You Are the Music: How Music Reveals What it Means to be Human (20 page)

‘The star’ – all of us

Amazing musical memory is not just the preserve of worldclass performers. My musical memory as a performer may be poor but there is nothing wrong with my memory for 1960s Motown, the Beatles’ back catalogue, or Beethoven’s symphonies. I can play them all in my head anytime, anywhere.

When I tell people that I study musical memory they often ask me why they have such a great memory for music and lyrics but can’t remember important dates or all the items on their shopping list. What makes musical memory so powerful by comparison?

Let’s start by blowing this illusion apart. Our memory for music on first hearing it is pretty terrible. Andrea Halpern and Daniel Müllensiefen
¹³
presented 63 undergraduate students with 40 unfamiliar short tunes. They then tested memory for these new tunes when they were buried among 80 test tunes. Memory performance across the undergraduates was only barely above chance; this was a difficult task.

Changing the instrument during the test made memory judgements even worse, indicating how fragile and shaky our first memories for new music can be. This result also suggests that when we first hear music we are particularly attracted to the ‘surface’ features, the nature of the actual sound or speed, and pay less attention to the deeper structures, harmonies and/or rhythms that are present.

To give you an idea of how poor this level of memory performance really is, we can compare it to our memory for photos. In 1973 Lionel Standing
¹⁴
presented volunteers with a collection of up to 10,000 novel photographs. He was hoping to demonstrate the limits of the long-term memory system and you would assume that with a maximum of 10,000 images he had gone rather overboard. In fact, he found an amazing level of performance and demonstrated that our memory for images is far superior compared to memory for words. People
who were shown all 10,000 new images were estimated to have retained 6,600 when tested only two days later.

How do we reconcile our poor initial memory for new music against the impression that we all have large and accurate musical libraries in our heads, filed with thousands of tunes, songs and instrumental pieces?

Research carried out on our long-term musical memories says we are right to have faith in our mental musical store. When we test recall of familiar and favourite music we see evidence for the amazing speed and accuracy of musical memory, even in individuals who have not had a single day of musical training.

Daniel Levitin
¹⁵
asked volunteers to select a favourite piece of music from a large collection of CDs that he had in his lab. Once the volunteers had selected their favourite track they were asked to have a go at singing it. Levitin recorded their performance and measured it for accuracy against the original CD. The majority of people demonstrated ‘absolute memory’ for their favourite recording, meaning that they were able to reproduce the first pitch of the track within a very small margin of error.

What is more, this finding has recently been well replicated across six different European labs.
¹⁶
On average singers produced their favourite tracks a little flat, but the majority were able to either hit the exact right note of a favourite pop song or at least get within a semitone. Overall, it appears that our memory for our favourite music is very good.

So how do we go from a shaky impression of new music to an almost accurate blueprint of our favourite pieces in memory? The keys are time and exposure. To go back to the comparison of memory for new music with that for photographs, the difference is that music is complex, multi-layered and evolves over time. There is just so much more to absorb on the first listen than there is to see in a photograph.

When we first hear a piece of music we are attracted to the surface features of the sound, the qualities of the singer’s voice and the timbres of all the different instruments. We are easily distracted by musical ‘surprises’, such as changes in key or tempo. It is the equivalent of being a kid in a large and exciting sweet shop – so much to enjoy but where to begin and what to try first?

Studies by Jay Dowling and colleagues
¹⁷
have shown how our memory for music develops over time after this first rush of listening excitement. Imagine this process as similar to that which happens when a house is built. Over the first few exposures to new music we put in the foundations of our house and create the skeleton structure: we build an idea of key (major or minor, and how these fluctuate) and the contour, the ups and downs of the tune. Over time, and with more exposure, we begin to add the bricks and windows to our house: we add in musical details such as the exact pitches, the different instruments. Finally we add the furnishings: the micro-variations in performance such as the accent of the singer or minute changes to tempo and volume.

It takes a lot of time and exposure to build a good musical memory but once it is built, like any well-built house, it is strong and long lasting. And unlike building a house, there is almost no effort required on the part of the listener. Your mind does the work for you; all you need to do is to listen.

Boosting musical memory

We remember the music that we hear in different ways. Some pieces of music take longer (i.e. more exposures) to learn well whereas some appear to bury themselves in our memory relatively quickly. One factor that influences the speed and accuracy with which we remember new music is personal preference.

It makes sense that we might better remember what we
perceive to be ‘good music’ compared to ‘bad music’, because memory and emotion have strong links in our minds. The music we like is more arousing, meaning that we experience greater attention capture and engagement, which is then associated with better memory. Positive mood is also associated with better memory for novel music compared to negative mood.

Stephanie Stalinski and Glenn Schellenberg
¹⁸
demonstrated the power of the relationship between musical memory and liking. Fifty-five volunteers rated how much they liked 24 novel musical extracts that each lasted around fifteen seconds. There was then a ten-minute break wherein they filled out questionnaires, followed by a memory test where each person heard 48 pieces of music, 24 from the first phase of the experiment and 24 which were new but similar to the first set. Each person decided whether a musical clip was completely new or something that they had heard before.

The results were clear. Despite our pretty poor memory for novel music, the people in this study could remember enough of the short tunes for the researchers to be able to assess the impact of liking. Musical clips that were liked were remembered significantly better than tunes that were rated as neutral or disliked.

The researchers went further. They wondered what would happen if they tried a longer delay, which is more like what would happen in the real world. They tried a 24-hour delay and found the same result: liked music was remembered better. The effect also held when they tried different listening situations, and even when they told people of the upcoming memory test and asked them to try to remember the music by creating visual memories while listening.

Over time it makes sense that we remember music that we like, as we will clearly choose to listen to this music more often. This research, however, demonstrates that the impact
of our initial reaction to music has a measurable effect on how well we remember what we hear after only one listen.

We also remember vocal music better than non-vocal music. Michael Weiss and colleagues reasoned that we may be more attuned to remember vocal music as compared to instrumental sounds as vocalisations are more biologically significant to us as humans.
¹⁹
They tested people’s memory for melodies that were either sung or played on various instruments, both familiar (piano) and less familiar (marimba, banjo).

The researchers chose 32 unfamiliar folk tunes from the UK and Ireland for their study, which took place in Canada. The short melodies were sung by a female vocalist, who performed all the notes on the syllable ‘la’. The researchers then generated instrumental versions of the vocal version that were perfectly matched in pitch, duration and amplitude. Each participant listened to sixteen of the melodies, four in each type of sound, followed by a questionnaire break and finally a memory test.

Memory was better for the vocal melodies compared to any of the other instruments, which did not differ from each other. Piano sounds were no better remembered than the marimba so we can rule out a simple effect of sound familiarity. The authors reasoned that, similar to liking, vocal music may stimulate higher arousal and attention engagement in listeners, an effect that may be driven by a biological propensity to be attracted to the human voice over musical instruments.

All the experiments in the last section demonstrate that multiple factors influence how well we remember new music, and no doubt there are many influences yet to be discovered and documented. All the factors, including our liking response and apparent attraction to vocal music, contribute to the eventual laying down of strong and long-lasting musical memories.

Name that tune – in 400 milliseconds or less

I keep saying that musical memories are super strong but where is the evidence?

One indicator of the power of musical memory is the ‘plink study’. Carol Krumhansl
²⁰
investigated the processes underlying musical memory by presenting people with the smallest examples of music that you can imagine; 300 and 400 millisecond ‘plinks’. If you go to your favourite music player and play a piece for just one second then you will appreciate how little you really hear in a plink of around a third of that time.

Krumhansl played people plinks of music from the 1960s through to the 2000s and asked them to name the artist and the title as well as the decade of release, style (a broader and more reliable category than genre), and the emotional content. Sounds impossible, right?

The volunteers in Krumhansl’s study had a good amount of musical training (around ten years) and were reasonably avid music listeners (an average of around 20 hours a week), but they were young adults and certainly not experienced musicians or music writers/critics. They listened to 28 musical plinks in total, chosen from top songs in lists by
Rolling Stone, Billboard
and
Blender.
Nearly 30 per cent of the participants claimed to be able to recognise the songs after hearing only a 400-millisecond plink, and of this group 95 per cent correctly identified the title and artist.

The best recognised plinks were Britney Spears’ ‘Baby One More Time’ and Aretha Franklin’s ‘Respect’. Also high in the recognition ratings were The Police’s ‘Every Breath You Take’ and John Lennon’s ‘Imagine’. At the lower end were Louis Armstrong’s ‘Wonderful World’, Coldplay’s ‘Viva La Vida’ and U2’s ‘Beautiful Day’.

The volunteers who spotted the plinks were also good at identifying the release decade, suggesting that this information
is recalled or inferred quickly along with the title and artist. Even when people were not able to spot the songs they were often able to identify emotional content and style. Krumhansl then tried a similar experiment with 300-millisecond plinks. As expected, accuracy went down but remained above 10 per cent.

Overall, the plink study shows that memory for wellknown music is incredibly detailed and reliable; a miniscule extract can trigger recall of the whole song to the extent that people can identify the artist and title, and decade of release. The fact that emotional judgements and style were not always so tightly linked to correct identification tells us that these aspects of musical memory may be stored differently in our minds.

Interestingly, there was also a preference for older songs in the study, from the 1960s and 1970s, even though the participants were young adults. This result could be interpreted as an impact of more distinctive styles and recording techniques on memory; a ‘uniqueness effect’. Or maybe songs from that era were associated with more iconic representations of history that were stronger in general episodic and semantic memory. Possibly there was an increase of exposure effect on memory, with the young adults having heard the music of their parents for longer than their own.

Or perhaps, as Krumhansl puts it, ‘it may be better music’: an interesting and, to my mind, not entirely fruitless argument.

Finally, Krumhansl used her study as a basis for estimating the capacity of musical memory. If you listen to twenty hours of music a week (like her volunteers) then you will hear an average of 22,000 songs every year (based on about three minutes’ length each). Although that includes many repetitions it is still reasonable to assume that an average memory for a listener of this frequency would extend to hundreds of
thousands, if not millions, of identifiable memorised pieces of music over a lifetime.

It’s a musical life

Another indicator of the power in musical memory is its longevity. David Rubin wrote about oral traditions
²¹
; how many folk, children’s and tribal songs have remained in human culture for hundreds if not thousands of years without ever being written down. It appears that music can survive perfectly well relying only on human memory over countless generations.

Musical memory also survives well over the course of a single human lifetime. My grandma is an example of someone with an excellent and longstanding musical memory. She loves to sing and will very often come out with a tune just because you utter a phrase that reminds her of a song. She has a particularly good memory for songs she heard as a child or young woman, a feature of general memory called
the reminiscence bump
.

Studies have shown, in fact, that older listeners retain many important music memory skills such as the ability to recognise alterations to melodies, even though performance is slowed slightly in line with theories of cognitive ageing.
²²
And as we shall see more in the next chapter, the music that provides comfort and stimulates memory most effectively in older people is most commonly the music that has been in their memory for the longest time, from their reminiscence bump.
²³