Stairway to Infringement

Intro: “The Hook”

This summer, an American jury found that “Stairway to Heaven” [hereafter Stairway] rockers Led Zeppelin did not infringe the song “Taurus” [hereafter Taurus], performed by the band Spirit. The plaintiff, Randy Wolfe—or rather, a trustee for the trust which owns the late Wolfe’s copyright—was Spirit’s songwriter, guitarist and vocalist, and the copyright owner of the musical work. Wolfe alleged that Stairway songwriters Robert Plant and Jimmy Page attended a Spirit concert, heard his 1967 composition Taurus, then stole the descending chord progression, arpeggio sequence, and harpsichord melody, using it as their 1971 classic ballad’s introduction.[1]

The case—conducted and concluded by jury like the questionableBlurred Lineslawsuitfaced some controversy. And, in the wake of the Blurred Lines ruling, it was difficult to forecast this case’s outcome. On June 23, however, the jury needed less than a day’s deliberation to deliver a reasonable judgment: scant evidence of copying and little substantial similarity between the songs meant no infringement.

It was reported that Wolfe’s attorney stated he “lost on a ‘technicality’” and plans to appeal. The technicality on which he lost, according to the trustee, was the substantial similarity comparison. It is a woeful misunderstanding of the significance of substantiality in non-literal copyright infringement to consider it a “mere technicality”.

Verse: Canadian copyright infringement

Applying US or Canadian law should lead to the same result in this case. In Canada, section 3 of the Copyright Act [the Act] grants a copyright owner the sole right to reproduce the work or a substantial part of that work. And, section 27(1) of the Act calls it an infringement when anyone, without the owner’s consent, does what the owner has the right to do with that work. When exact copies are not involved, infringement claims typically refer to substantial part reproduction. So, substantial similarity comparisons arise when two works which are dissimilar—like Stairway and Taurus—and courts must compare them to see if the impugned work reproduced a substantial part of the plaintiff’s work.

Therefore, to succeed in this type of infringement claim, the plaintiff (Wolfe’s estate) must prove two elements on a balance of probabilities:

  1. the work was copied; and
  2. sufficient objective similarity (the copied portion represents a substantial part of the plaintiff’s work).[2]

Whether a substantial part was taken is crucial to a finding of infringement. If it was not, the Act is not engaged,[3] and whether the defendant (Led Zeppelin) copied the plaintiff is irrelevant.

Copying part of a work can and does result in legitimate new copyrightable works. Also, because ideas are not copyrightable, using tools available to all authors cannot alone make infringement: creators can copy a colour palette from a painting; a series of rhyming words from a poem; a chord progression from a song.[4] Taking a particle of a work does not infringe.[5] So, while some cases unfortunately settle on this point, chord progression borrowing in pop music should not, typically, lead to infringement findings. If they did in other musical genres, many blues and jazz compositions would cease to be protectable, as they are genres built on borrowed chord progressions. Rock music, too, is not immune to borrowed chord progressions, but unless the borrowed progression represents a substantial expression of the first work, it would not be infringement. Sam Smith might have benefited from this point before settling with Tom Petty on this count, since the borrowed element in “I Won’t Back Down” is merely part of Petty’s verse, not the chorus (as it is used in Smith’s “Stay With Me”).

So, even if Led Zeppelin were proved to have directly copied the work—a claim which was “strenuously disputed”[6] though likely conceded when Page and Plant admitted having access to Spirit’s music—Stairway’s use of similar (not identical) arpeggios and one phrase’s chord progression would still have been insufficiently similar to any protected expression in Taurus. This point bears emphasizing because lay listeners are sometimes convinced of infringement upon hearing similar musical phrases, even though most innocent and deliberate rephrasing of musical ideas is allowed.

A problem that arises in courts is that the line between idea and expression is rarely clear and, particularly for non-literal works, rather hard to quantify. So, courts must embark on a determination of fact to compare and establish if a substantial part of the plaintiff’s work was reproduced in the impugned work. A deviation in the Canadian approach to a substantial similarity comparison presents a unique argument in this case which lacks in discussion of the US decision (though would still lead to non-infringement finding).


Interlude: Canada’s holistic test presents a stronger case for non-infringement

The Supreme Court of Canada (SCC) recently established a new test for comparing two works in a substantial similarity analysis.[7] The two works are compared cumulatively to see if the defendant unduly interfered with the plaintiff’s rights.[8] The holistic analysis is new, but it crystallized prior Canadian jurisprudence and does not represent a major departure from prior Canadian copyright cases, save for in one key area: the viewpoint from which the two works should be compared.[9]

Robinson c. Films Cinar inc [Cinar] shifted the comparative perspective from the traditional “layperson in the intended audience” to that of a person with a more knowledgeable viewpoint,[10] depending on the nature of the work.[11] The Cinar decision stated: [1] it is useful to compare the two works from the layperson’s view; but [2] the two works should be viewed from the person who can assess all the patent and latent aspects of a work because their senses and knowledge fully allow it.

This perspective deviates from the American one. There, after an extrinsic stage which asks first which elements are objectively similar,[12] the intrinsic stage just asks whether a typical lay observer would recognize the accused work as having been copied from the original.[13] Some courts do refer to the “intended audience” of the work,[14] a consideration infused into the Canadian comparison, but not all.


Chorus: Comparing the two works from the correct perspective should find no substantial similarity

A key argument to temper any similarity in this case found from the lay perspective is found in Cinar’s reasons beginning at paragraph [52]: “Stairway”, the impugned work, is by all accounts the more renowned work. Since this suit is based on 45-year-old compositions, any layperson in Spirit’s intended audience would recognize that arpeggio sequence as the introduction to Zeppelin’s rock classic, and more importantly, they would imbue it with significance (as the plaintiff’s expert did, though incorrectly to Wolfe’s composition).[15] It should be conceded that the impugned passage’s recognizability is not a result of its own inherent substantiality, nor of Spirit’s particular expression of it.

So, the question for the trier of fact, who must take the viewpoint of a person who can assess all patent and latent aspects, is whether Taurus’ A-minor descending sequence as described (the only protectable similar element found by the court in the extrinsic test)[16] would have been recognizable without thirty years of radio airplay and popular culture references. The trier must pretend they live in a world where Stairway never existed, and ask then if the borrowed element could comprise a substantial part of Taurus.


Bridge: the works being compared are musical compositions, not sound recordings

Note, however, that what the court compared was not two sound recordings (since the plaintiff does not own that copyright).[17] It had to compare Stairway to the musical composition, Taurus, as written (in musical notation). And while it is permissible for the court to consider Stairway’s sound recording (in Cinar, the court compared a realized animated television show to incomplete, unpublished materials), the comparison remained limited to Taurus’ transcription, and not its recording, which Wolfe has no right to protect.[18]

Yet, there is some Canadian case law which would militate against Zeppelin, if a similar case were to occur here. A 1991 decision found infringement on the basis that the first eight measures of the two songs had a “striking resemblance” emphasizing the qualitative importance of the first measures of the refrain “which are the ‘hook’ that the ear retains for the purpose of identifying a piece”.[19] This analysis favours Spirit, but only if one presumes that the opening progression of Taurus is the “hook” of that song. Again, hindsight (or hind-hearing) saturates the allegedly copied sequence with familiarity, but without Stairway’s fame, would it be so for Taurus?

Four arpeggios alone, non-identically copied, is a weak basis for an infringement claim. Inclusion of a matching melody line, similar rhythmic accompaniment, or other musical elements and Stairway could create a threshold case of substantial similarity. Wolfe’s expert attested that Taurus’ harpsichord melody resembled Plant’s vocal line in Stairway, albeit the court rejected its inclusion without giving reasons for doing so.[20]

One might surmise that comparing Stairway’s prominent vocal melody—a complex performance of lyrical content infused with a performer’s distinct vocal timbre—to a written lifeless harpsichord line (albeit perhaps with some direction for dynamics and articulation) was beyond the scope of the plaintiff’s protection.[21] Because the protected work is a musical composition and not a sound recording, a trier of fact should and did have difficulty establishing a lone dissimilar arpeggio sequence as comprising a substantial part of a composition. That difficulty is amplified when the impugned work is not to be compared to performance and production elements in the protected work.



Perhaps a Canadian court would find another legal peg in Cinar to rest a non-infringement finding on: in paragraph [40] the unanimous court wrote that “If the differences are so great that the work, viewed as a whole, is not an imitation but rather a new and original work, then there is no infringement.” Comparing Taurus—an airy two-and-a-half-minute baroque-style instrumental piece—to Stairway—an eight-minute hard rock ballad recording featuring several iconic guitar solos, lyrics, vocal melodies, drum riffs, and other protectable expression galore—should demonstrate that the latter is a new and original work. It should especially do so because, in 2016, it would be unfair to reward Taurus with a substantial part finding as a result of Stairway’s fame, and it would build a mechanism for less successful litigants to sue more successful artists for original works which legitimately employ common chords, melodies and rhythms: tools otherwise free for any composer to make music, but slowly locked away in spite of the Act’s purpose to encourage the creation and dissemination of new works.


Jordan Fine is Senior Editor of the IPilogue and Intellectual Property Journal and a JD candidate at Osgoode Hall Law School



[1] Skidmore v. Led Zeppelin, No. CV153462RGKAGRX, 2016 WL 1442461, (C.D. Cal. Apr. 8, 2016) [Skidmore].

[2] Copyright Act, RSC 1985, c C-42, ss 3(1), 14.1, 28.2 [C Act]; John McKeown, Fox on Canadian Law of Copyright and Industrial Designs, 4th ed, vol 1 (Toronto: Carswell, 2012) (online version updated in 2014) at 21:2 [Fox on Copyright]; Francis Day & Hunter v. Bron, [1963] 2 All ER 16 at 27 (CA) [Francis Day].

[3] C Act, Ibid, s 3(1).

[4] As David Vaver wrote in Copyright Law (Toronto: Irwin Law, 2000): “[a]nyone was (and is) free to paint funny-looking people holding guitars: what they cannot do is imitate Picasso’s expression of these subjects.”

[5] Robinson c. Films Cinar inc, 2013 SCC 73, 2013 CarswellQue 12345 [Cinar] at para [143].

[6] Skidmore v. Led Zeppelin, No. 2:15-cv-03462 RGK (AGRx), 2016 WL 1470076 (C.D.Cal. Apr. 4, 2016) (Trial Motion, Memorandum and Affidavit).

[7]Cinar, supra note 5 at [25].

[8] Ibid, at paras [35] & [143].

[9] Ibid, at paras [50]-[51].

[10] Ibid, at [51].

[11] Ibid at [51]-[55].

[12] Shaw v. Lindheim, 919 F.2d 1353, 1356–57 (9th Cir. 1990).

[13] Arnstein v. Porter, 154 F.2d 464, 467 (2d Cir. 1946).

[14] Dawson v. Hinshaw Music Inc., 905 F.2d 731, 733–36 (4th Cir. 1990).

[15] Skidmore, supra note 1 at *4.

[16] Ibid at *16.

[17] Skidmore v. Led Zeppelin, No. 2:15-CV-03462-RGK-AGR, 2016 WL 2771246 (C.D.Cal.) (Trial Motion, Memorandum and Affidavit); Ibid.

[18] Ibid.

[19] Grignon v. Roussel (1991), 38 C.P.R. (3d) 4.

[20] Skidmore, supra note 1 at *16.

[21] Ibid.

  1. Great post, examining the key issues in some depth and comparing the US and Canadian models well. I’d only add that the doctrine of “striking similarity” (where access can be inferred if the resemblance goes well beyond “substantial” similarity) has occasionally gotten some traction in the US, whereas in Canada demonstrating infringement without possibility of access has not, to the best of my knowledge, ever been successful.

    As a music theorist, what I find more interesting in this particular comparison (as one expert for the plaintiff also pointed out), is that a descending chromatic line (itself not musically particular) from the tonic to what is typically the dominant gets interrupted/abandoned on the submediant in both songs, leaving the expected progression incomplete. While probably not rising to the level of evidence of copying, it is a rather salient similarity between the two songs, and I can see why, from a musical basis, the judge would have found that there was a triable issue at hand.

  2. Thanks for the comment, Guillaume,

    If I’m not mistaken, “striking similarity” has been tossed around as a concept in a few cases…the one I’m thinking of is an Ontario small claims court in the 90’s called Drynan v. Rostad, but I think there may be a couple others. Still, I believe you’re quite right that Canadian courts typically demand a higher threshold for the copying aspect.

    I would agree that’s the correct approach. Particularly today, as access is becoming more easily inferred due to the Internet’s presence. I thought it was fascinating how Page, in his declaration (and referring to access to Spirit’s music), spoke of his physical record collection. He admitted that he had a copy of Spirit’s album but had never listened to it. Today, my Google play account grants access to terabytes of streaming music in addition to my own digitally downloaded or ripped-from-CD collection. Though google tracks my listening history, I’m unclear how I might disprove “access” to nearly any non-obscure song. Should the test change since access is as ubiquitous as spending time online?

    On the compositional note, I found the uniqueness of that musical phrase (and the unquestionable similarity) interesting too. It is certainly no I-vi-ii-V progression. What I’m especially curious about is in what circumstance that phrase (or any unique chord progression) could constitute a substantial part of a plaintiff’s song.

  3. After I read this thorough and interesting article, several things struck a chord with me:
    1. There is a 3-year statute of limitation to bring a claim of copyright infringement that starts running when the plaintiff knew or should have known of the alleged infringement. However, if the clock restarts at each new act of alleged infringement, the limitation period is meaningless in the music and film industry. With today’s seemingly infinite number of re-releases, online downloads and streaming sites the clock is perpetually restarting.
    2. Part of the judgment hinged on determining whether Led Zeppelin heard Taurus before writing Stairway to Heaven. This kind of analysis is already plagued with deductions and guesswork before adding on the fact that these works are 45-years old. This case highlights the absurdity of proving access in copyright infringement cases. In my opinion, the determination of whether a substantial part of the plaintiff’s work was reproduced is the logical first step (after proving title to sue) in any copyright infringement case. If it is answered in the affirmative, the trier of fact should determine the likelihood of independent creation in view of the substantial piece of the plaintiff’s work which has been reproduced.
    3. Justice Binnie, in a widely quoted passage from Théberge, states that the Copyright Act strikes “a balance between promoting the public interest in the encouragement and dissemination of works of the arts and intellect and obtaining a just reward for the creator”. Although the case at hand is American, this passage may offer further justification for the decision. Randy Wolfe, the writer of Taurus, tragically drowned in 1997 so it is his estate which filed suit. I wonder if the “just reward for the creator” side of the balance is significantly weakened when the actual creator is not the one claiming infringement. This contrasts with in Cinar v Robinson, where Claude Robinson claimed infringement of his own work, making him more of a sympathetic figure.
    4. In Cinar, there is a comparison between an aired TV show and literary submissions for a TV show. In Stairway, there is a comparison between a sound recording and a piece of musical notation. When compared to literary work, musical notation may comprise significantly lower amounts of “intelligible” elements, such as atmosphere, motifs and themes. The court in Cinar specifically notes the need of the expert to distill intelligible similarities between works. Will expert evidence going to the determination of “substantial part” be admitted in Canada where intelligible elements are sparse and perceptible elements are dominant?

  4. I have many opinions on these thoughtful comments, Justin, particularly my agreement with you on the weakness of justification for posthumous copyright protection (and on that I’ll just point to Thomas Macauley’s famous opposition to posthumous term extension “It is good that authors should be remunerated; and the least exceptionable way of remunerating them is by a monopoly. Yet monopoly is an evil. For the sake of the good we must submit to the evil; but the evil ought not to last a day longer than is necessary for the purpose of securing the good.”)

    Better responses demand more than I have time to commit to blog comments. But I will say (on the posthumous weakening of the “just reward” side of Théberge’s balance) that I can’t fathom why we continue to focus and define our copyright terms on dates of death. Heirs don’t create or innovate; so, according to copyright law’s purpose at least, they are not entitled to the fruits of their parents’ labour.

    If we want to protect authors’ families, and if we decide that copyright should be long enough to do so, surely there are better and more certain means by which we determine that period of time. Sound recordings (among other intellectual property rights), though not a traditional work of copyright, are protected on a fixed term from date of publication. So it is possible to conceive of a protection for creation which lasts for a known and finite stretch of time. Death on the other hand, albeit certain as an outcome, is otherwise entirely uncertain. So, we don’t know when or how authors will die, nor can we know how successfully they’ve obtained enough of their just reward during their lifetime to support their potentially penniless descendants, as in In re Dickens the oft-cited source of posthumous protection justification.

Leave a reply

Your email address will not be published. Required fields are marked *

1 + fourteen =