Murdoch University Research Repository

Welcome to the Murdoch University Research Repository

The Murdoch University Research Repository is an open access digital collection of research
created by Murdoch University staff, researchers and postgraduate students.

Learn more

Recognition memory for musical pitch and rhythm is resistant to interference

Wood, Tabitha (2019) Recognition memory for musical pitch and rhythm is resistant to interference. Honours thesis, Murdoch University.

PDF - Whole Thesis
Download (757kB) | Preview


Recognition performance for most stimuli decreases as the amount of interference between exposure and recognition increases. However, for a few select stimuli—music, poetry and line drawings—recognition performance shows little to no disruption by interference. The novel Regenerative Multiple Representations (RMR) conjecture proposes that for these select stimuli, our experience influences how they are perceived and encoded into memory. In the case of music, conformity to the compositional rules of the music that a listener is accustomed to, influences whether or not they perceive a stimulus as musical. If a stimulus is perceived as musical, the memory of this stimulus will consist of multiple memory representations which provides resistance to the cumulative disruptive effects of interference. The present study aims to get closer to the precise mechanisms that drive this resilience. In order to do this, we used melodies that conform to fewer compositional rules of Western music compared to musical stimuli that previously showed resilience; yet more rules compared to stimuli that previously showed no resilience. Recognition performance for 90 participants was measured across three experiments, with Experiments 2 and 3 isolating the pitch and rhythm information, respectively. Results from all three experiments showed resistance to the effects of interference with up to 106 intervening items. These findings suggest that the strength of enculturation on memory for melodies is so powerful that even the small amount of recognisable musical information (in the stimuli used in the present study), was enough to trigger the processes described by the RMR conjecture.

Memory is a complex and essential aspect of human experience. Recognition memory—deciding whether an object has been encountered before—has fascinated cognitive psychologists for decades. The current consensus is that recognition memory consists of two separate mechanisms: recollection and familiarity (Rugg & Yonelinas, 2003; Yonelinas, 2002). Recollection, refers to high confidence recognition of stimuli based on their specific features, e.g. when and where it was previously encountered (Yonelinas, 1994). Familiarity, on the other hand, refers to low confidence recognition based on remembering previous encounters, but being unable to recall specific details (Yonelinas, 1994). For example, running in to someone that you know you have met before, but being unable to recall when, where or the person’s name. For the present study, memory recognition based on familiarity is sufficient as we are more interested in the mechanisms of forgetting (i.e. interference) rather than the extent of the participant’s recognition (Yonelinas, 2001).

Decay and interference refer to two mechanisms of forgetting (Sadeh, Ozubko, Winocur, & Moscovitch, 2014). Decay denotes forgetting as a result of time passing, whereas interference, the focus of the present paper, infers forgetting as a result of additional stimuli encountered during presentation and retrieval of a given target object (Herff, 2017; Sadeh et al., 2014). A continuous recognition paradigm—the task used in the present study—is an explicit memory task that is not only used for testing memory recognition but also the effects of interference (Shepard & Teghtsoonian, 1961). The continuous recognition paradigm involves presenting a set of stimuli to participants, multiple times, in a continuous stream. This allows the researcher to manipulate interference levels by changing the number of intervening items (interveners) between the first and subsequent presentations of a given item (Shepard & Teghtsoonian, 1961). For example, Olson (1969) used a continuous recognition paradigm to investigate the effects of interference on memory for trigrams (random 3 character combinations). The experiment consisted of a set of 334 trigrams, namely 20 target trigrams repeated 1 to 9 times, and 314 distractor trigrams (interveners) resulting in between 0-70 interveners between any initial and subsequent presentations of a target trigram. The present study is different, however, as the entire set of melodies are target items. The structure of the continuous recognition paradigm used in the present study is discussed further in the Analysis subsection of the General Method section.

The continuous recognition paradigm is different to other explicit tests of memory recognition in which the stimuli are usually presented in blocks (i.e., a block of stimuli to learn and a test block in which stimuli from the learning block need to be identified) (Schaal, Javadi, Halpern, Pollok, & Banissy, 2015). However, the continuous recognition paradigm is more relatable to the continuous recognition of everyday life as there is no distinct learning and test phase (Herff, 2017; Schwartz, 2016). For example, everyday life consists of a constant stream of multiple incoming stimuli through all sensory systems, that act as interference on each other. Furthermore, it is not always apparent when any incoming information will need to be recalled.

Recognition performance for the majority of stimuli decreases due to the cumulative disruptive effect of interveners (Olson, 1969; Sadeh et al., 2014). For example, a disruptive effect of interveners is found in; verbally presented words with 1 - 32 interveners (Buchsbaum, Padmanabhan, & Berman, 2011); visually presented words with 0 – 64 interveners (Poon & Fozard, 1980); photographs of household objects with up to 12 interveners (Yeung, Ryan, Cowell, & Barense, 2013); short letter-number combinations presented visually and verbally with 0 – 32 interveners (Le Breck & Baron, 1987); positive and negatively valanced faces with 10 – 15 interveners (Treese, Johansson, & Lindgren, 2010); sentences with 4 - 8 interveners (Tillmann & Dowling, 2007) and visual rhythm - expressed through flashing lights with 6 – 8 interveners (Collier & Logan, 2000). However, the disruptive effect of interveners on recognition is not found, when using a few select stimuli.

Firstly, melodies that consist of compositional rules that match the musical rules developed and accepted in the participant’s culture (Herff & Czernochowski, 2019; Herff, Olsen, & Dean, 2018) are especially resistant to interference, even with up to 197 interveners (Herff, Olsen, & Dean, 2018). Compositional rules can be thought of as a type of grammar in musical language. Just like grammatical rules for language allow for a shared understanding, the expectations that compositional rules create helps people understand and enjoy music (Vuvan, 2012). For example, there are many compositional rules specific to Western music such as tonality (all pitches being in a musical key) and predictable rhythm. However, when melodies that do not conform to the musical rules from a participant’s culture are used, the disruptive effect of interveners surfaces (Herff, Olsen, Dean, & Prince, 2018).

Secondly, line drawings show no significant reduction in recognition performance as a result of interveners whereas photographs do (Berman, Friedman, & Cramer, 1991). Thirdly, memory for poetry shows resistance to forgetting as a result of interference (Tillmann & Dowling, 2007). The regenerative multiple representations (RMR) conjecture (Herff, Olsen, & Dean, 2018) is a novel theory that has been proposed to explain why some stimuli demonstrate resilience to interference.

Item Type: Thesis (Honours)
Murdoch Affiliation(s): Psychology, Counselling, Exercise Science and Chiropractic
Supervisor(s): Prince, Jon and Herff, S.
Item Control Page Item Control Page


Downloads per month over past year