A Philosophy of Cover Songs

Front matter

P COVER SONGS -

CS NEN = Es.

To access digital resources including: blog posts videos online appendices and to purchase copies of this book in: hardback paperback ebook editions

Contents

Introduction 1 What is a Cover?

The dictionary definition o- =se crre 0.00 02000. Fivé: problems a 22255 .chee cae mes rx eee E

The histOry Of COVELS «i emen mm Gi cn dee oe ae ged a Early days: esp d oat ge ee Ae sr terrence: eeu See E Hijacking continues... s eate eee Covers that hark back... aaa The lessons of istoty osat dee Da e h e ee

Epilogue

List of Figures

90: N OS DL Ws T9. ups

Twoadsfor'Fugitive ss sosoo eA 15 The first Top of the Pops albumcover.... 2... lees 45 The announcement of contest results . . . 2.0... 2... ee eee eee 49 TheKingBeatS$ gases a He SAG ates SG ees ay aoe dS se 56

To Cristyn, Christy, and Ron.

Their influence runs through the whole book.

Even though the words are mine, we found many of the ideas together.

Acknowledgments

In the late 20th century: With the resource of Napster, my brother Warren put together two dozen versions of the song ‘Stardust.’ Inspired by this, I assembled several similar compilations— 29 versions of ‘Don’t Get Around Much Anymore’, 12 versions of ‘Solitude’, and 8 versions of the theme from Shaft. I had great fun listening to different covers, exploring the potential of each song.

About a decade ago: My friend Christy Mag Uidhir reached out to me, said that he had some ideas about covers, and suggested that we should coauthor something. We worked at it but disagreed on some fundamental issues. My wife Cristyn and I had long conversations in which it became clear that neither Christy nor I had staked out the right position. Distinctions needed to be made. That collaboration led to a paper by the three of us (Magnus et al. 2013). When responding to that paper, Andrew Kania refers to the three of us as ‘the Mags’— he says, ‘to save some space!’ (2020: 238)

More recently: Christy suggested that it was time to say more about covers. Sure, I replied, but what should we say? He didn't know. I brainstormed lots of possibilities. The three of us soon had a draft of a new paper. I was having weekly lunch with Ron McClamrock, who used to play in cover bands. He brought a different perspective to the project and ultimately joined us as a coauthor (Magnus et al. 2022). It remains to be seen how this collaboration will be abbreviated— perhaps ‘4M’ or ‘the Mags feat. Ron McM”

In the course of this collaboration, Icame up with more ideas than would fit and wrote things which were cut for space. I had sabbatical coming up, giving me time to think about the bigger picture. The broader ideas and stray drafts formed the basis of this book.

Thanks to the Department of Philosophy at the University at Albany for making this book possible— for the semester of sabbatical in Fall 2021 which gave me a chance to focus on the project, but also for the willingness to tolerate my dilettantism.

The account I give here is built on the one I developed with my coauthors, and it would not have been possible without them. For feedback on drafts of the book, thanks to Christy Mag Uidhir, Ron McClamrock, Evan Malone, and Warren Magnus. Thanks also to Oz McClamrock, whose secondhand input about musical practice was useful at many stages of the project.

P.D. Magnus Albany, 2022

‘Now you're telling me you're not nostalgic.

Then give me another word for it.

You who are so good with words

And at keeping things vague.’

—Joan Baez, ‘Diamonds and Rust’ (1975)

‘Now you're telling me you're not nostalgic.

Then give me another word for it.

You who’s so good with words

And at keeping things vague.’

— Judas Priest, covering Joan Baez (1977)

‘Now you're telling me you're not nostalgic.

Well, then give me another word for it.

You were so good with words

And at keeping things vague.’

— Great White, as a tribute to Judas Priest (2008)

Introduction

Cover songs are a familiar feature of contemporary popular music. Musicians describe their own performances as covers, and audiences use the category to organize their listening and appreciation. However, philosophers have not had much to say about them.

A common philosophical approach is to consider historical positions— for example, asking what Plato or Kant said ona topic. That makes no headway here, because Plato and Kant had nothing to say about cover songs. How could they? A cover is a version of a song that was first recorded by someone else, so covers require the technology to record and play back music. If Plato or Kant wanted to listen to a song again, they had to find performers to play it again.

Nevertheless, philosophy provides a valuable toolbox for thinking about covers, and the philosophy of cover songs illustrates some general points about philosophical method. Why is it that people have been announcing the death of covers for as long as there have been covers, while musicians keep making them? To answer that, we need to introduce distinctions. There are different kinds of covers.

As muchas we need distinctions, however, we also need to recognize that honing our categories to diamond precision can be pedantic and doctrinaire. There are some distinctions which are not worth making.

A philosophical account of cover songs would be perverse if it were just an ethereal abstraction, so I discuss lots of different examples in this book. You may already be familiar with some of them. Others may be new to you. When it matters what a particular record sounds like, there is no reason to take my word for it. Most of the recordings that I discuss are readily available on the internet. You should listen and decide for yourself.

Please keep two things in mind about the examples.

First, you may find that you have different opinions about some of them than I do. Where the disagreement is incidental to the broader philosophical point, I ask you to substitute an example which you find more agreeable.

Second, despite all of the examples I address, there are many more that I have had to leave out. If you find yourself thinking about examples that I do not explicitly discuss, then I invite you to apply the distinctions and moves I make in the book. Part of the fun of thinking about cover songs is that there are interesting examples all over the place.

This book is divided into three parts, each containing two chapters. Although issues recur at different places, I have tried to make the chapters stand coherently on their own.

The first part is about how to think about covers. Chapter 1 reviews the history of covers and topples some possible definitions of ‘cover.’ Even though there is no clear definition, we can get by without one. I take cover songs to be the ones that are typically called that, and that is enough to get going. Chapter 2 introduces several distinctions which can help us understand covers better: First, between songs, performances, and tracks. Second, between mimic covers and rendition covers.

The second part is about appreciating covers. Chapter 3 uses the difference between mimic and rendition covers as the key to thinking about how we evaluate and appreciate them. Chapter 4 discusses covers which have an especially strong connection to the original— the originalis either alluded to by the cover or changes how we hear the cover. Evaluating and appreciating these covers turns out to be tricky.

The third part is about the metaphysics of covers and songs. Chapter 5 poses some puzzles about the metaphysics of cover versions. Although a cover is typically a version of the same song as the original, there are some interesting and striking counterexamples. Chapter 6 develops an analogy between songs and biological species. Like species, songs are historical individuals. This shows how to resolve the puzzle cases from the previous chapter and also helps in thinking about oddities like mashups, parodies, and instrumental covers.

Consider this recent Twitter thread:

Panda Lakshmi: Once I was singing Istanbul (not Constantinople) in my house and my mom starting singing along. Turns out the TMBG version is a cover!

Ellen Fuoto: So my 70 something year old brain is starting to wake up. You mean They Might Be Giants made a cover of the Four Lads hit from back when I was 6?

Uglysquirl: You two have blown my mind.. I’ma big fan of cover songs and TMBG and I never knew this was a cover.

gargoyle: I... With the.. But... I was today years old when I learned this. I’m going to need some alone time to deal with this crisis. ©

The song that they are talking about (‘Istanbul (Not Constantinople)’) was a hit in 1953 for the quartet the Four Lads, and it ‘found its way into our cultural lexicography as one of those songs that you knew you knew, but didn’t know where you knew it from’ (Treble 2018). The duo They Might Be Giants (TMBG) recorded a faster, livelier version for their breakthrough 1990 album Flood. The younger participants in this thread are surprised to learn this, because they just know it as a TMBG song. At the same time, the one older participant is surprised to learn about the TMBG version. Minds are blown. Even though the winking emoji in the last post makes clear that nobody’s life is deeply changed by this discovery, all the participants find it significant that the TMBG version is a cover.

Music audiences, which include you and me, use the concept of cover to understand certain songs, performances, and recordings. We take the difference between original and cover to be significant. But what is the difference? What does ‘cover’ mean?

The dictionary definition

A ‘cover’ is typically defined as a recording of a song that was first recorded by someone else. Something like this is given in many dictionaries and by some scholars. For example: Albin Zak provides a glossary entry defining a ‘cover version’ as ‘A recording of a song that has been recorded previously by another artist’ (2001: 222). Don Cusic writes, “The definition of a “cover” song is one that has been recorded before’ (2005: 174).

If it were that simple, this could be a short book. Inevitably, complications arise. Let's look at five of them.

Five problems

1. Consider the song 'Let It Be', written by John Lennon and Paul McCartney. Their band, the Beatles, had a hit with it when they released their version in 1970. However, the first released recording of the song— by a few months— was by Aretha Franklin. A website which generates its descriptions automatically labels the Beatles' version as a cover of Franklin's, and that is just what the usual definition would suggest. However, this seems absurd. If either version is a cover, then itis Franklin's. Lennon and McCartney were members of the Beatles who wrote the song with the intention of recording it, even though McCartney sent a demo to Franklin in hopes that she might record a version. It just happened that her version was released earlier.

One might think that the prior existence of the demo makes Franklin's version a cover, but many recordings— most in recent decades— exist as demos before there is a published version. To take just one example, consider Patsy Cline's 1961 hit ‘Crazy.’ The song was written by Willie Nelson, who was trying to get a singer to record and release it. He cut a demo record of ‘Crazy’ and played it in a bar in Nashville for Patsy Cline’s husband, who insisted he play it for Cline. She loved it and recorded her version the next week. Although Nelson had recorded a demo, almost nobody calls Cline's version a cover. It does not show up on internet lists of best cover songs or songs you didn't know were covers. Artists on YouTube typically list their versions of 'Crazy' as covers of Patsy Cline. So the existence of a demo does not seem to make Cline's version a cover.

However, consider ‘Girls Just Want to Have Fun’, a hit for Cyndi Lauper in 1983. It was written by Robert Hazard, and he recorded a demo version in 1979. Surprisingly, Lauper’s version appears on many of those internet lists. This may partly be confusion because Hazard’s demo was later published (to piggyback on the success of Lauper's version) but often these lists acknowledge that Hazard’s version was a demo. One comments, ‘Hazard’s recording never got past the demo stage, so I'll choose to consider Lauper’s version "technically a cover but sort of not really"' (Proximo 2017). When Lauper's album was selected for the National Recording Registry, a webpage at the Library of Congress included the comment, 'Lauper's take on Robert Hazard's "Girls Just Want to Have Fun" wasn't a mere cover, it was a transformation of the song into a joyous feminist anthem' (NRPB 2018). Something which is not a mere cover is more than a cover, rather than not being a cover at all.

Another example is the Crickets’ 1957 hit ‘Oh Boy’, which is often described as a cover of Sonny West's version (Londergan 2018). West, who cowrote the song, had recorded a demo of it under the title ‘All of My Love.’

Contrary to the simple definition, the existence of a demo version does not automatically make a version a cover— but maybe it does sometimes. Call this the problem of demo versions.

2. If we accept Cline's version of ‘Crazy’ as the original, then later recordings should count as covers. However, when Willie Nelson recorded it for his debut album the following year, it was not obviously a cover. Here common usage is unclear. Some people count Nelson's version as a cover (due to Cline's original) but others do not (due to Nelson having written the song). It is a vexed question.

It is also unclear how to think of cowritten songs. Consider two cases: First, the song ‘China Girl’ was cowritten by Iggy Pop and David Bowie, and Bowie played on and produced Pop's 1977 recording. Bowie recorded his own version in 1983 without Pop. Second, the song 'Layla' was cowritten by Eric Clapton and Jim Gordon. They recorded it with their band (Derek and the Dominos) in 1971. For MTV Unplugged in 1992, Clapton recorded an acoustic version which won the Grammy Award for Best Rock Song. The usual definition would call Bowie's 'China Girl' and Clapton's 'Layla' covers, and some people would agree (Leszczak 2014, Popdose 2011). Bob Leszczak, for example, describes the MTV Unplugged performance as Clapton having "covered his own song' (2014: 120). Other people are inclined to say that these are not covers.

So there are vexing questions about whether and how a person who wrote or cowrote a song can cover earlier recordings of it. Call this the problem of songwriters.

3. The typical definition only applies to new recordings. Yet a cover band is a musical group that just performs covers, and most cover bands perform live rather than recording. This shows that the word 'cover' is readily applied to live versions as well. There is an asymmetry, however, because something is not a cover if it is a recorded version of a song that has previously been performed live. Even though a cover may be a live version, the earlier original must be a recording. Call this the problem of live versions.

4. A cover of a song need not include any singing, and instrumental versions are regularly labelled as instrumental covers. Nevertheless, one would not call Miles Davis and Cannonball Adderly's 1958 version of 'Autumn Leaves' a cover. There were earlier released recordings by Yves Montand in 1946 (as 'Les feuilles mortes"), by Dizzy Gillespie and Johnny Richards in 1950 (as the instrumental ‘Lullaby of the Leaves’), and by others. The tune had become a jazz standard. When it is played today, one might compare the new performance to the famous Davis/ Adderly version, but one would not call the new performance a cover.

‘Autumn Leaves’ is not extraordinary in this regard. We treat jazz recordings differently than we treat rock recordings. This has prompted writers like Deena Weinstein (1998) and Gabriel Solis (2010) to argue that covers only exist in rock music. Weinstein writes, ‘Cover songs, in the fullest sense of the term, are peculiar to rock music, both for technological and ideological reasons' (1998: 138). However, this requires an expansive conception of what counts as rock. There are covers in pop music and contemporary country as well. Moreover, there are numerous points of interaction between jazz and rock (especially rock in this expansive sense). It is unclear how to draw the boundaries around the regions of musical or cultural space where covers are possible. Call this the problem of genre.

5. Contrast two cases: First, Kid Cudi's 2008 '50 Ways to Make a Record' follows the same melody and musical structure as Paul Simon's 1975 '50 Ways to Leave Your Lover’ but replaces Simon’s lyrics about love lost with ones about making music. Cudi's track is often described as a cover of Simon’s. Second, Weird Al Yankovic’s 1981 ‘Another One Rides the Bus’ follows the melody and musical structure of Queen’s ‘Another One Bites the Dust’ but replaces lyrics about being indomitable with ones about public transit. ‘Another One Rides the Bus’ is usually described as a parody of ‘Another One Bites the Dust’ and is not counted as a cover.

The percentage of words shared between the original and the parody does not seem to matter. There is parallel structure in the title and lyrics between Cudi and Simon but also between Yankovic and Queen. Perhaps the only thing which stops ‘Another One Rides the Bus’ from being a cover is that it is a parody, which in turn is because it is funny. And ‘50 Ways to Make a Record’ counts as a cover because it is not a parody, which in turn is because it is not funny. John P. Thomerson, who denies that parodies have to be humorous, seems to count all covers as parodies; he refers to typical cover band performances as ‘reverential parodies of classic rock and country hits’ (2017: 1). A definition of ‘cover’ should be able to make sense of this. Call this the problem of parodies.

Looking for the real definition

These problems are reasons to be unhappy with the usual definition, and we can use them as a toolbox to dismantle other definitions. For example, Andrew Kania defines a cover in this way: ‘A cover version is a track (successfully) intended to manifest the same song as some other track’ (2006: 412). This is vulnerable to all of the problems discussed above.

One might start tinkering with these definitions, adding clauses to resolve each of the problems. Yet that is not the only possible response.

An alternative approach supposes that the meaning of the term is determined by how it was introduced. The word ‘cover’ refers to a particular type of thing. So it has a real definition, the true nature of those things, regardless of what ordinary people or scholars might say when asked to define the term.

This approach was originally applied to proper names and to natural kind terms like ‘gold.’ The idea is that the word ‘gold’ was introduced to describe samples of gold, and it meant that kind of stuff. For centuries, people did not know what gold really was. They could not have given a true and informative definition. Only later did chemists develop atomic theory and physicists learn the structure of atoms, allowing us to characterize gold as a chemical element in terms of the number of protons in each of its atoms.

Nevertheless, that is what ‘gold’ meant all along (on this account). (My gloss of the view here is rather breezy. Key texts are by Saul Kripke (1972) and Hilary Putnam (1975), and decades of literature have followed.) Although the category of cover versions does not look like a natural kind, it has also been suggested that this approach to meaning applies to artifacts (Putman 1982). So maybe ‘cover song’ picks out that kind of recording or that kind of version. Unlike ‘gold’, which entered Old English from even older sources, the word ‘cover’ in the sense that interests us arose only in the late 1940s. So let’s turn away from puzzle cases and consider some history.

The history of covers

The term ‘cover’ first found widespread usage in the 1950s, corresponding to a shift in the record business.

Here is the simplified version: Before the 1950s, songs which everyone played became standards. This is natural when the paradigm case of music was live performance, both because performance is ephemeral and because it is done by whatever musicians someone has in front of them just at that time. Radio, initially dominated by live performance, did not immediately change this paradigm. After the 1950s, new versions of songs are often considered in relation to earlier recordings of that same song which are taken as canonical or original. The new versions are covers.

Early days

Initially, customers tended to seek out a particular song rather than a particular recording of that song by a particular artist. By covering a song, a record company could steal sales which would have gone to a competitor. As John Covach and Andrew Flory write, ‘When the original version appeared on a small independent label, a larger independent label (or a major label) could record a cover and distribute its records faster and more widely... They add that 'to some extent, this explains the greater success of these versions and why we call them "covers"' (2018: 87). Ray Padgett writes that covers in the 1950s were 'copycat recordings done quickly' and suggests two reasons these might have come to be called ‘covers’: First, a publisher might be ‘“covering its bets" by releasing its own recording of a popular song.’

Second, it was aimed to ""cover up" another version of the same song on a store’s shelves’ (2017: 4).

That is only part of the story. In a 1949 Billboard magazine article on small record labels, Bill Simon writes:

The original disking of Why Don't You Haul Off and Love Me?, cut for King [a small record label] by Wayne Raney, has hit 250,000, and versions are now available on all major labels. None of these, however, has approached Raney's mark. Another King disk, Blues Stay "Way From Me?, by the Delmore Brothers, is close to 125,000 in six weeks, and other companies have just begun to cover the tune. (1949: 18)

Here 'cover' has the sense of coverage. Just as a band might try to learn the popular songs that an audience member might request, a record company wanted to be able to have a version for sale. This reflects how songs work. A song can be performed by different artists. It is not matched one-to-one to the person who wrote it or the singer who made it famous.

From a commercial standpoint, there is no reason to make something original. It is easiest just to copy the interpretation and arrangement of a hit record, and the success of the hit suggests that it might be more commercially successful than trying something new. So there was a shift from making sure a label's library covered the repertoire to cutting records that just copied successful ones. In 1954, the chain store Woolworth's launched its own record label in the UK, Embassy Records. Their entire line was cheaply recorded knock-offs (Inglis 2005, Woolworths 2017).

Some in the industry commented on the contrast between earlier covers (new versions of a song recorded for coverage) and these new copy recordings. A 1955 Billboard article laments 'the duplication (rather than the covering) of successful disks' (1955a). Anarticle a few months later describes a New York radio station that ^will henceforth refuse to play "copy" records.' The article explains that this policy 'draws a clear distinction between "cover" records and "copy" records— defining the latter as those disks which copy— note for note— the arrangement and stylistic phrasing of the singer' (1955b). Nevertheless, the word 'cover' came to apply to both sorts of records— both recordings of the same song that used a different interpretation or arrangement and also those that copied the interpretation and arrangement of the original recording.

A further feature of music in this period was the centrality of rankings in trade magazines as a measure of commercial success. At least in the United

Figure 1: In the mid-1950s, Woolworth’s Embassy Records built a state-of-the-art studio in London for recording cheap covers.

Image courtesy of the Woolworths Museum.

States, this introduced complexities of race and class. The Billboard magazine rhythm and blues (R&B) chart had, prior to 1949, gone under a succession of other titles: ‘Harlem hit parade’, ‘race’, and ‘sepia.’ As the earlier names make clear, the chart was not meant to capture a particular style of music but instead a particular audience demographic— black people. The country and western chart, previously ‘hillbilly’ and ‘folk’, was also organized around a particular audience. As Covach and Flory note, ‘Rhythm and blues... charts followed music that was directed to black urban audiences, and country and western... charts kept track of music directed at low-income whites’ (2018: 85). This left the pop charts, although nominally just tracking popular music, focused predominantly on the white, middleclass market.

Covach and Flory put the point in terms of the music’s target audience, but a song could have success beyond just its target. The charts were constructed based on reports from radio stations and juke boxes (of what they were playing) and from record shops (of what they were selling). As a result, a song by a black artist could make it onto the pop charts if it had plays and sales in places to put it there. Similarly for country musicians. A song that made it onto multiple charts was called a crossover, and crossing over meant a distinct kind of commercial success.

Although some crossover hits were a single record appearing on multiple charts, others were the same song but recorded by different artists. Given the racial division of the charts, there are striking examples of white artists having pop hits with songs that had been R&B hits when recorded by black artists. The most famous example of this is probably Pat Boone’s 1956 pop version of Little Richard’s ‘Tutti Frutti.’ In that case, the cover by the white artist did not completely eclipse the original. Although Boone’s cover reached #12 on the Billboard pop chart, Little Richard’s reached #17 on the pop chart and #2 on the R&B chart. Regardless, this is just one instance of a broader pattern in which, as Denise Oliver Velez puts it, ‘Black music... was “borrowed,” “lifted,” “copied,” and made money for white artists, often garnering both commercial success and awards... while leaving the Black originators with far less, or nothing’ (2021). Singer-songwriter Don McLean describes it this way:

[I]f a black act had a hot record the white kids would find out and want to hear it on 'their' radio station. This would prompt the record company to bring a white act into the recording studio and cut an exact, but white, version of the song to give to the white radio stations to play and thus keep the black act where it belonged, on black radio. A ‘cover’ version of a song is a racist tool. (2004)

The word ‘cover’ suggests itself here perhaps as a contraction of ‘crossover.’ McLean leverages this as a definition, to argue that Madonna’s 1999 version of his 1971 song ‘American Pie’ should not be called a cover. Yet, common usage treats Madonna’s version as a cover. Although covers were sometimes used as racist tools, racism is not intrinsic to the concept of a cover as such. As Michael Coyle puts it, crossover covering of R&B hits by white artists ‘exploited racist inequality but did not arise because of it’ (2002: 144).

The word cover originally had a sense of coverage which was not in itself tied to race, and covers in that sense continued. Even when a cover eclipsed the original, it was not always about race. For example Sonny West cowrote and recorded ‘All My Love (Oh, Boy)’ (1957) and ‘Rave On’ (1958), but both were covered by Buddy Holly and the Crickets. Borrowed, lifted, and copied, but by white musicians from a white musician.

In the earlier, song-focussed market, songwriters and publishers would make money from sheet music as well as recordings. In the 1950s, the situation was changing. The only way for a country and western song to sell successfully as sheet music was if it crossed over to the pop charts (Gabler 1955). And soon enough sheet music would not be a central concern at all, as the primary product became the recording itself. Because of the changing marketplace, covers were a way for a song to get exposure to a broader audience. This was good for songwriters (who got a royalty from every sale, regardless of whose version was selling) but bad for performers (who profited only from sales of their records).

Coyle argues that this history fails to capture what covers really are. He writes ‘that no one in 1954 would have used the word “cover” to mean what we mean by it today.’ The sense of the word ‘cover’ that I’ve discussed so far in this section is what Coyle prefers to call hijacking a hit or just hijacking. Although hijacking was called covering in the 1950s, Coyle maintains that the word means something different now. He writes, ‘The notion of covering a song has changed radically in meaning because... the relation of writers to performers to audiences... has changed radically’ (2002: 136).