Initially envisaged as a text to accompany an undergraduate musical composition classes, Bob Snyder‘s “Music and Memory” gives a fascinating overview of the basics of both cognitive psychology and musical structure. Perhaps most interesting of all is his unifying perspective – that “cognitive structure creates constraints on possibilities for musical structure”. Under this model, understanding memory is not simply of passing interest to the musician, but of fundamental importance to understanding music itself.
The book does not concern itself with different types of sensory memory (visual, auditory, motor, etc.) but instead focusses on the interdependence between music and memory. Snyder states that for “music that has communication as its goal, the structure of the music must take into consideration the structure of memory.” At the heart of the book is the premise that there the three difference types of memory (sensory, short term, and long term) can be mapped onto three different levels of musical organisation (notes, melody/rhythm, and form), which occur over roughly equivalent timescales.
The first level of memory – so called echoic or sensory memory – operates on the a timescale of under a second and relates to raw sensory data. For auditory information, vibrations are converted by the inner ear to nerve impulses that represent the amplitude (volume) and frequency (pitch) of the original soundwaves. Basic categorisation of this information occurs before the brain gets involved, and sounds that occur less than a sixteenth of a second apart are perceived as a single event. At this fleeting level, all we can perceive musically are the basic building blocks of notes – pitch, loudness and timbre.
The second level of memory – short term memory (a component of working memory) – operates on a timescale of a few seconds (3-5 sec on average) and allows a number of events to be grouped and held in the consciousness at the same time. Musically this correlates to melodic or rhythmic phrases, and thus the capacity of working memory places a natural limit on the length of musical phrases. Acoustical groupings are determined either by temporal proximity, aural similarity or continuity of movement, i.e. notes that are played in fast succession, sound similar or move in the same pitch direction tend to get grouped together. While in working memory, short chunks of information can be pondered simultaneously and compared with information previously stored in long term memory. The timescale can be increased by rehearsal, i.e. repeating the same material to make it more familiar and facilitate permanent memorisation. Because of the importance of the interaction with long term memory to allow comprehension, what we already know plays a major role in determining what we see and hear. In musical term this means that our ability to appreciate new music is primarily dependent on experience, rather than any inherent property of musical genre; familiar music is easier to understand, and therefore to enjoy.
The third level of memory – long term memory – operates on a timescale of fractions of minutes through to many years, and generally involves chemical or structural changes in the brain. Musically, this correlates to large-scale form, i.e. sections of music, which is not perceived immediately but only in retrospect. “In listening to a whole piece of music, we are only able to consciously understand the relationship between different parts of the piece by having events come back into awareness from long term memory.” Most of the contents of long term memory is unconscious, so reconstruction of large-scale patterns takes much more effort than basic pattern recognition and may require repeated listenings.
The book also highlights the importance of musical categories and boundaries, which are really just constructs of the way our memory allows us to perceive sounds. Snyder describes interpretation as the “management of nuance” – variation within the boundaries of musical categories. A single piece played by multiple different performers is still perceived as being the same piece – with the same notes, phrases and sections – but with different dynamic emphasis, rhythmic accuracy and pitch deviation.
All of this has profound implications for composition as well as performance – such as natural limitations on phrase length and the importance of repetition and structure. Snyder goes as far as to categorise all music into two groups based on their memorability:
- Music that attempts to exploit memory, which contains chunks of similar material separated by clear boundaries and organised into an overarching hierarchical structure that can be efficiently learned. Functional tonality provides ample opportunities for establishing high-level musical architectures, and much of Western classical and popular music falls into this category.
- Music that attempts to sabotage memory, which flouts all or some of the structural principles outlined above, making both anticipation and recollection much more difficult. This includes both information-rich music (e.g. atonal music) and information-poor music (e.g. highly repetitive or slow music), such as contemporary Western experimental music and some religious chants.