The SNARC (spatial numerical associations of response codes) effect reflects the tendency to respond faster with the left hand to relatively small numbers and with the right hand to relatively large numbers (S. Dehaene, S. Bossini, & P. Giraux, 1993). Using computational modeling, the present article aims to provide a framework for conceptualizing the SNARC effect. In line with models of spatial stimulus-response congruency, the authors modeled the SNARC effect as the result of parallel activation of preexisting links between magnitude and spatial representation and short-term links created on the basis of task instructions. This basic dual-route model simulated all characteristics associated with the SNARC effect. In addition, 2 experiments tested and confirmed new predictions derived from the model.

The anterior intraparietal sulcus, and more specifically its horizontal segment (hIPS), is known to play a crucial role in the cognitive representation of numerical quantity. Whether the involvement of hIPS is restricted to the processing of numerical information or generalizes to non-numerical ordinal dimensions remains an open question. Using functional magnetic resonance imaging during comparison tasks, we demonstrate that the hIPS is equally responsive to numbers and letters, indicating that hIPS is also involved in the representation and processing of non-numerical ordinal series. This extends the numerical processing function of IPS into the realm of abstract knowledge processing.

Until a few years ago, perception of visual stimuli and selection of motor responses were considered as autonomous processing stages. For instance, Kahneman and Treisman (1984) suggested that the information about a visual stimulus in an experimental task is integrated into an episodic object file, which contains the visual features of the object, together with information about the location of the stimulus. Only when processing in the perceptual stage is finished is the information translated into specific response codes. In recent years, however, evidence has accumulated that object files contain not only perceptual information, but also response-related information. Treisman (1992) argued that repeated experience with objects leads to an integration of stimulus features with task or response information, and Danzinger and Robertson (1994) showed that the benefits of stimulus feature repetition depended on a consistent stimulus-response mapping. Probably the strongest antimodularity view of stimulusresponse associationsis the theory of event coding proposed by Hommel (1998aHommel ( , 1998b. According to this theory, an experimentaltask does not result in the creation of episodic object files but in the creation of episodic event files that integrate information about both the stimulus features and the accompanying response codes, so that response codes are immediately and automatically activated. Dehaene et al. (1998) reported an fMRI study that seems to agree more with Hommel's (1998aHommel's ( , 1998b theory than with the traditional modularity view. Participants had to judge whether a number was larger or smaller than 5 by pressing a key with the left or the right hand. Before each target, a prime was presented that in half of the trials was associated with the same response (i.e., both smaller or larger than 5) and in half of the trials with a different response (e.g., prime smaller than 5 and target larger than 5). The data showed a strong response compatibilityeffect (RCE): Responses to targets were faster when the primes were associated with the same response than when the primes asked for a different response. Brain-imaging data showed that the difference between compatible and incompatible trials was not limited to the visual and the association areas but could be traced well into the motor cortex, indicating that the primes had activated their accompanying response code. This is the more interesting because the authors presented the primes tachistoscopically immediately before the target, so that the participants could not use deliberate guessing strategies, and the response activation thus had to be an automatic process (see Neely, 1991, for a distinctionbetween automatic and strategic processes in priming).In this paper, we first describe a numerical task that evoked the same pattern of behavioral results as Dehaene et al. 's (1998), and then we show that the RCE we observed was not based solely on a learned association between stimulus-specific physical features and response codes but stemmed from...

No abstract

Today, it is generally accepted that unconscious stimuli can activate a response code, which leads to a response congruency effect (RCE) on a subsequent target. However, it is not yet clear whether this is due to the semantic processing of the primes or to the formation of direct stimulus-response (S-R) associations bypassing the semantic system. Recently, it was shown that even novel primes, for which no direct S-R links exist, can also evoke an RCE that is in line with the activation of response codes through semantics. In these experiments, the authors examined 3 alternatives for this RCE from novel primes and report a novel effect in unconscious priming. First, the authors show that this effect is not limited to a small set of numerical stimuli but also extends to letter stimuli (Experiments 1-3). Second, the authors show that the RCE is not a side effect of the prime-target distance effect, as has been reported before (Experiments 1-2). Third, the authors found that, for RCE to occur, overlap at the motor level but not at the semantic level was crucial (Experiments 2-3). Finally, in addition, the results showed a category match priming effect independent of RCE. This last result is evidence that novel unconscious primes activate their semantic category prior to the target and might be considered a good marker for semantic processing.