The theory of magnitude (ATOM), posits that time, space, and number are represented by a common magnitude system in which quantities are linked via a common metric (Walsh, 2003). This idea is supported by studies showing similarities in the behavioural signatures of the representations of different types of magnitudes. As for numerical magnitude, it is known that small and large numbers are represented spatially along a left-to-right mental continuum. Accordingly, participants are faster to respond to small numbers with the left hand and to large numbers with the right hand (SNARC; Dehaene et al., 1993). Moreover, cueing attention to the left or to the right impairs the processing of large and small numbers, respectively, suggesting a functional involvement of space for the processing of numerical magnitude (Stoianov et al., 2008). Recent publications revealed with SNARC-like paradigms various associations between space and other domains such as size, luminance, pitch, size, risk and even emotions (MacNamara et al., 2018). Angry and happy faces are responded faster with the left and right hand, respectively (e.g., Holmes & Lourenco, 2011). These studies concluded that ATOM should be extended to these domains, without considering the possibility that such associations might instead be mere response biases made salient by the experimental setup. First, we replicated the observation that cueing attention impacted on a task a numerical comparison task (i.e., more/less than five). We then showed that orienting the attention to the left or right had no impact on a task in which participants had to compare emotional faces (i.e., more/less happy than a neutral face). Altogether, our results support a functional mechanism for space in number processing but not for emotions. We conclude that the observation of SNARC-like effects should not be overinterpreted as the evidence for a functional and generalized spatial representation.