Koenderink J J
Biol Cybern. 1984;50(1):35-41. doi: 10.1007/BF00317937.
A nervous net may be studied by an external observer who notes, for example, such features as the spatial lay-out of receptive fields, the existence of somatotopic maps, etc. It is not often acknowledged that many of such features have no functional relevance to the machine itself. Thus somatotopy from the sensor to the motor apparatus may appear important (and in certain respects it is), but it is irrelevant to the function of the organism as an abstract machine. For the abstract machine only functional relations, but not spatial relations per se, count. It is shown how, proceeding from purely functional relations (cross-correlations of signals), a simultaneous order of neural elements may be constructed. Such an order has an objective existence for the abstract machine, not merely for an external observer. It is argued that sensory modalities and the cohesion within modalities (e.g. the visual field) must be understood in such a functional manner. Thus the two-dimensionality of the visual field is objectively present in the cross-correlation structure of optic nerve signals, it exists independently from an external observer's description of the retina as a two-dimensional receptor array.
外部观察者可以研究神经网,例如,该观察者会记录诸如感受野的空间布局、躯体感觉定位图的存在等特征。人们常常没有认识到,许多这样的特征与机器本身并无功能上的关联。因此,从传感器到运动器官的躯体感觉定位可能看起来很重要(在某些方面确实如此),但对于作为抽象机器的生物体功能来说,它是无关紧要的。对于抽象机器而言,只有功能关系本身才是重要的,而空间关系并非如此。本文展示了如何从纯粹的功能关系(信号的互相关)出发,构建神经元件的同步顺序。这样的顺序对于抽象机器来说具有客观存在性,而不仅仅是对于外部观察者而言。有人认为,感觉模态以及模态内部的连贯性(例如视野)必须以这种功能方式来理解。因此,视野的二维性客观地存在于视神经信号的互相关结构中,它独立于外部观察者将视网膜描述为二维感受器阵列而存在。