The more numerous the longer: how the integration between numerosity and time leads to a common neural response

Abstract

If you are stuck in a traffic jam, the more numerous the queuing cars are, the longer you expect to wait. Time and numerosity are stimulus dimensions often associated in the same percept and whose interaction can lead to misjudgements. At brain level it is unclear to which extent time and numerosity recruit same/different neural populations and how their perceptual integration leads to changes in these populations' responses. Here we used high-spatial-resolution functional magnetic resonance imaging with neural model-based analyses to investigate how the topographic representations of numerosity and time change when these dimensions are varied together on the same visual stimulus in a congruent (the more numerous the items, the longer the display time) or incongruent manner. Compared to baseline conditions, where only one dimension was changed at a time, the variation of both stimulus dimensions led to changes in neural population responses that became more sensitive either to the two features or to one of them. Magnitude integration led also to degradation of topographies and shifts in response preferences. These changes were more pronounced in the comparison between parietal and frontal maps. Our results while pointing to partially distinct representations of time and numerosity show a common neural response to magnitude integration.

Magnitude processing and integration entail perceptual processes independent from the task

“…Instead, the results seem more in line with recent findings of topographical cortical maps of different magnitudes, overlapping with each other (Y. Cai et al, 2021; Fortunato et al, 2023; Harvey et al, 2013, 2015; Harvey & Dumoulin, 2017; Hendrikx et al, 2024; Protopapa et al, 2019). Recently, it has indeed been proposed that the interaction between different magnitudes could arise from the overlap of neural populations sensitive to different dimensions (Hendrikx et al, 2024; Tsouli et al, 2022), without the need for a centralised mechanism or a common magnitude neural code (Walsh, 2003).…”

“…While evidence has now been accumulated against a linguistic/conceptual view of magnitude integration (Cai & Connell, 2015;Whitaker et al, 2022), other theories have proposed that magnitudes interact at a more cognitive rather than perceptual level, as a working memory interference (Z. G. Cai et al, 2018), or as a response bias (Yates et al, 2012). Finally, more in line with a perceptual interpretation of the integration effect, it has been recently proposed (Hendrikx et al, 2024;Tsouli et al, 2022) that the interaction could arise from the processing of different dimensions in partially overlapping cortical maps (Fortunato et al, 2023;Harvey et al, 2013Harvey et al, , 2015Hendrikx et al, 2022Hendrikx et al, , 2024Protopapa et al, 2019), but without involving a common neural code.…”

The link between space and time along the human cortical hierarchy

“…Compared to baseline conditions where only a single stimulus dimension was manipulated, response functions became more sensitive to both duration and numerosity and their preferences changed. Interest-ingly, preference changes were more pronounced in parietal and frontal regions [35]. Parietal cortex is the brain area where different kinds of integration processes are known to happen, for example multisensory [36] and visuo-motor [37], and where different visual magnitudes like stimulus size, duration and numerosity are processed.…”

“…In some ROIs the eccentricity map was better captured by two high borders instead of a low and a high border. To place borders, two main criteria were applied[35]: map continuity (i.e., map’s vertices should belong to the same spatial cluster) and progression continuity (i.e., map’s vertices should be arranged following one eccentricity gradient only). These ROIs were used to investigate the properties of BOLD responses when position and duration co-vary in the stimulus (see Properties of GST model ) and to study spatial and duration topographies (see Spatial relationship between eccentricity and duration maps and Spatial properties of eccentricity and duration maps ).…”

Temporal speed prevails on interval duration in the SNARC-like effect for tempo

“…Furthermore, previous literature did not discuss a possible confound between tempo and numerosity of auditory beats (for recent evidence on common mechanisms for numerosity and temporal processing, see Fortunato et al, 2023 ). Indeed, when participants judge fast tempo stimuli, they are exposed to a larger number of auditory beats compared to when they judge slow tempo stimuli.…”