Did Lana Del Rey copy Radiohead?

This week, Lana Del Rey stated that she is being sued for copying Radiohead’s 1992 song Creep in her 2017 release, Get Free.

It’s true about the lawsuit. Although I know my song wasn’t inspired by Creep, Radiohead feel it was and want 100% of the publishing – I offered up to 40 over the last few months but they will only accept 100. Their lawyers have been relentless, so we will deal with it in court.

— Lana Del Rey (@LanaDelRey) January 7, 2018

First, some facts…

• Both songs use the same chord sequence: | I | I |  | III | III |  | IV | IV | iv |  iv |
  • Creep is in G major, so | G | G | B | B | C  | C | Cm | Cm |
  • Get Free is in Bb major, so | Bb | Bb | D | D | Eb | Eb | Ebm | Ebm |
• They are both mid-tempo (Creep is around 92 BPM; Get Free is around 102BPM).
• They both have a similar rhythmic feel – straight 8s 4/4 time, in 8-bar sections (this is a similarity but an unremarkable one, given that it applies to a huge number of songs).

…and some history…

• Creep is part-borrowed from Albert Hammond’s The Air That I Breathe (1972) – later a hit for The Hollies. According to The Guardian, Radiohead gave Hammond and his co-writer Mike Hazlewood a credit in the Pablo Honey album liner notes.

Here are the three songs in reverse order of release:

• Get Free (2017)
• Creep (1992)
• The Air That I Breathe (1972)

https://www.youtube.com/watch?v=yzP4kQhlPBY

Some points to consider while listening:

• There are some arguable melodic similarities, most obviously in bar 3 of Creep’s chorus (and bar 3 of Get Free’s pre-chorus). Listen to [0:47]-[0:58] in Get Free and [1:00]-[1:09] in Creep.
• The Air That I Breathe’s verse has a subtly different chord sequence from Creep’s, because it resolves to the I chord in bar 6 of the 8-chord loop. It’s in Bb major, so compares to Creep as follows:
  • | Bb | Bb | D | D | Eb | Ebm | Bb    | Bb    | (Air That I Breathe)
    | Bb | Bb | D | D | Eb | Eb    | Ebm | Ebm | (Creep)
• Listen to Creep at [2:26]-[2:35] – Yorke’s falsetto melody over the guitar solo – and compare to Air That I Breathe [0:21]-[0:32]; this is the melodic section that is most similar between the two (and the probable reason for the Hammond/Hazlewood dispute and subsequent credit).
• One key point of argument for Del Rey’s side could be the extent to which elements of Creep are themselves protectable – the chord sequence particularly – and whether the apparent melodic similarities provide evidence of copying.
• I won’t comment here on whether I personally believe that copying has taken place, but rather will focus on the likely arguments that each side will use.

Radiohead’s side will want to assert that the allegedly similar elements of Creep (the chord sequence, and parts of the melody) are protected by copyright, and that the apparent similarities demonstrate that Get Free copies a ‘substantial part’ of Creep; Del Rey’s side could argue that the chord sequence is found in other works (e.g. Air That I Breathe and David Bowie’s Space Oddity to name two), and that the melodies are not exactly identical at any point (in terms of pitch and exact rhythmic placement of notes). They may also make a case for ‘independent creation’ – that it’s possible that the two groups of songwriters could have come to similar-sounding creative outcomes separately.
In listening, the question isn’t simply “do these songs sound similar”? It’s possible for two recordings to have similarities without necessarily infringing copyright (for example, if the similar elements in the defendant’s work appear in other songs that predate the plaintiff’s work). And independent creation might be a usable argument for LDR’s team. It’s reasonably easy to demonstrate that the chord pattern | G | B | C | Cm | was not originated by Radiohead, so the implicit question is this; if two separate songwriters were playing these chords at around 100BPM with the same groove and 8-bar form, what’s the likelihood that they might make similar melodic decisions independently of each other? Independent creation does happen, but so does inadvertent plagiarism, especially when the plaintiff’s work is very well known. Ascertaining the difference, when you have only the finished works as your evidence base, is rarely a simple process, tied up as it is with the ‘usual’ epistemological challenge of songwriting creativity research – i.e. it’s pretty much impossible for anyone to infer creative intent only from listening to music (Bennett 2013).
According to LDR’s tweet, the parties have been unable to reach an agreement, so it seems set to go all the way to court… and presumably then into the MCIR database.