Memories are generally classed in two main categories. Episodic memory is the recall of certain events or facts personal to us, such as what we had for breakfast. Semantic memory, meanwhile, involves shared ‘facts about the world’ – the meanings of words, say, or knowing the capital of France. Both types of memory may persist for a long time: we might, for instance, retain an episodic memory of our first day at school. Both types are also involved in processing music: we know that this song was written by Duke Ellington (semantic), but also that we first heard it played by Stan Tracey at Ronnie Scott’s (episodic). There is also a third category: procedural memory, which is remembering how to do something, like play the saxophone.
Each of these memory types can be hit or preserved selectively, and so the effects can be quite specific to different forms of dementia. Some people might be able to recall how a tune goes but not what it is called. Alzheimer’s, for example, generally does more damage to episodic than to semantic or procedural memory. One study compared brain scans of people with Alzheimer’s with young healthy subjects as they recalled familiar songs. The results showed that the activity of parts of the cortex linked to that long-term memory was relatively unaffected by the disease.
Other conditions can disrupt our emotional connection to music. People with dementias due to frontotemporal lobar degeneration (FTLD) often find it hard to identify the emotion expressed in others’ faces or voices – happiness, sadness, fear or anger – and they seem to share a difficulty in reading those qualities into music that most unimpaired listeners would identify. On the other hand, some types of FTLD can enhance the emotional impact of music, creating an abnormal craving for it that is dubbed ‘musicophilia’. This might be because, while a person’s ability to process the strictly musical structures is relatively intact, FTLD degrades their capacity to place this in an appropriate social context: they can’t so easily infer the social signals they get from others. That, after all, is a common attribute of obsessive behaviour.
Memory and musicality are complicated things on their own, let alone in relation. It’s straightforward enough to argue that without some kinds of memory there would be no music, but one might go further and argue that memory makes music as biologically and culturally inevitable as language. Memory is, at the most basic level, an evolutionary necessity, because it is what lets us exist effectively within our environment and take advantage of it, allowing organisms to use past experience to ready themselves for the future. Even bacteria, at some level, do that. Memory is what lets us, as humans, see patterns and regularities, and in that way anticipate what will come next. Music plays with this pattern-seeking instinct, this human delight in solving puzzles. But it is, and must be, a shared enterprise: a product of the group and the culture. Then it becomes a receptacle for communal memory and experience. For people like Joyce, it may supply a non-verbal means of maintaining a connection to the past and present of ourselves, our friends and family, and our society.