Abstract
The phenomenon of collinear facilitation demonstrates spatial integration and is thought to be mediated by lateral interactions. It was described psychophysically in the fovea (Polat & Sagi, Vision Res. 1993) but could not be found in extra-fovea eccentricities (3–4°) when using Gabor stimuli (e.g. Williams & Hess, JOSA A, 1998). However, physiological studies in the primary visual cortex of the cat (Polat et al, Nature, 1998) and the monkey (Kapadia et al, Neuron, 1995) showed increased firing rates for co-aligned stimuli around 4°. We tried to resolve this apparent conflict.
Here we show that the absence of peripheral facilitation is not a consequence of improper scaling, and that collinear facilitation is reduced or gone when targets are placed outside fixation but still within the fovea (1°). This raises the possibility that attention is involved in the facilitation mechanism. Indeed, it has been shown that lateral interactions are modulated by attention in the fovea (Freeman et al, Nature Neurosci, 2001). Therefore we manipulated attention at an eccentricity of 4 degrees using a double task which contained (1) Vernier acuity of the two flanking Gabor patches and (2) detection of the target located in between. d' for detection of the collinear target was ∼2 times higher than for the orthogonal one (d'c/d'o=1.97). When performing the detection task alone, d' for collinear and orthogonal targets were similar (d'c/d'o =1.22). These results show that collinear facilitation is present at 4°, provided that attention is appropriately allocated. This finding demonstrates again that lateral interactions are subject to attentional modulation. The orientation selectivity of this attention-induced facilitation implies that it is mediated by lateral interactions. It seems that differences in allocation of attention may underlie part of the distinction between foveal and peripheral vision.