Mechanisms and evolution of signal pattern recognition

描述（由申请人提供）：所提议研究的总体目标是获得有关信号模式识别的机制和演化的知识。该研究利用青蛙声学通信的优势作为模型系统来探索：（1）传递具有生物学意义的信息的信号物理特性的组合； (2) 神经系统处理和提取此类信息的工作原理； (3)信号模式识别的进化变化模式。 由于青蛙不需要听觉经验来产生或识别信号，因此该系统可以产生与婴儿识别和辨别语音中的类似问题有关的知识。也就是说，所有语言都来自一组共同的语音片段，而几乎没有语言经验的婴儿可以区分甚至在他们最终学习的语言中不会出现的语音片段。因此，这项研究解决了有关模式识别的先天机制的问题，这种机制使正常言语的发展成为可能，而这反过来又对人类心理健康很重要。 拟议的研究将使用计算机合成信号的回放来模拟男性交流的声音；雌性以积极的趋音性对这些声音作出反应，并辨别一种或多种特性的细微差别。该研究解决了基于信号频率内容和精细时间特性差异的判别标准的相互作用。这些相互作用将在两个密切相关的物种中进行量化；然后，这些结果将用于选择对其他七个或多或少密切相关的物种进行测试。这些数据与对所有九个物种的系统发育关系的独立估计相结合，将用于评估模式识别可能如何进化。还将确定可以解释进化趋势的机制和环境因素。最终，这项研究对于构建关于模式识别基础的感觉机制的可检验假设以及指导此类研究的执行非常重要。

DESCRIPTION (provided by applicant): The overall goal of the proposed research is to gain knowledge about mechanisms and evolution of signal pattern recognition. The research exploits the advantages of acoustic communication in frogs as a model system for exploring: (1) combinations of physical properties of signals that convey biologically significant information; (2) the operating principles by which nervous systems processes and extract such information; and (3) patterns of evolutionary change in signal pattern recognition. Because frogs do not require acoustic experience to produce or to recognize signals, this system can yield knowledge bearing on analogous problems in the recognition and discrimination of speech sounds by infants. That is, all languages draw from a common set of phonetic segments, and infants with little linguistic experience can discriminate among phonetic segments that do not even occur in the language that they will ultimately learn. This research thus addresses questions about inborn mechanisms of pattern recognition that make possible the development of normal speech, which, in turn, is important for human mental health. The proposed research will use playbacks of computer-synthesized signals that simulate male communication sounds; females respond to these sounds with positive phonotaxis and discriminate subtle differences in one or more properties. The research addresses interactions of discrimination criteria based on the frequency content of the signal and differences in fine-scale temporal properties. These interactions will be quantified in two closely related species; these results will then be used to select tests to be conducted with seven other species that are more or less closely related. These data, in conjunction with an independently derived estimate of the phylogenetic relationships of all nine species, will be used to assess how pattern recognition might have evolved. Mechanistic and environmental factors that could explain evolutionary trends will also be identified. Ultimately, this research will be important for framing testable hypotheses about sensory mechanisms underlying pattern recognition and for guiding the execution of such studies.