Learning the positions of words relative to a marker element.

Learning the positions of words relative to a marker element.
复制标题

学习单词相对于标记元素的位置。

DOI:
--
复制
发表时间:
1966
影响因子:
--
通讯作者:
M. Braine
M. Braine
中科院分区:
--
文献类型:
--
作者:
M. Braine

文献摘要

被引文献

相似文献

通过让孩子们接触一种微型的、语义上空洞的语言的子集,研究了位置条件定义--“g之后的第一”、“g之后的第二”--的学习。这包括两种短语类型,FA和GPQ,单独出现或以任何顺序出现。学生学习了这个结构,包括条件位置定义。句子补全测试问题中的错误表明,特别是在学习的早期,有混淆两个短语中对应位置的倾向。学习到的关系不能合理地用有限状态图来表示,一种包含下推存储的状态图似乎比另一种更合适。有人认为,有限状态图不能反映关联的性质,特别是远程关联,关联可能更好地被视为将或有项与短期存储中的条目相关联的规则,或者可能被视为对上下文敏感的重写规则。
The learning of conditional definitions of position—"first after g," "second after g"—was studied by exposing children to a subset of the sentences of a miniature, semantically empty, "language." This contained 2 phrase types, fA and gPQ, occurring alone or in either order. Ss learned the structure, including the conditional positional definitions. Errors in sentence-completion test problems indicated a tendency to confuse homologous positions in the 2 phrases, particularly early in learning. The relationships learned could not reasonably be represented by a finite state diagram, and 1 kind of state diagram incorporating a push-down store seemed more appropriate than another. It was argued that finite state diagrams do not capture the properties of associations, particularly remote associations, and that an association might better be regarded as a rule relating contingent items to an entry in a short-term store, or, perhaps, as a contextsensitive rewrite rule.