The free access to this article was made possible by support from readers like you. Please consider donating any amount to help defray the cost of our operation.
Uncovering the Flaws in A Philosophy of Cover Songs
Brandon Polite
Abstract
In this paper, I explore some flaws in P.D. Magnus’s book, A Philosophy of Cover Songs. First, I contest Magnus’s claim that in order to count as a cover, a version of a song must target a recording rather than a performance of that song. Second, I reveal a problem with his taxonomy of covers and defend “mimic covers” from his view that they aren’t worthy of aesthetic interest. Finally, I question the adequacy of his ontology of songs by considering a class of songs that his view cannot accommodate.
Key Words
cover versions, Hallelujah, music ontology, philosophy of music, songs, Todd Rundgren, versioning
1. Introduction
In 2022, P.D. Magnus released A Philosophy of Cover Songs, the first book-length treatment of and his first solo offering on the subject.[1] In the book, Magnus presents a helpful history of covers; refines his views on the nature and appreciation of covers from his earlier, coauthored work on the subject; and branches out into new territory by using his insights into covers to develop a novel account of the ontology of songs. Despite the book’s virtues—and there are many that I’ve adumbrated elsewhere[2]—in this paper I will explore what I take to be some of its flaws. In section 2, I contest Magnus’s claim that in order to count as a cover, a version of a song must target a recording rather than a performance of that song. This discussion enables me to offer a definition of covers, which Magnus declines to do in the book. In section 3, I reveal a flaw in Magnus’s taxonomy of covers and defend what he calls “mimic covers” from his view that they aren’t worthy of aesthetic interest. In section 4, I question the adequacy of the ontology of songs that falls out of his examination of cover versions, by considering a class of songs that his view cannot accommodate. While Magnus’s approach to covers is generally on the right track, and his book offers deep insights that surpass his earlier work, his account will be lacking until these critical shortcomings are addressed.
2. Dispensable recordings
In Chapter 1, Magnus surveys the history of covers to clarify the concept. The term ‘cover’ first specifically referred to recordings of songs by one artist or group that were previously recorded and released by another artist or group. Over time, the extension of the term expanded to include live performances of previously recorded songs. Magnus’s historical survey doesn’t yield a definition of ‘cover,’ but it does provide him with what he considers to be a necessary condition: “to be a cover something needs to be a musical version”—either a recording or live performance—“which covers an earlier recording.”[3]
It’s plausible, though, that the term’s extension has expanded further in a way that challenges Magnus’s claim that an earlier recording is necessary for a version to count as a cover of it. To show that this is plausible, consider the following hypothetical:
Hypothetical: Two bands are on tour together. Band 1 (the opener) plays a new song concert after concert. Band 2 (the headliner) learns the song and one night performs their own version of it during their set.
Let’s play with this hypothetical a bit to pump our intuitions about the practice of covering. Consider the following scenarios:
Scenario 1: Band 1 has released a recording of the song—it’s the lead single from their forthcoming album and is charting well.
Scenario 2: Band 1 has recorded the song but hasn’t released it yet—it’ll be the lead single from their forthcoming album. They’ve played it for Band 2 on the tour bus between gigs.
Scenario 3: Band 1 has recorded the song, hasn’t released it yet, and hasn’t played it for Band 2. Indeed, Band 2 doesn’t even know the recording exists.
Scenario 4: Band 1 hasn’t recorded the song yet—they’ve only ever played it live.
Scenario 1 is a paradigm case of a cover version. Band 2 has unquestionably covered Band 1’s song.
Scenario 2 also seems to be a case where Band 2 covers Band 1’s song—including by Magnus’s lights. There’s a definitive studio recording of the song; the fact that it hasn’t yet been released to the general public shouldn’t affect whether Band 2’s version of it counts as a cover or not. If Magnus were to deny this, then he’d need to amend his necessary condition so that cover versions require “previously released recordings.” However, it isn’t clear how such an amendment could be justified without begging the question.
Scenario 3 is a more problematic case. For Magnus, covers target earlier recordings. Even though such a recording exists in this scenario, Band 2 doesn’t intentionally target it, since they don’t know that it exists. Instead, they are targeting the song itself, specifically, the version of it performed live by Band 1 night after night. But perhaps the mere existence of the recording is enough to push Band 2’s version over into cover territory on Magnus’s view; again, he isn’t explicit that the covered recording must have been released to the public. If this is correct, then Band 2’s ignorance of the recording shouldn’t preclude their version of the song from counting as a cover.
This conclusion is plausible. Imagine that I’ve never heard any recorded version of Leonard Cohen’s “Hallelujah”—his or anyone else’s. However, I have heard dozens of versions of the song performed by others at the many open mics I regularly perform at. I have a good ear and memory, learn to play multiple versions the song, and then start performing my own rendition of it during some of my open mic sets. Everyone in attendance thinks that I’m covering Leonard Cohen, I think of myself as doing so, and it seems pretty straightforward that we’re all correct about this.
If this way of thinking about the open mic case is correct, then Band 2’s version should also count as a cover. In both cases there is a definitive studio recording; so, at the very least, they both satisfy Magnus’s necessary condition to count as covers.[4] If he were to deny that Band 2’s version counts as a cover, then he would need to think of some significant disanalogy between it and the open mic case—which does very much seem to be a case of covering—to do so. Alternatively, he could deny that both cases genuinely involve covering. If he were to do this, then he would need to provide an argument to show that neither version is a cover, perhaps adding a second necessary condition to the effect that the individual or group producing the cover must be familiar with and respond explicitly to the recording they’re covering. This would avoid the Gettier-style problem that Scenario 3 raises for his account of covers. Doing so, however, would make the epistemic state of the person or group producing a version of a song an essential part of whether it counts as a cover of that song, which he seems to want to avoid throughout the book. As such, this move is one that Magnus might not want to make, meaning he could well be committed to what I take to be the plausible view that Scenario 3 is a genuine case of covering.
Scenario 4 wouldn’t count as a cover by Magnus’s lights, since there is no earlier recording targeted in Band 2’s version; yet, Scenario 3 should lead us to be skeptical of this conclusion. If Band 2 really can cover Band 1’s song without ever having heard the recording of it (Scenario 3), as I think there’s good reason to believe they can, this means that recordings are at least sometimes dispensable to the practice of producing genuine covers. The version in Scenario 4, therefore, should count as a cover version just as much as the one in Scenario 3 does. If this is correct, then Magnus has misidentified earlier recordings as required for versions of songs to count as covers. As such, it seems at least plausible to suggest that Magnus hasn’t located a necessary condition for covers at all, but instead something that is merely coextensive with most covers.
If what I’ve just argued is correct, then this could point us toward a plausible definition of ‘cover’:
Cover =df a performed or recorded version of a song by one artist or band previously performed or recorded by another artist or band.[5]
An important qualification is needed to make this definition viable: namely, it needs to be restricted exclusively to the popular music tradition, which includes not just pop music but also rock, country, reggae, and so forth. It would be strange to claim that someone has “covered,” say, a Schubert lied, a jazz standard, a traditional folk song, or a national anthem merely by performing or recording a version of it. The practice of covering doesn’t exist within those traditions. However, someone absolutely could cover someone else’s version of these sorts of non-popular songs.[6] Just think of how many covers of Jimi Hendrix’s (1969) version of the “Star-Spangled Banner” there are. Or consider the Animals’ “House of the Rising Sun” (1964), which covers Bob Dylan’s recording of “House of the Risin’ Sun” (1962). In covering Dylan’s version of the song, the band brought a traditional folk song into the popular tradition. Dylan himself, however, arguably didn’t previously “cover” Dave Van Ronk’s (not-yet-recorded) version of the song, since both were situated within the folk tradition; instead, he stole Van Ronk’s arrangement in recording it for his (Dylan’s) debut album.[7]
The main virtues of the definition I’ve proposed—indeed, of having a definition at all, rather than a mere necessary condition, which is all Magnus provides us with—are that it fits well with the ordinary usage of ‘cover’ and enables us to answer, for any version of a song, whether it definitively count as a cover or not. Further, the definition is broad enough to encompass cases like Scenarios 3 and 4, which Magnus can’t account for because they’re cover versions that target performances rather than recordings.
3. Rethinking the value of mimic covers
In the seminal essay, “Judging Covers,” Magnus, together with coauthors Cristyn Magnus and Christy Mag Uidhir, offered a taxonomy of covers according to which there were four distinct categories: mimic, rendition, referential, and transformative.[8] Mimic covers aim to sound exactly like the target version, whereas renditions do not. Referential covers refer in some way or other to the target version; and transformative covers so radically depart from the target that they count as new, but obviously derivative, songs ontologically distinct from the originals.
In A Philosophy of Cover Songs, Magnus revises this taxonomy after recognizing its limitations. The key distinction is now between mimics and renditions, with referential and transformative covers now subsumed under the latter category. (In fact, Magnus drops the terminology of “transformative covers” altogether, simply talking about those renditions that don’t instance the target version in its place.)
Of course, there’s nothing to preclude covers that refer to their targets without also instancing the same song, which we might call transformative referential covers. Indeed, one of the greatest covers of all time, Aretha Franklin’s (1967) version of Otis Redding’s “Respect” (1965), is of just this sort. Keep this caveat in mind as I proceed. The basic idea is that Magnus has broadened his notion of rendition cover as a catch-all that includes all covers except for mimics.
On the specific decision to “demote” referential covers to a subcategory of rendition covers, Magnus writes:
In earlier work, my collaborators and I … considered [referential covers] as a separate category from rendition covers.[9] I now think of rendition covers more broadly, so that referential covers are a particularly interesting kind of rendition cover.[10]
But there’s a glaring problem with this move: there are clear examples of mimic covers that refer to the recordings they target. The most obvious examples of what we might call referential mimic covers are found on Todd Rundgren’s 1976 album Faithful. On the album’s first side, Rundgren includes stunningly faithful—hence, the album’s title—mimics of canonical rock tracks by the Beatles, Beach Boys, Yardbirds, Jimi Hendrix, and Bob Dylan. These mimics are certainly “superfluous duplicate[s]” of the original recordings, which is what Magnus believes to be the highest achievement possible for mimic covers,[11] in the sense that they don’t need to exist. However, by means of their superfluousness they are able to do critical artistic work—namely, they point out the superfluousness of earlier rock music.
Media culture theorist Steven Bailey argues that Rundgren’s faithful mimics specifically succeed by ironically embodying the previous generation of rock music’s highest aspirations.[12] “What is particularly interesting about them,” Bailey writes,
is that the material includes music that was regarded as technologically adventurous (e.g., “Good Vibrations,” “Strawberry Fields Forever”), as well as “Most Likely You’ll Go Your Way and I’ll Go Mine,” which is valued for its spontaneity and lack of polish. While one might view the replication of “Good Vibrations” as a tribute to Brian Wilson, an earlier studio wizard, the Dylan and Hendrix covers reveal deeper implications concerning the use of technology and the reproducibility of genius.[13]
He continues:
[T]he irony here is that his utilization of studio technology and personal virtuosity—normally tools for maximizing self-expression—produce a work free of personal commitment. By turning technology upon itself, Rundgren takes the Benjaminian aura-smashing potential of mechanical reproduction one step further. Not only is art reproducible, it no longer requires an original. “Good Vibrations” by Rundgren becomes indistinguishable from “Good Vibrations” by the Beach Boys, and thus the technology that allowed Brian Wilson to create his pop masterpiece renders the achievement meaningless.[14]
Since I see no reason to disagree with Bailey’s analysis of Rundgren’s covers as referential mimics, we need to revise Magnus’s taxonomy so that referential covers can exist on either side of the mimic-rendition binary.[15]
The preceding observations indicate that mimic covers can be more interesting and worthwhile than Magnus is willing to give them credit for. At the very least, mimics can be more than mere technical exercises whose value resides solely in how closely they resemble the recordings they target, which is what Magnus limits them to; and this would be true even if the close resemblance is an indispensable means toward achieving those higher artistic and aesthetic ends.
About mimic covers, Magnus writes:
[T]here is a limit to how great a mimic cover can be. Even if you admire the artisan who makes a replica, you admire merely their craft. The perfect replica would be interchangeable with the original—as good as the original, but no better. Contrawise, there is no limit to how bad a mimic can be.[16]
But Magnus doesn’t believe rendition covers are good only insofar as they depart from their targets; they can also be good, like mimics, because of those aspects of their targets they preserve:
For a mimic, which aims at matching the original in all respects, the differences are all failures of skill. For a rendition, however, the differences reflect interpretive choices—that means that the similarities, too, reflect interpretive choices.[17]
One mode of appreciating a rendition cover, then, is by considering it in relation to its target: what it preserves and how it diverges. Magnus calls this the “etiological mode” of appreciating renditions.[18] Choosing to preserve the target version in all respects, of course, as Rundgren did with his Faithful covers, is an interpretative choice just as much as choosing to preserve only some of them is. As this is true for all mimics, it would seem that the etiological mode is open to mimics just as much as it is to rendition covers—that is, we can enjoy them for more than their mere craftsmanship and think about them in relation to their targets: what they might be saying about them, and so on.
But there is another way to appreciate renditions that Magnus does not to believe applies to mimics:
[A] rendition cover can be a great version of the song apart from any consideration of the original. It can be beautiful, powerful, or moving—features that it just has on its own without considering its relation to the original.[19]
Magnus refers to this as the “aesthetic mode” of appreciating rendition covers.[20] But this mode is also open to mimics. Hardcore Swifties, for example, obsessively listen to Taylor Swift’s re-recordings of her previously released tracks—her “Taylor’s Versions” of them—in the etiological mode: poring over their similarities and differences. But these mimics can definitely be appreciated on their own terms; whether they’re covers or not, while debatable,[21] is irrelevant to my point. New Swift fans and casual listeners, and even long-time Swifties, although perhaps with some difficulty, can enjoy them without thinking about Swift’s re-recordings in relation to the original versions—which is how Swift wants us to treat them, that is, as replacements for the originals. And any listener who confuses a Taylor’s Version for the original will simply enjoy it as if they were enjoying the original. This is true of any especially good mimic—which, it should be noted, many of Swift’s are not by design. Listeners can enjoy the trippy bluesiness of Rundgren’s faithful mimic of “If Six Was Nine,” for example, without appeal to Hendrix’s (1967) original version of the song. Its trippy bluesiness, of course, certainly derives from Hendrix’s version; however, this and its other aesthetic qualities are no less available to listeners because of this.
The preceding discussion reveals that mimic covers offer us more aesthetic rewards than Magnus believes they do, perhaps as many as renditions do, and thus that they aren’t as fundamentally different as he believes. While it’s important to note their differences for the purpose of evaluation, since the bar for bad mimics is much lower than it is for bad renditions, because any departure from the target lessens a mimic’s quality, it’s equally important not to overstate their differences, as I’ve argued that Magnus does. Ultimately, recognizing the similarities between them allows us to appreciate the artistic value in both.
4. Songs without lineages?
A highlight of Magnus’s book is the ontology of songs he provides in its concluding chapter, according to which songs are best understood by way of analogy to biological species. Magnus specifically argues that songs, like species, are historical individuals, meaning they’re determined by ancestry relations rather than their intrinsic properties. Belonging to a particular species requires that an organism stand in the right causal and historical relationship to other members of that species—that is, they are part the same hereditary lineage. Conceiving of species in this way accommodates the biological facts that species come into existence, evolve, and go extinct at definite places and times.
Magnus thinks of songs along the same lines. He writes:
My idea is to see a song as a lineage of versions connected by relations of inspiration and copying. Just as the members of a species are organisms of common descent, the versions of a song are performances/recordings with historical continuity and causal dependency.[22]
In particular, Magnus conceives of songs as akin to certain plant species, like wild strawberries, that reproduce in two ways: by dropping seeds and by sprouting runners.[23] A plant produced from a seed isn’t genetically identical to either of its parents, whereas one produced from a runner is a clone of its parent plant. To better understand how songs reproduce analogously, consider the following example. “Comme d’habitude” was a mega-smash French pop hit from 1967, composed by Jacques Revaux, with lyrics by Claude François and Gilles Thibaut. François’s original recording—the initiator of this particular lineage—has been played millions of times since its release, and he and others have performed relatively straight renditions of it live many times. The copies of the studio master, whether on vinyl, cassette tape, CD, mp3, and so on, are the “runners” the song spawned, as they are its sonic clones. The straight and other renditions of the song are the “seeds” it dropped, as they all vary slightly or not-so-slightly from the original version and each other, that is, there’s some sonic variation between them.
In 1968, struggling songwriter David Jones’s (aka, David Bowie’s) publisher asked him to write English lyrics for the song. This resulted in the song “Even a Fool Learns to Love.” However, this version of the song went nowhere, as Bowie’s publisher rejected it, although both his lyrics and the demo he recorded for it were eventually released.[24] A year later, Paul Anka wrote his own English lyrics for the song, which became arguably the most important release of Frank Sinatra’s career, “My Way” (1969), a song that Sinatra performed live hundreds of times and has been covered both live and on record by countless artists. Whether Bowie’s and Anka’s versions count as members of the “Comme d’habitude” species or new species that it spawned, that is, ontologically distinct songs, is debatable. I’m inclined to think the latter, but perhaps there are good reasons to think they really are mere instances of “Comme d’habitude.” One might point to the fact that Revaux, François, and Thibaut received songwriting credit alongside Bowie and Anka, respectively, to start making this case; however, I won’t pursue that line of thinking any further here.
Two years later, in 1971, Bowie released his breakthrough album, Hunky Dory, which contains his masterpiece, “Life on Mars?”. On album’s back cover, Bowie indicates that the track was “Inspired by Frankie.” Indeed, the track—which has itself been listened to and covered countless times —is a parody of Sinatra’s “My Way” borne out of the rejection of “Even a Fool Learns to Love.”[25] While “Life on Mars?” may ultimately derive from “Comme d’habitude,” there’s a much weaker case for the claim that it’s an instance of that song than for either of its predecessors. Indeed, I think it’s straightforwardly an ontologically distinct song.
In any event, this is quite a complex musico-historical family tree—with some incestuous cross-pollination thrown in for good measure! The new songs—if they really are new songs—would have emerged from “Comme d’habitude,” Magnus argues, when they spawned separate lineages of their own, that is, when they inspired other artists to produce their own versions intentionally targeting them rather than “Comme d’habitude.”[26]
Magnus’s account of songs as historical individuals is incredibly interesting and highly plausible. It’s certainly more plausible than, say, its Platonist rivals, which contend that “Comme d’habitutde” has always existed and will always exist. However, I’m concerned that it cannot accommodate songs with less robust pedigrees than the ones just described—indeed, songs with seemingly no pedigrees at all. These would be songs that don’t belong to any clear lineage because they exist solely in a single performance. Over my decades of guitar playing, I’ve made up dozens of songs (words and all) almost entirely on the spot, with little-to-no working out—mostly brief little ditties, but some longer and more involved—that I never played again because, for one reason or other, I didn’t bother to record or remember them. (The most prevalent reason, I should add, is that they weren’t any good.) Given all of the many musicians who have been, are, and will be, it’s safe to estimate that hundreds of millions of these extemporaneous song-like entities have existed and will exist, however brief their lifetimes may be. Whatever their nature, they’re certainly inspired by earlier songs; however, we wouldn’t be able to trace the lineages for many (perhaps most) of them as clearly as we could for “Life on Mars?” and, moreover, they initiate no lineages of their own. These musical entities very much appear to be songs. Yet, because they don’t belong to any clearly delineated lineages and fail to spawn lineages of their own (no recorded copies, no subsequent versions), it isn’t clear that they would count as songs by Magnus’s lights. At the very least, Magnus can’t definitively tell us that they really are songs, which is certainly a flaw of his account.
For biological species, nonreproducing mutations that don’t initiate separate lineages of their own are genealogical dead-ends. They don’t belong to a species all their own, but instead belong as “freak” members of the species they spawned from. But it would seem wrong, in the musical case, to say these “freak” songs are instances of the songs that inspired them. If I play an extemporaneous song very clearly inspired by three earlier songs, it isn’t clearly an instance of any or all of them—it’s its own thing. But what kind of thing is it? I would like to say that it’s a song, but Magnus’s account lacks the resources to affirm that it is. Perhaps Magnus could say that this just an important disanalogy between species and songs, and there may be no fact of the matter about whether or not these musical freaks count as songs. However, an ontology that can’t tell us which entities count as songs and which do not wouldn’t be a very compelling ontology of songs. As such, Magnus needs to do more work to make his ontology of songs fully acceptable to those of us not of a Platonist persuasion.
5. Coda
In this paper I have uncovered some flaws in Magnus’s account of cover versions. Along the way, I proposed what I take to be a prima facie plausible definition of covers, suggested an important revision to Magnus’s taxonomy, defended the aesthetic merits of mimic covers, and inquired into an underexplored area in the ontology of music: the nature of extemporaneous songs. These contributions not only highlight the limitations of Magnus’s account but also pave the way for a more nuanced understanding of the nature of cover versions and how we appreciate them. None of the criticisms I have made are fatal to Magnus’s overall approach to covers, nor were they intended to be. I offer them in the spirit of friendly amendments or else to suggest areas in need of some careful rethinking and further exploration. Ultimately, my aims here have been to enrich the ongoing discourse and to encourage more rigorous analysis and appreciation of the complexities within the philosophy of popular music.
Brandon Polite
bpolite@knox.edu
Brandon Polite (he/him) is professor of philosophy at Knox College in Galesburg, Illinois. His research interests focus on aesthetics, especially the philosophy of music. He is editor of the book, Taylor Swift and the Philosophy of Re-recording: The Art of Taylor’s Versions (Bloomsbury, 2025). He is also host of the YouTube series, Polite Conversations: Philosophers Discussing Art (https://www.youtube.com/c/PhilosophersDiscussingArt).
Published February 9, 2025.
Cite this article: Brandon Polite, “Uncovering the Flaws in A Philosophy of Cover Songs,” Contemporary Aesthetics, Volume 23 (2025), accessed date.
Endnotes
[1] P.D. Magnus, A Philosophy of Cover Songs (Open Book Publishers, 2022). https://doi.org/10.11647/OBP.0293.
[2] Brandon Polite, “Book Review: A Philosophy of Cover Songs,” Journal of Aesthetics and Art Criticism 81, no. 1 (2023), 109–112, https://doi.org/10.1093/jaac/kpac072.
[3] Magnus, A Philosophy of Cover Songs, 28.
[4] The case of “Hallelujah” is, in fact, overdetermined, as there are at least three definitive studio recordings of the song: Cohen’s (1984) original, John Cale’s arguably canonical (1991) version, and Jeff Buckley’s (1994) cover of Cale’s version.
[5] Perhaps I should instead make this just a necessary condition and not also a sufficient one, thereby following Magnus in not wanting to give a definition of covers. For the purposes of this article, though, I’m going to boldly suggest that the concept of a cover is definable simply because I think the definition I provide is highly plausible.
[6] I say “non-popular” rather than “unpopular” to indicate that I’m referring to songs that weren’t originally produced within the popular tradition rather than to songs that very few people like.
[7] https://americansongwriter.com/the-time-bob-dylan-stole-dave-van-ronks-arrangement-for-house-of-the-rising-sun/. Alternatively, we might think that 1960s folk music, due to its widespread mainstream success, is distinct from traditional folk music and really belongs within the popular tradition. If so, then it would be right to say that Dylan actually did cover Van Ronk’s version of “House of the Rising Sun” and stole credit for Van Ronk’s arrangement in doing so, if what I said about Scenario 4 is correct. To determine which option is the correct one in this particular case, we’d need a clear account of music genres, which goes well beyond the scope of this paper.
[8] Cristyn Magnus, P.D. Magnus, and Christy Mag Uidhir, “Judging Covers,” Journal of Aesthetics and Art Criticism 71, no. 4 (2013), 361-370.
[9] Here citing Magnus, Magnus, Mag Uidhir, “Judging Covers.”
[10] Magnus, A Philosophy of Cover Songs, 74.
[11] Magnus, A Philosophy of Cover Songs, 127.
[12] Steven Bailey, “Faithful or Foolish: The Emergence of the ‘Ironic Cover Album’ and Rock Culture,” Popular Music and Society 26, no. 2 (2003), 141-159.
[13] Bailey, “Faithful or Foolish,” 153.
[14] Bailey, “Faithful or Foolish,” 153-154.
[15] Note, though, that the referential-transformative hybrids discussed above, with Franklin’s “Respect” as a paradigmatic example, only fall under the rendition side of the binary. I can’t think of a single example of a mimic cover that does not instance the same song as its target. Indeed, I believe that such a cover would be impossible in principle.
[16] Magnus, A Philosophy of Cover Songs, 58.
[17] Magnus, A Philosophy of Cover Songs, 60.
[18] Magnus, A Philosophy of Cover Songs, 65.
[19] Magnus, A Philosophy of Cover Songs, 60.
[20] Magnus, A Philosophy of Cover Songs, 65.
[21] To better understand why, see Cristyn Magnus, P. D. Magnus, Christy Mag Uidhir, and Ron McClamrock, “Tell Me Why This Isn’t a Cover,” in Taylor Swift and the Philosophy of Re-Recording: The Art of Taylor’s Versions, ed. Brandon Polite (Bloomsbury, 2025).
[22] Magnus, A Philosophy of Cover Songs, 115.
[23] Magnus, A Philosophy of Cover Songs, 113.
[24] https://www.davidbowieworld.nl/mijn-bootlegs-2-2/outtakes-promo/attachment/david-bowie-even-a-fool-learns-to-love/.
[25] https://ig.ft.com/life-of-a-song/life-on-mars.html.
[26] Magnus, A Philosophy of Cover Songs, 115-116.