Summary: New research reveals that numbers in our visual field can subtly distort how we judge spatial positions, showing that perception is shaped by both numerical magnitude and object-based processing. In traditional line-bisection tasks, smaller numbers shifted perceived centers leftward, supporting the idea of a mental number line.
But when shapes became more complex, such as squares, the numerical effect disappeared, replaced by strong upward biases linked to object recognition systems in the brain. These findings show that spatial-numerical association interacts with deeper visual pathways rather than acting alone. The work offers fresh insight into how the brain integrates symbolic and spatial information.
Key Facts
- Number Size Alters Perception: Smaller numbers shifted perceived line centers leftward, while larger numbers shifted them rightward or downward depending on orientation.
- 2D Shapes Break the Pattern: In squares, magnitude effects vanished and an object-driven upward bias emerged instead.
- Object-Based Processing Dominates: Visual recognition pathways can override numerical magnitude signals, reshaping spatial attention.
Source: Tokyo Metropolitan University
Researchers from Tokyo Metropolitan University have studied the relationship between numerical information in our vision, and how it affects our perception of space.
Volunteers were asked to identify the center of lines and squares filled with numbers; how far they were from the true center revealed unexpected biases.
Crucially, their work with squares showed how our perception of space is a complex interplay between “object-based” processing and our processing of numerical information.
Numbers in our vision can affect how we see the world in unexpected ways. Known as spatial-numerical association, the way in which our brain processes numbers can introduce “attentional biases.”
For example, for people from a left-to-right writing culture, a simple game where players press one of two buttons with the lower number on it reveals faster response when the lower number is on the left; the opposite is true using larger numbers.
Sometimes, the information does not need to be numbers, but something that indicates magnitude, like a brighter light, or louder noise. In fact, similar trends have been seen in animals and insects.
This suggests that a “mental number line,” some left-to-right mapping of magnitude-related information onto space, might be a deeply ingrained characteristic in nature. Understanding how these biases work is crucial not only to gain a better idea of how our brains work, but in optimizing how information is presented in the real world.
A team of researchers from Tokyo Metropolitan University led by Professor Masami Ishihara and doctoral student Ryo Hishiya have been using “bisection tasks” to probe spatial-numerical association. Standard bisection tasks ask volunteers to estimate the center of a line or bar.
When the bar is painted over with smaller numbers, they found that volunteers systematically put the center more left than for larger numbers. This confirms the left-to-right “mental number line” found in previous work.
When they tried the same exercise with vertical bars, however, volunteers put their point lower for larger numbers, contrary to the “bottom-to-top” association expected from previous works. There is clearly more at work than just the mental number line.
In a new approach, the team proceeded to repeat the experiment with squares i.e. two-dimensional shapes. Curiously, they found that the effect of number magnitude disappeared.
Instead, the presence of numbers was enough to induce a strong upward bias, and a weak leftward bias; an absence of numbers led to a stronger bias in the horizontal direction, likely due to pseudoneglect, a known natural bias in attention toward the left.
The team propose that this vertical bias reflects the impact of the ventral visual stream, the part of our brain trying to recognize objects (in this case, numerical strings) which also tends to push attention upward. In this case, “object-based” processing seems to show a dominant effect over our processing of the value of numbers.
While future work remains, the team’s experiments offer a new glimpse into the subtle asymmetries that color our view on the world.
Key Questions Answered:
A: They shift perceived centers of lines depending on numerical magnitude, reflecting an internal mental number line.
A: Object-based processing took over, producing strong vertical biases that overpowered numerical effects.
A: The brain integrates symbolic numbers with structural object-processing streams, meaning spatial perception is shaped by multiple interacting systems.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- Journal paper reviewed in full.
- Additional context added by our staff.
About this vision and perception research news
Author: GO TOTSUKAWA
Source: Tokyo Metropolitan University
Contact: GO TOTSUKAWA – Tokyo Metropolitan University
Image: The image is credited to Neuroscience News
Original Research: Open access.
“Numerically induced attentional biases in horizontal, vertical, and two-dimensional shapes” by Masami Ishihara et al. Scientific Reports
Abstract
Numerically induced attentional biases in horizontal, vertical, and two-dimensional shapes
Previous studies have demonstrated that numerical magnitudes can induce attentional biases mainly in a horizontal space.
The present study aimed to clarify spatial-numerical association in horizontal, vertical, and two-dimensional square stimuli composed of visually aligned strings of relatively smaller Arabic numerals (i.e., 1 or 2) or larger numerals (i.e., 8 or 9).
Neurologically and psychiatrically healthy participants, all of whom were Japanese speakers, were instructed to identify the veridical center of the stimuli.
The results indicated that, with horizontal stimuli, participants placed their subjective midpoint farther left when smaller numbers were presented compared to larger numbers, consistent with the predicted left-to-right mental number line.
With vertical stimuli, however, smaller numbers induced biases upwards, inconsistent with the predicted bottom-to-top representation.
For the square, the stimuli with number strings elicited biases with a stronger vertical component, whereas the stimuli without number strings elicited biases with a stronger horizontal component.
The upward biases for the square stimuli with number strings may reflect the activation of object-based processing rather than the cognitive function of numerical processing.
These findings suggest that distinct mechanisms may predominate over spatial-numerical associations with regard to the biases in the two-dimensional plane.
