The Melody of Hypnotize: Looking at Biggie’s Masterpiece From a Fresh Perspective
It’s a little odd that the most popular genre in the world over the past few decades is still challenged over its right to call itself “music”. There are some who still argue that rap is not a genuine form of music because it supposedly lacks certain elements like melody or harmony. There are already very good arguments that explain why this assertion doesn’t work, and it’s not an argument I think is worth addressing anyway since hip-hop easily stands on its own merits and doesn’t need approval from the likes of Ben Shapiro. However, I think there is an interesting question being implied in this argument: is rhythm the only musical element that we should look for in rap?
A majority of the analysis I am aware of on rap music focuses on its poetic elements and cultural relevance. Producers often talk about the musical elements they use for their craft, but vocal performers don’t seem to have a formal framework for the decisions they make when performing their music. This is mostly because there is no demand for such a framework: rap is primarily an aural tradition and formalizing it risks missing some of the identity that makes it unique as an art form. However, considering how much experimentation and growth came from jazz composers who thought seriously and analytically about their music, I think it may be worthwhile to begin looking at rap in a formal way. If nothing else, it can give us a greater appreciation for the artistry that goes into making it.
The purpose of music theory in my mind isn’t to tell musicians how they ought to write their songs, but rather to make connections between form and meaning so we can better understand the “language” of music. Probably the most common example of this is the idea that the major scale sounds “happy” and the minor scale sounds “sad”. There are also conventions like functional harmony which set up expectations for the listener that composers can play around with. As someone who often listens to rap music on a primarily musical level, I take it for granted that rap has musical elements that portray meaning outside of the lyrics. I hypothesize that there are underlying patterns that rap artists use intentionally or implicitly in their music, and that it is possible to analyze these elements in a systematic way that allows for a deeper understanding of the art form.
There has already been some work on this. The earliest I am aware of comes from Martin Connor, a music teacher from Philadelphia. Most of the work he has done focuses on transcribing the rhythms of rap musicians. The rhythm of a vocal performance can be understood in terms of its relationship to the beat, poetic elements, and prosodic elements like inflection, tone, and stress. For example, a rapper may emphasize end rhymes by placing them on strong beats and putting extra volume and stress on those syllables. A separate attempt to create a highly-detailed systematic analysis of rap music was undertaken in 2015 by musicologist Nathanial Condit-Schultz. His MCFlow corpus includes 124 transcriptions of rap songs, all done by ear. These transcriptions record information about phonetics, stress, phrasing (prosodic boundaries), intonation, rhythm, and rhyme.
Most interesting to me personally is the more recent work being done on intonation and pitch. The human voice is an incredibly versatile and expressive instrument. It feels highly reductive to me when people make the assumption that rap is “only” percussion. Kanye West reportedly focuses on making sure his raps have a memorable “melody”, whereas Brother Ali visualizes “the delivery and the tone and the pitch” of his raps before even writing lyrics to them. In his comprehensive doctoral thesis on rap flow, Komaniecki places pitch techniques on a spectrum from more song-like to least song-like. By using relative differences in pitch, rappers can emphasize important words and rhyme schemes, create quasi-melodic motifs, or add layers of rhythmic complexity to their music. Komaniecki transcribes rap rhythms with individual syllables notated as either relatively high or low in pitch. In contrast, Doctoral Fellow Ethan Hein has taken a more precise approach of measuring the exact pitch of each syllable. This allows him to closely analyze the pitch contours of entire phrases and the magnitude of pitch contrasts. These analyses demonstrate that rappers use a diverse array of pitch-based techniques to develop unique flows.
Inspired by these efforts, I decided to experiment with my own framework. Using the free language analysis software Praat, I set out to record the fundamental pitch of each syllable as its closest note and transcribe that as a piece of sheet music. In theory, this approach would combine the utility and regularity of the Komaniecki system with a level of detail closer to that of Hein’s. This allows for rhythm, pitch, and linguistic elements to be observed as a cohesive whole. It should be consistent enough to compare different artists to each other, while still being flexible enough to incorporate additional information on dynamics, timbre, or anything else that may be theoretically relevant. It is also possible to automate, meaning that a large corpus could be created with a relatively small amount of labor. This is at a cost to simplicity and accuracy, but those parameters can be adjusted to a certain extent.
For this initial analysis, I decided to transcribe the first verse of Notorious B.I.G.’s hit single “Hypnotize”. Biggie is known for his iconic, smooth flow and catchy lyrics. I predicted that his music would exhibit melodies presented in a virtuosic way that could serve as a baseline for other analyses. The rhythms of Hypnotize had already been transcribed by Martin Connor, so I chose it as the most expedient option. To make the results as consistent and repeatable as possible, I measured the average pitch at each vowel, so that each syllable would only have one pitch regardless of movement in pitch across the syllable. If a pitch was in between two notes so that it was not immediately apparent which note the syllable should be written as, I would transcribe it as a quarter tone. Quarter tones split the octave into 24 rather than 12 notes so that the note between A and A# would be A half sharp and the note between A and Ab would be A half flat. For notation I used the common Arabic system where half sharp is represented with a hash (#) sign with only one vertical line, while half flat is a represented with a backwards flat symbol, looking like a “d” rather than a “b”. This allowed for a greater level of precision while still keeping the score readable.
The results of this method are… interesting. After listening to the melody enough times it does start to sound a bit like the song, but from the offset it’s clear to me that a lot of important information is lost in the transcription process. That being said, there are still some insights that can be taken from the data. In general, Biggie does demonstrate a degree of expressiveness through pitch, and is certainly not rapping in “monotone”. The majority of notes land between C and F, the range of a perfect fourth. A is by far the most common note, and I see it as a sort of center of gravity or baseline that other tones can contrast with. It appears to sit roughly in the middle of the range, suggesting some balance between high and low contrasts. The instrumental of this song is built on an E minor chord, and A is the fourth scale degree of E minor. This may or may not be relevant, but I would not be surprised if rappers harmonized their vocals with the instrumental to some degree.
The first thing that really popped out to me doing this analysis was the number of motifs in the song. Motifs in this case are not a specific set of notes, but rather a sort of general note shape that repeats throughout the piece. For example, the pink motif is generally a 1/16th note followed by a 1/8th note at a descending pitch, then a 1/16th note rest. The most incredible motif to my view is the burgundy example, which not only retains the same shape but also the exact same notes all three times it occurs. The dark green motif is also the same pitch the first four times it’s used. Obviously the pitches are rounded to the nearest note and not exactly the same for all three examples, but the fact they are so close is impressive in itself and suggests to me that motifs do exist in Biggie’s performance.
The light green and blue motifs span about two-thirds of the song, whereas the pink motif covers three-fourths. Counting the burgundy and red motifs as a single motif has some validity in my opinion, since the only difference between the two is the pitch contrast between the first and final notes in the overall phrase. If we do this, the motif also spans about two-thirds of the song. All of these motifs are introduced in the first four bars, with only two additional motifs which are introduced in the last four bars. All of the motifs introduced before these final ones are only a quarter note long. The gray motif is three quarters notes long, and the gold motif is two. It’s possible that a large number of short motifs early on create a burst of momentum, and the repetition of material lets the listener get used to the melodic concepts which are then expanded upon with new material as they progress through the verse.
Motifs tend to be placed close together and repetition is common. The first string of motifs is two and a half bars long, and afterwords the motifs are almost always strung in pairs or triplets. Motifs placed by themselves are never more than a quarter note beat away from the next motif. Spaces between motif clusters are generally a bar apart. This regularity is fascinating to me, and suggests some consistency to the usage of motifs.
The melodic contours do not always match up to what you would expect poetically or prosodically. Stress is usually associated with higher pitch and volume, but often times stressed syllables land on notes that descend pitch-wise. For example, in the fourth bar the rhymes highlighted in red are all highly stressed, but melodically they encompass the highest and lowest notes in the phrase. Low notes tend to denote the ends of phrases, but the second and third phrases actually end on their highest notes. Sometimes phrases start and end in the middle of motifs; other times the motifs seem to dictate the phrasing. There appear to be important relationships between melody and the other elements of the song, but the melody itself functions independent of any of those relationships. I think this within itself proves that rap does have an inherent melodic component, albeit one that cannot be accurately modeled using the traditional Western pitch system.
The results of this investigation far exceeded my expectations. I did not expect there to be genuine melodic motifs, especially ones that repeat so often and fill up so much of the song. This is a result that would not have been nearly as apparent using the methods already developed by other researchers. I think this method has a lot of potential, and I’m hoping other musicologists begin to adopt it and move towards more sophisticated analyses of rap music. I’m happy with the results of the 24 note scale, but I imagine other notation systems would reveal different insights and are worth experimenting with. I also think it would be interesting to automate the transcription process to aid in compiling large datasets. This is entirely feasible, albeit at a cost to precision.
There are a few things this system could handle better. Most notably, I feel that accuracy suffered immensely in the last bar. As P. Diddy’s backup vocals overlapped with Biggie’s, it became more difficult to accurately measure a single pitch. One way to handle this is to allow for more transcribing by ear rather than an entirely “objective” method. For the sake of repeatability I only reported the pitches I got through PRAAT, although in some cases I found the “objective” pitches to be different from what I was hearing subjectively. It is also essential that systems for analyzing multiple pitches are found, since doubling vocals is such a common feature of rap music and may in fact have its own properties. Based on the preliminary results from this analysis, I believe it is likely that harmonies exist in rapping just as melodies do. The cross over vocals between Biggie and P. Diddy resulted in the lowest transcribed note by far, suggesting that Diddy was rapping at a different, perhaps lower, vocal register than Biggie. With features like “hype” vocals at play, polyphony may be a far more prominent feature in rap music than one might expect.
The wonderful thing about using sheet music is that it is extremely flexible, and individuals can transcribe music to any level of detail they like. Systems for measuring volume, pitch bends, glissandi, timbre, articulation, and other features are something I really look forward to seeing in other people’s work. I highly encourage anyone reading this to try out this system and experiment with it. If you have any questions or need help, do not hesitate to reach out. I think this is a very exciting time for music theory. There’s a whole world of music out there that has barely been explored; these are only the first steps on a fresh new journey into the unknown.