I’ve recently discovered the American song-poem. This was a form of vanity publishing songwriting that was most common between the late 1960s and early 1970s. The lyrics were written by amateur poets, who would respond to small ads by music production companies promising to set them to music for a fee. There is an excellent collection of materials and information about the genre at the American Song-Poem Music Archives.
These (mostly awful-sounding!) recordings are useful in my songwriting creativity research, because they demonstrate the relationship between process and product, and also because they are created using a known process constraint (e.g. Bennett 2012). All the songs ostensibly took a lyrics-first approach (in that the poem was sent in to the production company) but in many cases it is obvious that the singer is crowbarring the lyrics into the melodic phrase. This suggests to me either that the songs were done in one take with no rehearsal (unlikely, given the comparative sophistication of some of the arrangements) or, more probably, that the backing tracks were pre-recorded and then re-used many times with different lyrics/clients. The lyric scansion on many of the songs is horrendous, with rushed phrasing to get to the end of the line and some phrases that start or finish at a point in the bar that is obviously ‘wrong’.
But the session singers are competent and the backing tracks well-made, in an easy-listening kind of way. You can sense that the recordings were made in a hurry – there are moments when the keyboard and bass are playing different chords from each other on the downbeat, and the band corrects the mistake in real time – the chords are usually correct by the middle of each bar. These are competent pro session players, probably working to very small fees and tight timescales. The strings are mostly played on a Mellotron by the sound of it (and the low-budget approach suggests this was likely).
I find a strange beauty in the mismatch between obvious doggerel in the lyrics and the cheesy professionalism of the backing tracks. And in some cases the vocalists are clearly trying to work too many syllables into a backing track that can’t hold them all in each phrase. Here’s Rat A Tat Tat, America. Listen at [0:53] where the ‘marshy bog’ line has to be added after the melodic line has finished.
Some of them work quite well, possibly more by accident than design, or at least by some excellent crafting by the session singers. Most of the scansion in ‘I Like Yellow Things’ works pretty well, although the B section is a tougher listen than the lists in the verses.
Aside from the scansion (which I suggest becomes a valid musical gesture of its own after you listen to three or four of these songs back to back) I’ve also become intrigued by the melodies. We know that these songs were recorded to a budget and to a deadline, and that the session singers were probably working to a pre-recorded backing track with a minimum of rehearsal. So they’re going to be looking for melodies that work reliably with the chords, and are less likely to take melodic risks. The melodies have two characteristics that suggest a first-take approach to pitch. They rely heavily on static notes, often the root note of the current chord (as in I Like Yellow Things). And they repeat melodic phrases without the variation that would be needed to maintain interest. For an example of over-using the same melodic phrase take a listen to Do You Know The Difference Between Big Wood and Brush? The melody has a good ascending chromatic line that sings well for the first four notes (the words ‘do you know the…’). My guess is that this phrase was written at the same time as the backing track, and the singer then used this basic framework to (try to) scan the new lyric over it later. But we then get the ugly scansion of the four syllables in the words ‘difference between’ (more crowbarring), and the singer rather apologetically repeats the root note of Bb for the last three of these in order to hit the start of the next ascending phrase over ‘big wood and brush’. The chromatic ascent (in the key of Bb, the notes are D-Eb-E-F) is occasionally abandoned on the fly in favour of a diatonic approach (D-Eb-F) in order to get the lyric phrase to fit.
In some songs the singer appears to abandon the idea of a properly constructed melody, and simply sings up and down the scale until the lyric line comes to an end. In the case of The Moon Men, I get the impression that the band (drums, bass, sax, electric piano) is following the singer. I also note that this track seems to have early 80s production values (that might even be a Dx7 keyboard part?), suggesting that the practice continued beyond its apparent peak in the late 60s and early 70s.
The lyrics are perhaps the least interesting aspect of song-poems to analyse, being mostly written by beginners. It is worth observing one common trait, which is the tendency to write about subjects with no emotional content. I Like Yellow Things is a good example – the title says it all, and the lyric is simply a literal list of yellow things. Here are another couple of examples of the same phenomenon – Green Fingernails (sung with wonderful sincerity by Gene Marshall) and Listen Mister Hat.
And sometimes there was no melody at all – several of the recordings are just poems spoken over music. Run Spook Run (performed by the master of the genre, Rodd Keith) is particularly interesting because it contrasts the beat-poem approach of the verses with a properly arranged titular chorus, complete with Andrews Sisters style backing vocals.
There were times when the session singer was so unrehearsed or rushed that different choruses have different melodies. Compare The Palace Roses at [0:11] and [1:46] – same lyrics, entirely different melody. It’s clear that by the time singer Todd Andrews got to the final chorus he had forgotten his own melody from the first one!
Commercially, Song Poems could be seen as a nasty scam designed to extract money from gullible amateur songwriters. But these recordings show that there are some cases where, within the constraint of terrible lyrics and time-limited studio sessions, the production companies are doing their best to make something that sounds good. Whatever you think of the material (and I think the most we can say is that it has arguable kitsch charm) it’s an interesting insight into the creative process of word-setting and the psychology of vocal improvisation.
It seems apt to end this blog post with Rodd Keith’s strangely prophetic I Died Today, in which he apparently predicts his own death…