Complex Worlds from Simpler Nervous Systems

Description

Anyone curious as to extent to which various human cognitive and neural capabilities are can exist in nervous systems that are much smaller and simpler than humans will gain a lot from the perusal of this book. But more importantly, the book also offers a glimpse of how these nervous systems are able to deal with their environment in ways that perpetuate the survival of the organisms that possess them. Their abilities in many ways surpass those of humans, but the comparison with humans should really not be the focus of attention. The most important thing to gain from the reading of this book is that nervous systems have evolved in ways that are advantageous to the organism. As two authors in the book expressed it, "the abilities of an animal seem to be governed largely by what it needs to pursue its lifestyle." All of the articles in this book are interesting, but for lack of space only three of them will be reviewed here.

The authors of the article "Exploration of Cognitive Capability in Honeybees: Higher Functions Emerge from a Small Brain", give a brief but fascinating overview of the research that has been performed in the neural and learning capabilities of honeybees. It is incredible fact, as brought out in the article, that the brain of the worker honeybee is only one cubic millimeter in volume, has a mass of only 1 mg, and has less than a million neurons. In spite of these dimensions however, honeybees are still able to process visual and motion information in ways that are very similar to the way that humans do. Indeed the honeybee is able to engage in pattern recognition, perception, and the learning of complex tasks. Honeybees are able to take pattern presented to them, train on these patterns, and use what they have learned to evaluate new patterns presented to them. Most interestingly, the authors describe experiments that show that honeybees are able to perceive some of the illusions that humans do. Other abilities discussed include learning to negotiate complex mazes, and are able to count landmarks as they do. Furthermore, they make use of rules that worked in the past in order to navigate through mazes. Thus bees exhibit a remarkable ability to construct concepts. The authors also mention the exciting prospect of constructing a learning machine that is capable of performing behavior similar to the honeybee. Given the size of the honeybee brain, this certainly seems like a goal that could be readily accomplished.

In the article "In the Mind of a Hunter: The Visual World of Praying Mantis", the authors present the mantis as being an insect that is very complex from the standpoint of its ability to process information, being manifested in what the authors refer to as "plastic behaviors." Anyone who has observed a praying mantis in a garden or other places outdoors cannot help but be fascinated by their behavior. This article puts these behaviors on a neurological foundation, and the picture the authors paint is a very interesting one. The reader learns of the compound eyes of the praying mantis, which allow visualization in every direction. The range of light intensity (four log units) allows the mantis to distinguish between different objects. Amazingly, their eyes have about nine thousand sampling units or `ommatidia' as the authors call them. But it is the "prey recognition" algorithm used by mantids that is of primary interest to the authors. They have found through their research that this algorithm depends on the simultaneous assessment of a collection of stimulus parameters. From the standpoint of its nervous system, prey recognition is accomplished by a movement-sensitive cell called the lobula giant movement detector (LGMD). The LGMD is presynaptic to the descending contralateral movement detector (DCMD). They mention the construction of an artificial neural network of the LGMD-DCMD systems that learns to respond to the same types of stimuli that mantids recognize as prey, but unfortunately do not discuss it in any detail.

The author of the article "Motion Perception Shapes the Visual World of Amphibians" discusses how frogs and toads are able to catch their prey, avoid predators, and find mates without the benefit of eye movements. The emphasis in the article is in on how these different entities are classified and discriminated, how retinal images of moving objects are discriminated from self-induced moving images, on whether or not toads employ concepts or engage in learning, and how toads analyze visual stimuli without the benefit of a cerebral neocortex. The distinction between prey and nonprey is correlated with the geometry of the object relative to the direction of movement. In order to justify what is happening at the neuronal level, the author describes the properties of the retinal ganglion cells (which mediate the output of the retinal network) and the neurons of the retinal projection fields in terms of their receptive fields. A table is given along with extension discussion of their properties. Toads also make use of the odor of their prey, and the author discusses, with a detailed diagram, the brain structures involved in visual-olfactory learning. Most interesting is the author's discussion of backpropagation artificial neural networks used to model the feature detection abilities of amphibians. A two-layered artificial neural network is trained to classify and evaluate objects of different lengths moving in prey and nonprey configurations.