Our binocular vision means that we see horizontally. Fatter frames are the most natural image format. In other words, they are the least intrusive and most accommodating to the eye. The naturalness of horizontal vision reinforces the eye's desire to scan from side to side, and a corresponding reluctance to scan up and down. There is an assumption that the bottom of the picture is a base; a level surface on which other things can rest. Most things are longer in one direction than in another, and it is natural to align the main axis of an image with the longer sides of the rectangular frame.
A square format render often suffer from lack of direction. Patters and other formless arrangements fit well into a square frame because the frame has no directional emphasis - very few renders lend themselves well to square compositos.
The question of which aspect ratios are perceived as the most comfortable is a study in its own right, but in principle, there seems to be a tendency towards longer horizontally, but less elongated for vertically composed images. The most common render is the proportions 3:2, but it does depend on the subject you are rendering.