Response Codes). SNARC effects have subsequently been reported in a variety of tasks, such as phoneme detection of digits’ names (Fias, Brysbaert, Geypens, & d’Ydewalle, 1996), digit magnitude classification (Bächtold, Baumüller, & Brugger, 1998), and even for the midpoint localization of long digit strings (Fischer, 2001). These findings led Fischer, Castel, Dodd, and Pratt (2003) to ask whether the presentation of an irrelevant digit at fixation could evoke an attentional shift to the left or right visual field. Consistent with this idea the presentation of a nonpredictive low digit (e.g., 1 or 2) facilitated target detection on the left while the presentation of a nonpredictive high digit (e.g., 8 or 9) facilitated target detection on the right. This finding is particularly interesting in that numbers, which under some circumstances have spatial meaning in the real world (e.g., spatial coordinates on a map, the left-to-right ordering of numbers on rulers and tape measures, the ascending/descending nature of house addresses on most North American city streets), do not have the same strong spatial connotation that other spatial cues do, such as arrows.
That irrelevant numbers influence the allocation of attention in a target detection task leaves open the question of whether other related stimuli may have a similar effect, or whether the effect observed by Fischer et al. (2003) is specific to numbers. Although it was originally believed that the SNARC effect was restricted to numerical values (Dehaene et al., 1993), it has recently been demonstrated that a SNARC effect is obtained for other ordinal stimuli such as letters of the alphabet, days of the week, and months of the year (Gevers et al., 2003, Gevers et al., 2004).1 In these studies participants were presented with items at fixation and were required to make an order-relevant decision (e.g. does this month occur before or after July) or order-irrelevant decision (e.g., does this month end in the letter ‘R’). Critically, a SNARC effect was observed in both the order-relevant and order-irrelevant tasks, as participants were faster to respond to left ordinal information (e.g., months occurring before June) when they responded with their left hand relative to their right hand, and faster to respond to right ordinal information (e.g., months occurring after June) when they responded with their right hand relative to their left hand. Gevers et al. suggested that these findings are evidence that the mental representation of ordinal sequences, and not just numbers, is spatially coded. Moreover, the finding that the SNARC effect is observed in a task in which ordinal information is irrelevant was taken as evidence that the spatial component of ordinal sequences is automatically activated.
Given that numbers—much like letters, days, and months—convey ordinal meaning, it is important to determine whether numerical sequences and non-numerical ordered sequences share similar processing mechanisms, or whether the aforementioned findings are separable. For example, while Gevers et al. (2003) provide evidence that non-numerical ordinal sequences activate spatial representations, a recent study by Zorzi, Priftis, Meneghello, Marenzi, and Umilta (2006) with neglect patients suggests that numbers are processed differently than other ordinal sequences. To this end, the present study adopts the attention paradigm used by Fischer et al. (2003) to investigate whether the ordinal sequences of letters, days, and months will also influence the allocation of attention. If an attentional SNARC effect is observed when this range of irrelevant ordinal information is placed at fixation, it would suggest that numerical and non-numerical ordered sequences share similar processing mechanisms. If, on the other hand, an attentional effect is specific to numbers it would suggest that numbers may represent a special class of ordinal information that, in turn, will provide important insights into the processes that underlying both the SNARC and attentional SNARC effects.
Section snippets
Participants
Thirty undergraduate students from the University of British Columbia underwent individual 45-min sessions, receiving course credit as remuneration for participating in the study. All students had normal or corrected-to-normal vision and were naïve about the purpose of the experiment.
Apparatus and procedure
The experiment, programmed in Visual C++, was individually conducted on Pentium IV PC’s with VGA monitors in a testing room equipped with soft lighting and sound-attenuation. Participants were seated approximately
Results and discussion
Errors occurred on less than 2.3% of all trials and these trials were eliminated from all subsequent analyses. Reaction times (RTs) and standard deviations for targets appearing at each target location as a function of cue condition are presented in Table 1. Moreover, Fig. 2 presents the RTs as a function of cue-target congruency. For all four stimulus types (numbers, letters, days, and months), RTs were collapsed for the left and right values (e.g. RTs for targets following ‘a’ and ‘b’ were
Experiment 2
In Experiment 1, we observed an attentional SNARC effect for number stimuli, replicating previous work, but failed to observe such an effect with letters, days, and months.
Results and discussion
Given the fact that the present experiment consisted of two tasks (target detection followed by before/after decision), two types of errors were possible: target detection errors and before/after judgment errors. Target detection errors occurred on less than 1.8% of all trials while before/after judgment errors occurred on less than 0.3% of the trials. All trials in which an error occurred were eliminated from all subsequent analyses. Reaction times (RTs) and standard deviations for targets
General discussion
The purpose of the present study was to determine whether nonpredictive central cues conveying ordinal information would influence the manner in which attention is allocated during a target detection task. Recently, Fischer et al. (2003) extended the earlier work of Dehaene et al. (1993) demonstrating that the presentation of an irrelevant number at fixation influences the manner in which attention is shifted across the visual field: left target detection is facilitated when a low number is
References (23)
- et al.
Stimulus-response compatibility in representational space
Neuropsychologia
(1998) Uninformative symbolic cues may bias visual-spatial attention: Behavioral and electrophysiological evidence
Biological Psychology
(1997)- et al.
The mental representation of ordinal sequences is spatially organized
Cognition
(2003) - et al.
The mental representation of ordinal sequences is spatially organized. Evidence from days of the week
Cortex
(2004) Linguistic and conceptual control of visual spatial attention
Cognitive Psychology
(1995)- et al.
The spatial representation of numerical and non-numerical sequences: Evidence from neglect
Neuropsychologia
(2006) - et al.
The mental representation of parity and number magnitude
Journal of Experimental Psychology: General
(1993) - et al.
The importance of magnitude information in numerical processing: Evidence from the SNARC effect
Mathematical Cognition
(1996) - et al.
Processing of abstract ordinal knowledge in the horizontal segment of the intraparietal sulcus
Journal of Neuroscience
(2007) Number processing induces spatial performance biases
Neurology
(2001)
Perceiving numbers causes spatial shifts of attention
Nature Neuroscience
Cited by (90)
When A is greater than B: Interactions between magnitude and serial order
2022, Consciousness and CognitionThe influence of the location of ordered symbols on the ordinal position effect: The involvement of the task performed
2020, Acta PsychologicaCitation Excerpt :Numerous studies have indicated that the spatial stimulus-response compatibility effect (or the Simon effect) was very strong and could be observed even in more space irrelevant task contexts (Fitts & Seeger, 1953; Kornblum et al., 1990; Simon & Small, 1969; Yamaguchi & Proctor, 2011; Yamaguchi & Proctor, 2019). However, although several studies captured the ordinal position effect in the processing of ordinal symbols in stimulus order irrelevant task contexts (Gevers et al., 2003; Gevers et al., 2004; Wang et al., 2019), Dehaene et al. (1993) did not, nor did Casarotti et al. and Dodd et al. when the ordinal sequence of ordinal symbols was irrelevant to the classification task performed (Casarotti, Michielin, Zorzi, & Umiltà, 2007; Dodd, Stigchel, Leghari, Fung, & Kingstone, 2008). These results imply that the ordinal position effect is weaker than the spatial stimulus-response compatibility effect.
On the genesis of spatial-numerical associations: Evolutionary and cultural factors co-construct the mental number line
2018, Neuroscience and Biobehavioral ReviewsCitation Excerpt :While the SNARC is the most oft-cited metric of spatial-numerical associations in cognition, it is important to note that phenomena other than the classic SNARC demonstrate the widespread influence of SNAs on cognition. For example, the implicit association between numbers and sides of space can induce covert attentional shifts, with attention orienting to the left or right following small or large cues, respectively (Dodd et al., 2008; Fischer et al., 2003; Galfano et al., 2006; Goffaux et al., 2012; Nicholls et al., 2008; Ranzini et al., 2009; Ristic and Kingstone, 2006; Salillas et al., 2008; Schuller et al., 2014; Schwarz and Keus, 2004). This finding has been called the attentional-SNARC (Att-SNARC).
- ☆
This research was supported by a Killam post-doctoral fellowship and a Natural Sciences and Engineering Research Council (NSERC) post-doctoral fellowship to M. Dodd and NSERC grant to A. Kingstone. We would like to thank Wieske van Zoest, Ipek Oruc, Sharon Morein-Zamir, Michael MacIsaac, and two anonymus reviewers for their insight at various stages of this project. We would also like to thank Kristin Divis for assistance in collecting the data.