The basic idea of tempo is simple: how many beats there are per minute. More beats per minute means the music is faster, fewer beats per minute means the music is slower. The image below shows a tempo map of “Dear Prudence” by the Beatles that I made with Ableton Live.
The song’s tempo ranges between 70 and 90 beats per minute over the course of the tune, with noticeable speeding up at the end.
For contrast, here’s a tempo map of the first minute and a half of “God Make Me Funky” by the Headhunters. As you can see, the Headhunters keep much steadier time; they only fluctuate between 92 and 95 BPM.
So, beats per minute, there’s not much more to tempo… or so you would think. There can be some confusion about what constitutes a beat! Before we get into that, though, let’s briefly touch on the Italianate tempo markings that European classical composers used. These markings long predate the invention of the metronome, so they are necessarily subjective. Also, they don’t just tell the performer how fast to play; they also give indications of feel. Lento and Andante describe the same beats-per-minute range, but Lento is more somber and formal, while Andante is more animated.
- Largo – slow and broad (40–66 bpm)
- Adagio – slow with great expression (44–66 bpm)
- Lento – slow (52–108 bpm)
- Andante – at a walking pace, moderately slow (56–108 bpm)
- Moderato – at a moderate speed (108–120 bpm)
- Allegro – fast and bright (120–156 bpm)
- Vivace – lively and fast (156–176 bpm)
- Presto – very fast (168–200 bpm)
European classical tradition and Anglo-American popular music are different in many ways: instrument textures, form, harmony, and so on. The biggest difference, though, is in tempo. Classical music treats tempo as an expressive dimension of the music. Performers and conductors will vary the tempo widely from one phrase to the next, or even from one note to the next. Even at a steady tempo, classical musicians use a lot of rubato. The word means “stolen”, because you might slow down a bit at the beginning of a phrase and speed up at the end of it, so the beginning is stealing some of the ending’s time.
Pop might use a little rubato here and there for intros and endings and occasional special effects, and one pop song in a thousand will have a conspicuous tempo change, but otherwise, musicians aim to keep things metronomic and steady. The Beatles might have intended to speed up a bit over the course of “Dear Prudence”, but probably not, it’s just what happened. At a measure-to-measure level, they were certainly trying to maintain a consistent groove. Pop musicians do use subtle microtiming within the metronomic framework (swing, rushing, dragging), but that microtiming is only meaningful because the underlying grid stays consistent.
If pop music hardly ever uses tempo changes as a structural or expressive feature, why do you so often hear people describe songs as “speeding up”? The reason is that non-musicians (and some musicians) confuse tempo with density. Listen to “We Found Love” by Rihanna and Calvin Harris at 1:44, which is a typical EDM buildup section. The snare hits every quarter note, then every dotted eighth note, then every eighth, then every sixteenth. This increase in density combines with the rising pitch of the synth and the rising cutoff frequency of the filter to create tension before the drop. But the track is not speeding up at all, it’s just getting busier.
For a more subtle version of this phenomenon, listen to “Up For The Down Stroke” by Parliament. The first section is at about 93 BPM, with hi-hats on the eighth notes. At 2:14, the groove changes, so now there are hi-hats on every sixteenth note (and there’s less swing.) The greater frequency of drum hits might make this part seem “faster,” but the tempo stays exactly the same.
In more recent dance and hip-hop, there can be some significant confusion about what the tempo is, because it’s not necessarily obvious what constitutes a “beat”. This has always been true; some people might want to count the Parliament track above at 47 BPM or 186 BPM. Most people intuitively feel the beat as being as close to 100 BPM as possible, but not everyone does all the time. As the average tempo of grooves get slower and the time feel gets more subdivided, it gets to be more ambiguous. Dubstep and trap play on this uncertainty often. Dubstep does it more conspicuously, switching to halftime feel during the beat drop, implicitly halving the tempo.
Trap keeps things steady, but some elements always imply halftime while other elements imply doubletime. This ambiguity is present from the early stages of building a track; most producers who want a beat at 65 BPM will set their DAW session tempo to be 130 BPM and will do everything in halftime. This was originally a practical consideration – it’s easier to program 16th notes than 32nd notes. But it also bakes in this uncertainty about which power of two is the most salient. Is “Versace” by Migos at 65 BPM or 130 BPM?
It’s worth zooming out for a second to consider the difference between the actual rhythms played by instruments in a piece of music and the implied pulse. In dance and pop, you often have some drum or percussion sound on every beat or subdivision, but the hi-hat pattern doesn’t necessarily correspond to the listener’s sense of the pulse. The hi-hat might play twice for every pulse, or it might play on every other pulse, or three times on every pulse, or on every third pulse.
The rhythm of the drums (or other instruments) is the same for every listener, but the pulse implied by that rhythm is subjective. Unless the rhythm is very complicated or uneven, listeners can usually tap out a steady pulse, but different listeners might detect different pulse rates. The pulse rate that a listener experiences is called the tactus. Peter Martens explains that performers and composers might have one tactus in mind, while listeners experience a different one. The bottom number in time signatures reflects what the composer considers a beat to be, but I am living proof that listeners often can’t hear the intended time signature.
Last thing: during my lifetime, there have been two big shifts in pop music tempos, one conspicuous, one subtle. The conspicuous one is the steady slowing of average tempo. E. Glenn Schellenberg and Christian von Scheve found that the average pop song tempo decreased from the 115-120 range in 1960 to the 90-100 range in 2010. The trend has only continued since then due to the popularity of trap, with its extremely slow tempos. The trend continues backward in time too, since jazz uses faster average tempos than rock, R&B or hip-hop. As the tempos got slower, the pulse moved from eighth notes to sixteenth notes. Slower tempos provide more opportunities for subdivision. What does it mean? I don’t know, but it must be significant.
The subtle shift is the move from imperfect human timekeeping to perfect computer timekeeping. This doesn’t just reflect the growing popularity of electronic dance music and hip-hop. Rock, country and other non-electronic styles are also often recorded to a click track or drum machine, or are quantized after the fact. At this point, if you hear uneven timekeeping in a commercial recording, it was a deliberate stylistic choice. Note that computers did not force this change upon us. You can easily program your click track to change tempo, or to follow the tempo of a live performer. If you look at my Ableton tempo maps at the top of the post, consider that it would be effortless to force the Beatles’ performance to be perfectly metronomic, but it would also be easy to have electronically produced elements of a song follow the Beatles’ performed tempo. People just like robotic timekeeping, or at least, the music industry thinks that we do.
Comments are closed.