Herff Steffen A, Olsen Kirk N, Dean Roger T
1 The MARCS Institute for Brain, Behaviour and Development, Western Sydney University, Penrith, NSW, Australia.
2 Department of Psychology, Macquarie University, Sydney, NSW, Australia.
Q J Exp Psychol (Hove). 2018 May;71(5):1150-1171. doi: 10.1080/17470218.2017.1318932. Epub 2018 Jan 1.
In many memory domains, a decrease in recognition performance between the first and second presentation of an object is observed as the number of intervening items increases. However, this effect is not universal. Within the auditory domain, this form of interference has been demonstrated in word and single-note recognition, but has yet to be substantiated using relatively complex musical material such as a melody. Indeed, it is becoming clear that music shows intriguing properties when it comes to memory. This study investigated how the number of intervening items influences memory for melodies. In Experiments 1, 2 and 3, one melody was presented per trial in a continuous recognition paradigm. After each melody, participants indicated whether they had heard the melody in the experiment before by responding "old" or "new." In Experiment 4, participants rated perceived familiarity for every melody without being told that melodies reoccur. In four experiments using two corpora of music, two different memory tasks, transposed and untransposed melodies and up to 195 intervening melodies, no sign of a disruptive effect from the number of intervening melodies beyond the first was observed. We propose a new "regenerative multiple representations" conjecture to explain why intervening items increase interference in recognition memory for most domains but not music. This conjecture makes several testable predictions and has the potential to strengthen our understanding of domain specificity in human memory, while moving one step closer to explaining the "paradox" that is memory for melody.
在许多记忆领域中，随着中间插入项目数量的增加，人们会观察到在一个物体的首次和第二次呈现之间识别表现有所下降。然而，这种效应并非普遍存在。在听觉领域，这种干扰形式已在单词和单音符识别中得到证实，但尚未使用诸如旋律等相对复杂的音乐素材加以证实。事实上，越来越明显的是，在记忆方面音乐呈现出引人入胜的特性。本研究调查了中间插入项目的数量如何影响对旋律的记忆。在实验1、2和3中，在连续识别范式下每次试验呈现一首旋律。在每首旋律之后，参与者通过回答“旧的”或“新的”来表明他们在实验中之前是否听过该旋律。在实验4中，参与者对每首旋律的感知熟悉度进行评分，且未被告知旋律会重复出现。在使用两个音乐语料库、两种不同记忆任务、转调和未转调旋律以及多达195首中间插入旋律的四项实验中，未观察到除第一首之外中间插入旋律数量产生干扰效应的迹象。我们提出一种新的“再生多重表征”猜想，以解释为什么中间插入项目会增加大多数领域而非音乐领域在识别记忆中的干扰。这一猜想做出了几个可检验的预测，并且有可能加强我们对人类记忆中领域特异性的理解，同时朝着解释旋律记忆这一“悖论”又迈进了一步。
