Bigelow James, Poremba Amy
University of Iowa, Iowa City, IA, USA.
University of Iowa, Iowa City, IA, USA.
Hear Res. 2015 Sep;327:69-77. doi: 10.1016/j.heares.2015.05.002. Epub 2015 May 14.
Recent studies using the delayed matching-to-sample (DMS) paradigm indicate that monkeys' auditory short-term memory (STM) is susceptible to proactive interference (PI). During the task, subjects must indicate whether sample and test sounds separated by a retention interval are identical (match) or not (nonmatch). If a nonmatching test stimulus also occurred on a previous trial, monkeys are more likely to incorrectly make a "match" response (item-specific PI). However, it is not known whether PI may be caused by sounds presented on prior trials that are similar, but nonidentical to the current test stimulus (item-nonspecific PI). This possibility was investigated in two experiments. In Experiment 1, memoranda for each trial comprised tones with a wide range of frequencies, thus minimizing item-specific PI and producing a range of frequency differences among nonidentical tones. In Experiment 2, memoranda were drawn from a set of eight artificial sounds that differed from each other by one, two, or three acoustic dimensions (frequency, spectral bandwidth, and temporal dynamics). Results from both experiments indicate that subjects committed more errors when previously-presented sounds were acoustically similar (though not identical) to the test stimulus of the current trial. Significant effects were produced only by stimuli from the immediately previous trial, suggesting that item-nonspecific PI is less perseverant than item-specific PI, which can extend across noncontiguous trials. Our results contribute to existing human and animal STM literature reporting item-nonspecific PI caused by perceptual similarity among memoranda. Together, these observations underscore the significance of both temporal and discriminability factors in monkeys' STM.
近期使用延迟样本匹配(DMS)范式的研究表明,猴子的听觉短期记忆(STM)易受前摄干扰(PI)影响。在该任务中,受试者必须指出被记忆间隔分开的样本音和测试音是否相同(匹配)或不同(不匹配)。如果一个不匹配的测试刺激也出现在前一次试验中,猴子更有可能错误地做出“匹配”反应(特定项目PI)。然而,尚不清楚PI是否可能由先前试验中呈现的与当前测试刺激相似但不完全相同的声音引起(非特定项目PI)。在两项实验中对这种可能性进行了研究。在实验1中,每次试验的记忆内容包括一系列频率范围很广的音调,从而将特定项目PI降至最低,并在不同音调之间产生一系列频率差异。在实验2中,记忆内容取自一组八种人造声音,它们在一个、两个或三个声学维度(频率、频谱带宽和时间动态)上彼此不同。两项实验的结果均表明,当先前呈现的声音在声学上与当前试验的测试刺激相似(尽管不完全相同)时,受试者会犯更多错误。仅前一次试验的刺激产生了显著影响,这表明非特定项目PI比可跨越不连续试验的特定项目PI持续性更弱。我们的结果为现有的关于由记忆内容之间的感知相似性导致非特定项目PI的人类和动物STM文献做出了贡献。这些观察结果共同强调了时间和可辨别性因素在猴子STM中的重要性。