Research on memory often describes the remarkable longevity of music. However, memory for music is not uniform. Cook (1987) found that participants were not able to tell apart excerpts that modulated from those that did not when the excerpt was longer than 1 minute in length. This suggests that participants were no longer able to remember, and compare, musical keys after a relatively short period of time. Farbood (2016) and Woolhouse et al. (2016) further explored the limitations of memory for tonal structures finding that, in fact, harmonic memory only lasts up to 21 seconds after modulation. However, this research was done using homophonic stimuli—arpeggios or quarter-note chords—that may not be representative of the music participants would be listening to regularly. The focus of this project was to explore how the addition of certain musical features, such as melodic or rhythmic figurations, may influence harmonic memory. Observing these possible influences may provide us with insight into the processes responsible for auditory memory and how it differs from other domains, such as speech or vision. Chapter 1 explores prominent memory literature and music cognition experiments that support, or address concerns with, common memory models. Here, I introduce a cognitive system which reconciles music research with models by memory specialists such as Baddeley and Snyder. Chapter 2 presents a detailed account of background empirical literature, including Farbood (2016) and Woolhouse et al. (2016). Though fundamental to the exploration of temporally nonadjacent harmonic memory, this research is potentially limited in its generalizability due to the homophonic nature of the stimuli. Chapter 3 explores this limitation by testing the effects of adding surface features—melodic and rhythmic components often used for elaboration in composition—on memory for large-scale tonal structures. Results found that harmonic memory is, indeed, enhanced and prolonged by these elaborative components, lasting up to 33 seconds, well past the limit found in previous research. Farbood (2016) further claimed that harmonic memory is significantly interrupted by new, highly harmonic excerpts. However, results from Woolhouse et al. (2016), Spyra et al. (2021) and those from Chapter 3 all question this claim as they employed stimuli that was highly harmonic. Chapter 4 investigates the contradiction by testing whether functional diatonic, functional chromatic, or random sequences degraded harmonic memory for an original key. Functional diatonic intervening information resulted in increased harmonic memory, directly contradicting Farbood’s original findings. In Chapter 5, these results are explored in terms of prominent memory models in the field of cognition, supporting standard models of memory such as that by Baddeley and Hitch (1974) or Atkinson and Shiffrin (1968), as well as my proposed cognitive system. This is further elaborated by discussing the process of undergoing a musical judgement task from perception through to decision-making. In summary, this project suggests that more generalizable stimuli containing realistic musical features produce a significant boost in harmonic memory. Furthermore, this arguably calls into question standard practices in analysis that categorize surface features as hierarchically less important than ’deeper’ harmonic events, and thus, potentially less important from a cognitive perspective. Which is to say, this evidence suggests that these features may play a vital role in remembering nonadjacent harmonic structures. / Dissertation / Doctor of Philosophy (PhD) / Memory for music is often celebrated for its longevity. Music is a complex stimulus, however, and not all of its characteristics are remembered equally well. Past research has found that participants were not able to remember musical keys after a surprisingly short period of time: Farbood (2016) and Woolhouse et al. (2016) found that harmonic memory—i.e., memory for a key—lasts up to 21 seconds after a key change. Compared to nursery rhymes remembered from childhood bedtimes, this is remarkably limited. Yet this research did not fully explore which musical characteristics affect harmonic memory as it was done using simple musical stimuli: compositions made of blocks of chords. Whereas a string of chords might sound pleasant, it may not be representative of the type of music that people listen to regularly (with complex melodies and instrumentation). The focus of this project was to explore musical factors, such as melodies or rhythms, and measure how they interact with musical memory. Observing specific aspects of the stimulus gives us a window into the complexities of human memory, particularly that of the auditory domain. Chapter 1 provides an overview of memory literature with a focus on common memory models and the musical research that supports them or contributes to their development. Here, I propose a cognitive system which integrates prominent models that otherwise describe different stages of processing complex auditory stimuli. Chapter 2 presents a detailed account of background empirical literature. This provides a basis for a series of experiments outlined in Chapters 3 and 4. These experiments investigate how components of music influence harmonic memory. Components include Surface Features, or ornamentations in music such as melodies or rhythms, and Harmony, the structure of the key itself which can make an excerpt sound more, or less, familiar. Results suggest that memory is significantly enhanced and prolonged by the addition of surface features. Furthermore, harmony that most resembles culturally familiar compositional practices also provides a memory boost when compared to random or somewhat ambiguous sequences. In Chapter 5, the implications of these results are explored with regards to the general memory models discussed in Chapter 1. Results support standard models of memory and my proposed cognitive system, as demonstrated by following the processing of my experimental musical stimuli from sound to executive function. This project suggests that more complex and musically realistic stimuli produce a significant memory boost. This puts into question traditional practices in music analysis which separate surface features into hierarchically less important positions when, in fact, the musical surface may be vital to our processing of auditory stimuli.

Lexical categories like noun and verb are foundational to language acquisition, but these categories do not come neatly packaged for the infant language learner. Some have proposed that infants can begin to solve this problem by tracking the frequent nonadjacent word (or morpheme) contexts of these categories. However, nonadjacent relationships that frame categories contain reliable adjacent relationships making the type of context (adjacent or nonadjacent) used for category acquisition unclear. In addition, previous research suggests that infants show learning of adjacent dependencies earlier than learning of nonadjacent dependencies and that the learning of nonadjacent word relationships is affected by the intervening information (how informative it is and how familiar it is). Together these issues raise the question of whether the type of context used for category acquisition changes as a function of development. To address this question, infants ages 13, 15, and 18 months were exposed to an artificial language containing adjacent and nonadjacent information that predicted a category. Infants were then tested to determine whether they 1) detected the category using adjacent information 2) only detected the nonadjacent dependency, with no categorization, or 3) detected both the nonadjacent relationship and the category. The results showed high individual variability in the youngest age group with a gradual convergence towards detecting the category and the associated environments by 18 months. These findings suggest that both adjacent and nonadjacent information may be used at early stages in category acquisition. The results reveal a dynamic picture of how infants use distributional information for category acquisition and support a developmental shift consistent with previous infant studies examining dependencies between words.

Includes an exploration of working memory models from a musical standpoint. / A musical key can only be retained in memory for approximately 20 seconds. However, other factors may be influencing the strength of this retention. The current study tests the influence of time and number of events (chords) in an intervening key on the deterioration in memory of a nonadjacent key. Stimuli first established a major key using traditional harmonic rules, then modulated to an intervening key that was either 6 or 9 seconds in duration and formed from either 4 or 6 chords. Stimuli then returned to the original key in a probe cadence. Participants were asked to rate this cadence in terms of its sense of closure. It was revealed that there is a significant negative effect of time on the probe cadence though no effect of number of events was found. This suggests that spending more time in an intervening key, and not the number of intervening chords, diminishes the memory of the original key. However, it is unclear from this study where in memory the nonadjacent key relationship is processed. Relevant literature is examined to form a working hypothesis with the goal of strengthening future studies with a capable foundation in memory research and theories. / Thesis / Master of Science (MSc) / We have a very short memory span for musical keys. Humans can only remember a key for approximately 20 seconds after it’s changed. However, it is unclear whether it is time spent away from the key that makes us forget or if it is the number of chords we heard in a second, interrupting key. The current study tests this distinction using nonadjacent key relationships—in which a key is established, a different key interrupts for a random amount of time and number of chords before returning to the original key for two target chords. Results confirmed previous findings as there was a significant effect of time on memory. However, there was no effect of number of chords, suggesting that it is the length of time spent away from a key, not number of chords that has an effect on memory.