Putting more than one song together to make a bigger structure is a standard part of the arranger’s craft. There are all kinds of interesting things to think about the how you join the songs together, but today my interest is on the more fundamental level of how you choose them.
I think about medleys in two types. One is a collection of tunes that share a common origin or set of associations. Say, selections from a show, or songs by (or made famous by) a particular artist or group, such as my Madonna Medley or Meatloaf Medley. I tend to think of these as show pieces, useful as they offer a longer span of musical time without a break than single songs would, in much the same way that classical concert programming places more substantial works as focal points amongst the shorter items.
The thing about this kind of medley is that, in order for them to make sense to the listener, they need to know what the connection between the songs is. You’ll be offering a variety of key and tempo and feel over the course of the medley to offer interest, but there won’t be any inherent narrative connection between the lyrics of the different songs. So the genre tends to pick pretty mainstream, well-known material, and you’d also tend to introduce it in performance so that anyone who didn’t know the originals would still know to listen to it in this mode.
The other type of medley is the medley-as-self-contained artwork, in which the songs are chosen to have a specific narrative connection such that they work together irrespective of whether the listener knew either of the songs before. This is the kind of medley specified in the culture of barbershop contest, in which medleys are judged as if they are individual songs - that is, with expectations of having a balanced form, which offers both unity and variety.
These kinds of medley may still draw on material from a common origin - Mary Poppins, The Wizard of Oz and Oliver have all been sources for arranging commissions for me over the years. But they tend to be more tightly constructed, reprising songs to create a larger scale ternary-style form, or linking introduction and tag to bring a sense of closure and self-containment to the structure.
When discussing a potential new commission recently with a chorus director (look out later this year!), we found ourselves juggling some interesting factors relating to these questions of origin and audience knowledge. The starting point was a selection of songs all made famous by a single artist, but my feeling was that, whilst two of them would fit together to make a coherent narrative, the third (which was in some ways the strongest as a barbershop contest vehicle) only made sense if knew that they were all sung by the same person. So it wouldn’t have that sense of autonomy, of being a self-contained entity.
The next possibility we discussed was a medley of this third, strongest song with a tune from a different artist. This made great narrative sense, but actually gave us the inverse problem. If you knew both originals, the artistic worlds of the two artists were sufficiently different to be a significant distraction. Different styles of personal presentation, appealing to different fan bases, associated with different kinds of images in wider popular culture. Though I’m sure both artists would have had a great deal of respect for each other, it would be making the audience do too much work to imagine them both into the same medley.
So, here’s the take-away about external knowledge: it mustn’t be required for the medley to make sense, but if it is known, it mustn’t get in the way.
Another thing we discovered during this conversation was that some songs are so strong that they insist on being the prime song of a medley, they won’t let themselves be subordinated into second position. Two songs may have some fantastic lyrical or thematic link, but if they both want pole position, they won’t work together in the same medley.
The last point I’d like to make is actually the one I thought I was going to be making first, as it’s something I’ve been thinking about for many years already, but don’t think I’ve written about before. It’s about how the songs in a type 2 medley work best when they have not only some kind of narrative/thematic link, but the second song actually changes the meaning of the first, so you hear it differently on reprise.
One of the best examples of this I’ve heard is Paul Davies’ medley of ‘Do You Know What it Means to Miss New Orleans’ with ‘Basin Street Blues’. When the first song comes back, you have a much greater sense of what it is you miss. And of course, Cambridge Chord Company first brought this to context the season after Hurricane Katrina, showing how real-life events can themselves change the meanings of songs we thought we knew...