ITR: Linguist-Machine Collaboration for Comparative Linguistics

在美国国家科学基金会的支持下，约瑟夫·E·格里姆斯博士将进行为期两年的语言学研究，以扩大比较语言学家的能力，方法是将他们做出模式识别判断的专业知识与一种计算工具相匹配，该工具可以组织他们所有的数据，以回应这些判断。这种分工将使语言学家能够掌握来自更多语音种类的更大数量的数据，因为它为他们的学科提供了信息技术基础设施。尽管在语言比较方面的显著工作是手工完成的，但这需要花费数年的时间；事实上，对于世界上大多数人来说，关于语言关系可能是什么的更有根据的猜测，而不是那些关系的可靠证据。该项目的语言学家-机器合作伙伴关系将使研究人员能够迅速进行分析(包括添加可能在下一代灭绝并消失于科学领域的语言的数据)，遵循相互竞争的假说，像单独合作一样轻松地合作，并通过公共档案向其他人提供准确的文件。所有这些都是用传统方法很难做到的。为了找出人类群体是如何分化和迁徙的，学者们不仅必须利用书面和口头历史，还必须利用可恢复的史前，这主要是通过比较语言学、考古学和遗传学来获得的。如果语言学家能够填补他们数据中的空白，并更准确地重建我们的祖先语言肯定是什么样子，这将导致更好地洞察所有语言(包括英语)随时间变化的过程，并更好地理解祖先语言和古代文化模式之间的联系。以这种方式看待语言历史也增加了我们对人类语言能力的理解的时间维度，并帮助我们珍视语言多样性，而不是谴责它。拟议的工具将促进这一领域的扎实研究，这也可以帮助传统社会认识到比口头历史本身给他们带来的更深层次的遗产。在更直接的范围内，它可以确定在哪些领域，可以很容易地将以具有共同文化的密切相关的一组语言之一开发的教育材料和文学作品转变为同一组其他语言的这些材料的初稿。

With National Science Foundation support, Dr. Joseph E. Grimes will conduct two years of linguistic research to extend the capabilities of comparative linguists by matching their expertise in making pattern recognition judgments with a computational tool that organizes all their data in response to those judgments. This division of labor will allow linguists to stay on top of greater quantities of data from more speech varieties than has ever been possible, because it provides an information technology infrastructure for their discipline. Though notable work in language comparison has been done by hand, it consumes years; indeed, for much of the world there are more educated guesses about what the linguistic relationships might be than there is well-martialed evidence for those relationships. This project's linguist-machine partnership will empower investigators to do their analysis expeditiously (including adding in data from languages that could become extinct in the next generation and be lost to science), to follow out competing hypotheses, to work collegially as easily as individually, and to offer precise documentation to others through public archives. All these things have been difficult to do using traditional methods.To find out how human groups have diverged and migrated scholars must draw not only on written and oral history, but also on recoverable prehistory, which is accessible mostly through comparative linguistics, archeology, and genetics. If linguists are enabled to fill the gaps in their data and to make more precise their reconstructions of what our ancestral languages must have been like, it will lead to better insight into the processes by which all languages (including English) change over time, and to better understanding of connections between ancestral languages and ancient culture patterns. Seeing language history in this way also adds the time dimension to our understanding of humanity's capacity for language, and helps us to value language diversity rather than deplore it. Solid research in this area, which the proposed tool will promote, can also help traditional societies to become aware of a deeper heritage than what their oral history alone gives them. On a more immediate scale, it can identify areas where educational materials and literature developed in one of a closely related group of languages that share a common culture can be easily morphed into preliminary drafts of those materials for other languages in the same group.