Texts on visual perception typically begin with the following premise: Vision is an ill-posed problem, and perception is underdetermined by the available information. If this were really the case, however, it is hard to see how vision could ever get off the ground. James Gibson’s signal contribution was his hypothesis that for every perceivable property of the environment, however subtle, there must be a higher order variable of information, however complex, that specifies it—if only we are clever enough to find them. Such variables are informative about behaviorally relevant properties within the physical and ecological constraints of a species’ niche. Sensory ecology is replete with instructive examples, including weakly electric fish, the narwal’s tusk, and insect flight control. In particular, I elaborate the case of passing through gaps. Optic flow is sufficient to control locomotion around obstacles and through openings. The affordances of the environment, such as gap passability, are specified by action-scaled information. Logically ill-posed problems may thus, on closer inspection, be ecologically well-posed.

关于视觉感知的文本通常以以下前提开始：视觉是一个不适定问题，感知是由可用信息决定的。然而，如果情况确实如此，就很难想象视觉如何能够落地。詹姆斯·吉布森的信号贡献是他的假设，即对于环境的每一个可感知的属性，无论多么微妙，都必须有一个更高阶的信息变量，无论多么复杂，来指定它——只要我们足够聪明，能够找到它们。这些变量提供了有关物种生态位的物理和生态限制内的行为相关属性的信息。感官生态学充满了启发性的例子，包括弱电鱼、独角鲸的獠牙和昆虫的飞行控制。我特别阐述了穿越间隙的情况。光流足以控制障碍物周围和通过开口的运动。环境的可供性（例如间隙可通过性）由行动规模的信息指定。因此，经过仔细检查，逻辑上不适定的问题可能在生态上是适定的。