Institute of Psychology, Department of General Psychology, Humboldt University of Berlin, Rudower Chaussee 18, Berlin, Germany.
Mem Cognit. 2009 Dec;37(8):1103-19. doi: 10.3758/MC.37.8.1103.
In the present study, we examined the hypothesis of task-specific access to mental objects from verbal working memory. It is currently assumed that a mental object is brought into the focus of attention in working memory by a process of object selection, which provides this object for any upcoming mental operation (Oberauer, 2002). We argue that this view must be extended, since the selection of information for processing is always guided by current intentions and task goals. In our experiments, it was required that two kinds of comparison tasks be executed on digits selected from a set of three digits held in working memory. The tasks differed in regard to the object features the comparison was based on. Access to a new mental object (object switch) took consistently longer on the semantic comparison task than on the recognition task. This difference is not attributable to object selection difficulty and cannot be fully accounted for by task difficulty or differences in rehearsal processes. The results support our assumptions that (1) mental objects are selected for a given specific task and, so, are accessed with their specific task-relevant object features; (2) verbal mental objects outside the focus of attention are usually not maintained at a full feature level but are refreshed phonologically by subvocal rehearsal; and (3) if more than phonological information is required, access to mental objects involves feature retrieval processes in addition to object selection.
在本研究中，我们检验了从工作记忆的言语工作记忆中特定任务获取心理对象的假设。目前的假设是，一个心理对象通过对象选择的过程被带入工作记忆的焦点，为任何即将到来的心理操作提供这个对象（Oberauer，2002）。我们认为，这种观点必须得到扩展，因为信息的选择总是受到当前意图和任务目标的指导。在我们的实验中，要求在工作记忆中保持的三个数字中选择两个数字，执行两种比较任务。任务的区别在于比较所基于的对象特征。在语义比较任务中，访问新的心理对象（对象切换）比在识别任务中需要更长的时间。这种差异不能归因于对象选择的难度，也不能完全用任务难度或复述过程的差异来解释。结果支持了我们的假设，即（1）心理对象是为特定任务选择的，因此，它们是以特定的任务相关对象特征来访问的；（2）不在注意力焦点内的言语心理对象通常不会以完整的特征水平来保持，而是通过默读的方式在语音上得到刷新；（3）如果需要更多的语音信息，那么访问心理对象除了需要对象选择之外，还需要进行特征检索过程。
