In the first part of Chapter 5, we focused predominantly on the elements that make up melody and how Gestalt principles might be used to perceive these elements as a coherent whole. Continuing the discussion on the perception of melody and musical pitch, Tan et al. (2010) round out this chapter by focusing on memory for melody and potential neural bases for pitch and melody perception.
Memory for pitch and melody
Memory for pitch
The claim that most people can remember melodies is quite an uncontroversial one. But how do we actually remember these melodies? One important question concerning melodic memory is whether memory for melodies is based on absolute pitch (i.e., pitch heights and chromas) or relative pitch (i.e., intervallic relationships). Given the rarity of perfect pitch, or the ability to recognise and/or produce individual pitch chromas on demand without being provided a reference pitch, many have ruled out absolute pitch as the basis for pitch memory and concluded that melodies are remembered via relative pitch representations in the brain.
But is this true? Some evidence has suggested that most people do, in fact, recall melodies with respect to absolute pitches. For example, Levitin (1994) asked nonmusicians to sing their favourite song from memory and found that participants' responses were always within 2 semitones of the melody's original key. Importantly, while they were unable to correctly name the pitch chromas (i.e., they do not have perfect pitch), they nonetheless demonstrated good pitch memory.
One possible explanation for the above results is that participants were simply producing those melodies in those keys because they were comfortable within their singing ranges. Hence, the results are not because participants had good pitch memory, but rather, because of limitations in their ability to sing out notes of different pitch heights. This might be a valid interpretation if not for other studies that extended these results to the context of pitch perception. Here, Schellenberg & Trehub (2003) found that most people could still identify the original key of popular TV show theme songs even when they had been shifted up or down by up to 2 semitones. Given conceptual replications in both domains of pitch production and perception, Tan and colleagues seem to argue that memory for pitch really is based on absolute pitch representations.
Memory for melodic contours and intervals
Some evidence suggests that memory for intervals is strongly influenced by melodic contour, tonality, and key. For instance, Dowling (1978) provided melody pairs to listeners and asked them to report if the melodies were transposed versions of each other or if they were different. Here, melodies that are transposed constitute those that differ in key but have the same within-melody intervallic relationships (and by necessity, the same contour). Admittedly, I was quite confused with this section, but I'm going to try my best to communicate the results here.
Generally, Dowling found that participants (and especially musically-trained listeners) correctly identified atonal melodies that preserved the contour of the first melody. In other words, when the second melody implied no particular key (hence different intervals, no tonality, no key) but possessed the same melodic shape as the first melody, participants reported that the melodies were NOT transposed versions of each other.
So far, so good. What about for tonal comparisons, or based on the definition used in the study, melodies that a) imply a key, b) are within the same key, c) have the same contour, but d) have different intervallic relationships? Here, participants were unable to recognise that the melodies had different intervals and incorrectly reported that they were direct transpositions of each other.
This indicates that under tonal comparisons (i.e., comparing melodies that imply the same key), preserving melodic contour is sufficient to get participants to report remembering hearing the same within-melody intervallic relationships. On the other hand, under atonal comparisons (i.e., comparing a tonal to an atonal melody), participants do not report remembering hearing the same intervals even when the melodic shapes are the same.
My personal interpretation of this is that, at least initially, intervals within a melody are not remembered, perhaps due to the sheer volume of them. Instead, it is easier to remember the more global nature of a melodic contour, which has been established as one of the more salient features of a melody. This might explain why tonal comparisons with the same shapes are identified as transpositions even when they don't have identical intervallic relationships. That said, our exposure to music (in my context, Western diatonic music) leads to a strong preference for tonal music that implies a particular scale -- we generally like music that seems contextualised, which may also explain why atonal music often sounds so foreign and so full of tension. This might then explain how participants were so quick to identify atonal comparisons in Dowling's study, even when the contour remained the same.
One more thing I will note is that I think the initial confusion over the results of the study came from the author's definition of tonality. While Tan and colleagues defined tonality as whether or not a melody implies a scale or key in the earlier part of the chapter, it seems that in Dowling's experiment, it has the additional criterion of being within the same key. I don't agree with this -- two melodies can be tonal yet still imply different keys. Maybe I've just misread this whole section. If anyone would like to give their input, it would be greatly appreciated!
To round out this section, Tan and colleagues caution that the above results likely only apply to unfamiliar melodies. Once participants become familiar with the melody, they become better at identifying that melodies under a tonal comparison are not direct transpositions of each other (Dowling & Bartlett, 1981). Given this, we might need to rephrase the opening claim in this section -- memory for melody seems to be influenced by melodic contour, tonality, key, and melody familiarity.
Tonal schemas
Recall my claim that atonal melodies are highly salient because of our general preference for tonal music. One way in which music research has tried to explain this is by using the concept of tonal schemas. A schema is basically a representation or internal structure within our minds which are used to make sense of incoming input. The general process goes as follows: 1) our sensory organs receive sensations from the world, 2) these sensations are ambiguous, vague, and have no meaning, 3) these sensations are compared to mental schemas in long-term memory, 4) which then transform them into meaningful, unambiguous information -- our perception of the world. I will say here, before we get carried away, that I fully reject the existence of and need for mental representations and schemas. I talk about it briefly in the first few paragraphs here, but for the sake of honestly covering the Tan and colleague's book, I'll continue for now.
Back to the text. So, given the above description of schemas, tonal schemas basically imply an internal representation in our minds, which is used to process incoming pitch information to tell us whether or not a note is 'right' or 'wrong' within the tonal context of a melody. To elucidate what this tonal schema might look like, Krumhansl (1979) asked listeners to report how similar pairs of notes sounded after they had been presented with a short melody that established the tonal context of the notes. Using some multidimensional analysis (which I don't fully get either), the author managed to fit the results onto a conical shape as seen in Fig. 1.
In gist, similarity between notes is represented by the geometric distance between pitches. In the context of the C major scale, Tan and colleagues point out 3 main features of Fig. 1. Firstly, between levels, the first (lowest) level is closer to the second than the second level is to the third (highest) level. This makes sense, given how the first 2 levels make up the notes of the C major scale, whereas the highest level contains out-of-key notes. Secondly, within levels, pitches in the first level are highly similar, as seen from the smaller diameter of the level and the distance between notes. This is contrasted by the larger diameter of the third level, which suggests that out-of-key notes are not only highly dissimilar from within-key notes, but also from each other. Finally, the general arrangement of the levels signifies a downward pull towards the first level, like gravity pulling objects towards the ground. This indicates that notes in the first level are generally more stable than those in other levels, a conclusion further supported by how these notes make up the triad of the tonic chord (in this case, C major), which is often referred to as the most stable 'home' chord of any key.
Fig. 1 Conical representation of within-key pitch similarity (Krumhansl, 1979)
While Fig. 1 provides a within-key representation of what our tonal schema might look like, Fig. 2 represents a between-key version of this schema. Commonly referred to as the circle of fifths, similarity between keys is once again represented as a function of distance. Similar generally share more common notes. For example, the scales of C and G major differ only in the pitches of F and F#, whereas the C and F# major keys only share the notes of B and F/E#.
Fig. 2 The circle of fifths (Tan et al., 2010)
Implicit memory and priming
There's a small section on implicit memory, which I will cover briefly. Essentially, there seems to be evidence for harmonic priming, such as when chords are processed faster when participants are first provided with another harmonically-related one (Tillmann et al., 2003), and also for repetition priming, such as quicker response times when participants are made to sing the last pitch of a melody if the to-be-sung note is encountered in the first part of the melody than if it wasn't (Hutchins & Palmer, 2008). The general explanation here says that perceiving a pitch/chord leads to quicker access to similar pitches/chords from memory. Simply put, the initial auditory stimulus 'primes' the individual for faster processing, recall, and recognition of subsequent musically-related stimuli.
Neural bases of pitch and melody perception
The final part of the chapter summarises neuroscientific findings for melody and pitch recognition. Broadly, a variety of methods (e.g., fMRI, ERP, lesion studies) find that: a) pitch perception involves activation in the left temporal lobe areas and b) identifying tonal context involves more right hemisphere activity. More generally, some have argued that the left hemisphere is more concerned with the more specific features of pitch, whereas the right hemisphere is specialised for global processing of melody. For instance, a review by Peretz & Zatorre (2005) suggested that melodic contour is more dependent on the right hemisphere, while the recognition of specific intervallic relationships is more associated with the left hemisphere.
Concluding remarks
So far, we've been concerned mainly with melody perception. However, it is essential to note that the research discussed so far primarily focuses on melody in relation to more frequency-related features, such as pitch and intervals. Melody does not simply constitute WHAT is being played, but also HOW, and more specifically, for how long. Hence, in the next chapter, we'll focus more on aspects related to musical time, such as tempo and rhythm.
References
Dowling, W. J. (1978). Scale and contour: Two components of a theory of memory for melodies. Psychological Review, 85(4), 341–354. https://doi.org/10.1037/0033-295x.85.4.341
Dowling, W. J., & Bartlett, J. C. (1981). The importance of interval information in long-term memory for melodies. Psychomusicology Music Mind and Brain, 1(1), 30–49. https://doi.org/10.1037/h0094275
Hutchins, S., & Palmer, C. (2008). Repetition priming in music. Journal of Experimental Psychology Human Perception & Performance, 34(3), 693–707. https://doi.org/10.1037/0096-1523.34.3.693
Krumhansl, C. L. (1979). The psychological representation of musical pitch in a tonal context. Cognitive Psychology, 11(3), 346–374. https://doi.org/10.1016/0010-0285(79)90016-1
Levitin, D. J. (1994). Absolute memory for musical pitch: Evidence from the production of learned melodies. Perception & Psychophysics, 56(4), 414–423. https://doi.org/10.3758/bf03206733
Peretz, I., & Zatorre, R. J. (2005). Brain organization for music processing. Annual Review of Psychology, 56(1), 89–114. https://doi.org/10.1146/annurev.psych.56.091103.070225
Schellenberg, E. G., & Trehub, S. E. (2003). Good pitch memory is widespread. Psychological Science, 14(3), 262–266. https://doi.org/10.1111/1467-9280.03432
Tan, S., Pfordresher, P., & Harré, R. (2010). Psychology of Music: From Sound to Significance. http://ci.nii.ac.jp/ncid/BB01824497
Tillmann, B., Janata, P., & Bharucha, J. J. (2003). Activation of the inferior frontal cortex in musical priming. Cognitive Brain Research, 16, 145–161. https://doi.org/10.1037/e537102012-736
No comments:
Post a Comment