concurring:
I agree with the panel’s conclusion that our decision in Intervest Construction, Inc. v. Canterbury Estate Homes, Inc. (“Intervest”), 554 F.3d 914 (11th Cir. 2008), drives this ease and requires affirmance of the district court’s entry of judgment under Rule 50. But I think that Intervest represents a wrong turn in our Circuit’s copyright jurisprudence. Specifically, Intervest holds that judges are necessarily better able than juries to resolve whether the “average lay observer” would find “substantial similarity” between two architectural .works. But we ask juries to answer this same question in all kinds of other copyright cases. Because I do not see a basis for exempting copyright cases involving architectural works from jury trials simply because the question of “substantial similarity” may be close, I respectfully disagree with Intervest and would steer clear of its holding, were we not bound by it.
To establish copyright infringement, a plaintiff must prove (1) that it owns a valid copyright and (2) that the defendant copied original — meaning “protectable” — elements of the work. Oravee v. Sunny Isles Luxury Ventures, L.C., 527 F.3d 1218, 1223 (11th Cir. 2008). As the Supreme Court has explained, “The mere fact that a work is copyrighted does not mean that every element of the work may be protected. Originality remains the sine qua non of copyright; accordingly, copyright protection may extend only to those components of a work that are original to the author.” Feist Publ’ns, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340, 348, 111 S.Ct. 1282, 1289, 113 L.Ed.2d 358 (1991).
Where, as here, a plaintiff lacks direct proof of copying, the plaintiff may establish the element of copying by showing that the defendant “had access to the copyrighted work and that the works are ‘substantially similar.’ ” Herzog v. Castle Rock Entm’t, 193 F.3d 1241, 1248 (11th Cir. 1999). The test for “substantial similarity” is not at all technical. To the contrary, “substantial similarity” exists where “an average lay observer would recognize the alleged copy as having been appropriated from the copyrighted work.” Original Ap*1328palachian Artworks, Inc. v. Toy Loft, Inc., 684 F.2d 821, 829 (11th Cir. 1982) (emphasis added).
Because a copyright owner holds the rights to only those portions of his or her work that are original, the copyright owner must demonstrate “both [that] the similarities between the works are substantial from the point of view of the lay observer and [that] those similarities involve copyrightable material.” Oravec, 527 F.3d at 1224 (internal alteration omitted); see id. n. 5. In other words, the “substantial similarity” must exist at “the level of protected expression.” Id. at 1227.
Whether two works are “substantially similar” at the level of protected expression seems to me to be an inherently subjective and fact-bound inquiry. “At the most narrow, focused level, two works will almost always be distinguishable, and at the broadest level of abstraction they will almost always appear identical.” Baby Buddies, Inc. v. Toys R Us, Inc., 611 F.3d 1308, 1316 (11th Cir. 2010). As a result, “[l]ists of similarities between the two works are inherently subjective and unreliable, particularly where the list contains random similarities, and many such similarities could be found in very dissimilar works.” Herzog, 193 F.3d at 1257 (internal quotation marks omitted).
Indeed, long ago, Judge Learned Hand explained why the “substantially similar” inquiry will nearly always be subjective and ad hoc:
Upon any work ... a great number of patterns of increasing generality will fit equally well, as more and more of the incident is left out. The last may perhaps be no more than the most general statement of what the [copyrighted work] is about, and at times might consist only of its title; but there is a point in this series of abstractions where they are no longer protected, since otherwise the [copy-righter] could prevent the use of his ideas, to which, apart from their expression, his property is never extended. Nobody has ever been able to fix that boundary, and nobody ever can.
Nichols v. Universal Pictures Corp., 45 F.2d 119, 121 (2d Cir. 1930). Because “substantial similarity is an extremely close question of fact” the determination of which is by necessity subjective and ad hoc, “summary judgment has traditionally been frowned upon in copyright litigation.” Latimer v. Roaring Toyz, Inc., 601 F.3d 1224, 1232 (11th Cir. 2010) (internal citation omitted); see also Peter Letterese and Assocs., Inc. v. World Inst. of Scientology Enters., 533 F.3d 1287, 1302 (11th Cir. 2008) (“Historically, courts have hesitated to make determinations as to infringement or non-infringemertt on a summary judgment motion because of their reluctance to make subjective determinations regarding the similarity between two works.” (internal quotation marks and alteration omitted)).
Nor is the question of whether “substantial similarity” exists at the level of “pro-tectable expression” a question unique to copyright actions involving architectural works. Rather, it is one that we must answer in every copyright action, regardless of whether the object of the copyright is a book, a piece of artwork, or an architectural work.
Books and movies, for example, often include non-copyrightable elements such as scenes a faire — ’“sequences of events which necessarily follow from a common theme,” or “[incidents, characters, or settings that are indispensable or standard in the treatment of a given topic.” Herzog, 193 F.3d at 1248. So in cases involving those media, juries regularly must separate the unprotected elements from those that are copyrighted before determining whether “substantial similarity” exists between the original and the alleged copy.
*1329Cases involving “compilations”— “work[s] formed by the collection and assembling of preexisting materials or of data that are selected, coordinated, or arranged in such a way that the resulting work as a whole constitutes an original work of authorship,” 17 U.S.C. § 101— likewise require the factfinder to determine whether “substantial similarity” exists between the protectable elements of a compilation and an alleged copy. We have had no problem finding juries capable of distinguishing between protected elements and unprotected elements in these types of cases. See, e.g., BUC Int’l Corp. v. Int’l Yacht Council Ltd., 489 F.3d 1129 (11th Cir. 2007).
In Intervest, however, we departed from the rule of having juries decide the inherently fact-bound issue of whether two works are “substantially similar” and crafted a new rule for cases involving “architectural works.” 554 F.3d at 920-21. Specifically, we held that judges are generally better able to conduct this inquiry at summary judgment than jurors are at trial. Id. at 919-20.
In arriving at this new rule, we first opined that the Copyright Act of 1976’s definition of “architectural work” “closely parallels that of a ‘compilation’”; so, as with compilations, “any similarity comparison of [architectural works] ... must be accomplished at the level of protected expression' — that is, the arrangement and coordination of’ “common windows, doors, and other staple building components.” Id. at 919. Based on this analogy, we cited the Supreme Court’s decision in Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340, 349, 111 S.Ct. 1282, 1289, 113 L.Ed.2d 358 (1991), for the proposition that copyright protection in an architectural work, like a compilation, is “thin.” Intervest, 554 F.3d at 919.
Then we opined that judges, not jurors, are best equipped to conduct the “substantial similarity” inquiry in cases involving architectural works. Id. at 920-21. We reasoned,
[A] judge is better able to separate original expression from the non-original elements of a work where the copying of the latter is not protectable and the copying of the former is pro-tectable. The judge understands the concept of the idea/expression dichotomy and how it should be applied in the context of the works before him. As we have observed: “This distinction — known as the idea/expression dichotomy — can be difficult to apply, as there is no bright line separating the ideas conveyed by a work from the specific expression of those ideas.” Oravec [v. Sunny Isles Luxury Ventures, L.C.], 527 F.3d [1218,] 1224 [ (11th Cir. 2008) ]. Moreover, in examining compilations wherein only the arrangement and coordination of elements which by the nature of the work (here architectural floor plans) are sure to be common to each of the works and are not copyrightable themselves (special [sic] depictions of rooms, doors, windows, walls, etc.), the already difficult tasks may become even more nuanced. Because a judge will more readily understand that all copying is not infringement, particularly in the context of works that are compilations, the “substantial-similarity” test is more often correctly administered by a judge rather than a jury — even one provided proper instruction. The reason for this is plain — the ability to separate protect-able expression from non-protectable expression is, in reality, a question of law or, at the very least, a mixed question of law and fact. It is difficult for a juror, even properly instructed, to conclude, after looking at two works, that there is no infringement *1330where, say, 90% of one is a copy of the other, but only 15% of the work is protectable expression that has not been copied.
Id.
I think we lost our way in Intervest. While I agree with the majority that our decision in Intervest compels us to affirm the district court’s decision in this case, I believe the unique rule we crafted for architectural works is unmoored from traditional copyright jurisprudence. I find no reason that survives scrutiny which warrants treating the “substantially similar” inquiry in copyright cases involving “architectural works” differently than the “substantially similar” inquiry in other copyright cases.
True, in Feist, the Supreme Court explained that the protection afforded factual compilations is “thin” in the sense that the raw materials of a compiler’s medium— namely, facts — are not themselves copyrightable. 499 U.S. at 349, 111 S.Ct. at 1289. But this is just another way of saying that the factfinder must consider whether “substantial similarity” exists at the level of protectable expression — original content — only, and the amount of protectable expression relative to total content in a compilation is less than in a more original type of work. Whatever protectable expression a compilation contains, however, remains subject to the same copyright protection as the original content in all other types of copyrighted work. See id. at 349, 111 S.Ct. at 1290 (“[Cjopyright assures authors the right to their original expression, but encourages others to build freely upon the ideas and information conveyed by a work. ... This principle, known as the idea/expression or fact/expression dichotomy, applies to all works of authorship.”) (emphasis added).
Indeed, when Congress amended the Copyright Act to include “architectural works” in 1990, it did so because it recognized that “[architecture is not unlike poetry” and “concluded that the design of a work of architecture is [therefore] a ‘writing’ under the Constitution and fully deserves protection under the Copyright Act.” H.R. Rep. No. 101-735 (1990), as reprinted in 1990 U.S.C.C.A.N. 6935, 6941. Put simply, the protectable elements of an “architectural work,” though they may be fewer and therefore “thin” relative to other works, are nonetheless entitled to the full protection of the Copyright Act.
So the “thinness” or “thickness” of protected expression in a type of work merely defines the frame of reference for the “substantial similarity” inquiry. See Oravec, 527 F.3d at 1224 (explaining that the similarity inquiry targets the similarity between the protectable expression of a copyrighted work and the expression in an allegedly infringing work). But once we are looking at protectable expression, I see no meaningful difference in the substance of the “substantially similar” inquiry applicable to works entitled to “thick” protection, such as novels, and that inquiry applicable to works entitled to “thin” protection, such as architectural works. Regardless of the type of work at issue, a copyright plaintiff will always be required to demonstrate substantial similarity “at the level of protected expression.” Id. at 1227.
To the extent that Intervest suggests that the protectable expression in an architectural plan is somehow subject to less protection than the protectable expression in any other kind of copyright-protected content, the Second Circuit aptly summed up the problems with our approach in In-tervest.
Labeling architecture a compilation obscures the real issue. Every work of art will have some standard elements, which taken in isolation are un-copyrightable, but many works will have original ele*1331ments — or original arrangements of elements. The challenge in adjudicating copyright cases is not to determine whether a work is a creative work, a derivative work, or a compilation, but to determine what in it originated with the author and what did not. Intervest fails to do this. ...
Courts should treat architectural copyrights no differently than other copyrights. This is what Congress envisioned
Zalewski v. Cicero Builder Dev., Inc., 754 F.3d 95, 104 (2d Cir. 2014).
But the most troubling aspect of our decision in Intervest is our conclusion that judges are more able to conduct the inherently factual' and subjective “substantially similar” inquiry in architectural-works cases than jurors. Intervest, 554 F.3d at 920-21. I respectfully disagree with this determination.
The Seventh Amendment guarantees parties like Turner a jury trial. While the Copyright Act does not explicitly provide copyright plaintiffs a right to a jury trial, it does permit plaintiffs to recover either actual or statutory damages for violations of the Act. 17 U.S.C. § 504. And when a plaintiff seeks to recover either actual or statutory damages under the Act, the Seventh Amendment guarantees that plaintiff a right to a jury trial. Feltner v. Columbia Pictures Television, Inc., 523 U.S. 340, 355, 118 S.Ct. 1279, 1288, 140 L.Ed.2d 438 (1998); see id. at 346, 118 S.Ct. at 1284.
The Seventh Amendment, in turn, requires, where it applies, “that enjoyment of the right of trial by jury be not obstructed, and that the ultimate determination of issues of fact by the jury be not interfered with.” In re Peterson, 253 U.S. 300, 310, 40 S.Ct. 543, 546, 64 L.Ed. 919 (1920). Of course, it has long been recognized “that when the evidence given at the trial, with all inferences that the jury could justifiably draw from it, is insufficient to support a verdict for the plaintiff, so that such a verdict, if returned, must be set aside, the court is not bound to submit the case to the jury, but may direct a verdict for the defendant” without violating the Seventh Amendment. Randall v. Baltimore & O.R. Co., 109 U.S. 478, 481, 3 S.Ct. 322, 324, 27 L.Ed. 1003 (1883); see also Capital Traction Co. v. Hof, 174 U.S. 1, 13-14, 19 S.Ct. 580, 585, 43 L.Ed. 873 (1899) (“ ‘Trial by jury,’ in the primary and usual sense of the term at the common law and in the American constitutions, ... is a trial by a jury of 12 men in the presence and under the superintendence of a judge empowered to instruct them on the law and to advise them on the facts, and ... to set aside their verdict, if, in his opinion, it is against the law or the evidence.”).
But we do not remove factual determinations from a jury simply because the factual inquiry is “nuanced” or “difficult.” Cf. Intervest, 554 F.3d at 920. Instead, we issue specific instructions to educate the jury on the nature of its inquiry and presume that the jury follows those instructions. Fed. R. Civ. P. 51; Jamerson v. Sec’y for Dep’t of Corr., 410 F.3d 682, 690 (11th Cir. 2005) (“[W]e presume that juries follow instructions .... ”). And, where necessary, we ask the jury to return special verdicts breaking out each factual determination the jury must make to guide the jury in its task. Fed. R. Civ. P. 49. Parties also may present expert testimony to assist the jury in its evaluation of the evidence. See Fed. R. Evid. 702.
In fact, we routinely entrust juries with highly technical and complex factual inquiries, including, for instance, issues such as whether one party infringed another’s software patent, see Telecom Tech. Servs., Inc. v. Rolm Co., 388 F.3d 820 (11th Cir. 2004); whether a party possesses monopoly power in a relevant market or has sufficient economic power to coerce another into *1332buying a tied product in Sherman Antitrust Act cases, see Tech. Res. Servs., Inc. v. Dornier Med. Sys., Inc., 134 F.3d 1458, 1465-66 (11th Cir. 1998); and whether a party has established loss causation in a § 10(b) case under the Securities and Exchange Act of 1934, see Rousseff v. E.F. Hutton Co., 843 F.2d 1326, 1329 (11th Cir. 1988).
And we ask juries to make significantly weightier factual determinations than whether two works are “substantially similar,” including, for instance, a criminal defendant’s liability. Indeed, the Sixth Amendment demands that juries, not judges, make the weightiest of all factual determinations: whether capital punishment is warranted. Ring v. Arizona, 536 U.S. 584, 589, 122 S.Ct. 2428, 2432, 153 L.Ed.2d 556 (2002).
In short, when factual determinations are “nuaneed” or “difficult,” we educate juries and provide them with the tools necessary to do the job. We do not take the issues away from the jury and defer to judges. I am aware of no case law that stands for the proposition that we may more readily remove factual determinations from the jury’s purview when those determinations are, in the view of the judge, “difficult.”
But even if such precedent existed, I would not see the sense in generally removing the issue of “substantial similarity” from the jury in any kind of copyright case, including those involving architectural works. The standard for whether two works are “substantially similar” is whether “an average lay observer would recognize the alleged copy as having been appropriated from the copyrighted work.” Original Appalachian Artworks, Inc., 684 F.2d at 829 (emphasis added).
It is not clear to me why judges would be “better able to separate original expression from the non-original elements of a work where the copying of the latter is not protectable and the copying of the former is protectable.” Intervest, 554 F.3d at 920. I think it unlikely that many judges have architectural or even design experience. So we have no more practical, real-world understanding of the significance of particular design elements necessary to make a determination about whether a given architectural work is substantially similar to another at the level of protected expression than does a jury.
That is not to say that we cannot make informed decisions regarding these issues, upon reviewing appropriate evidence and applying the correct standard. Of course we can. But so can juries. Indeed, the very nature of the standard — “average lay observer” — suggests as much.1
I likewise respectfully disagree with any notion that judges have a special grasp on “the concept of the idea/expression dichotomy and how it should be applied in the context of the works before [them].” Id. No degree of mastery of the “concept of the idea/expression dichotomy” renders a judge better able to determine whether an average lay observer would recognize an alleged copy as having been appropriated from a copyrighted work.
Here, had we applied the same rules of copyright law that we use in cases involving copyrights on other types of works, the Rule 50 motion would have been properly denied because a material issue of fact existed. Specifically, an average lay observer could have concluded that, the Turner plans are substantially similar to HDS-2089, evidencing unlawful copying.
*1333Even a cursory glance at the blueprints reveals that, as HDS’s CEO testified at trial, the Turner plans are “virtually line for line” copies of HDS-2089. See App. This fact led HDS’s expert witness2 to testify that “the overall organization of traffic patterns for the home[s]” appears to be identical. HDS’s expert further testified that “the overall shape, the massing, the individual layout of the rooms is the same. They all have the same shape, width, and length. They have essentially the same massing .... They have the same organization ....” In other words, the selection and arrangement of common building elements in the Turner plans, in general, is nearly identical to the selection and arrangement in HDS-2089.
And beyond that, HDS’s expert also noted that the plans contain certain odd features that are hard to explain if Turner did not unlawfully copy from HDS-2089. For example, one of the bathroom walls in both HDS-2089 and the Laurent plan is slightly thicker than the surrounding walls. This feature has a practical purpose in HDS-2089, but it makes no sense in the Laurent plan: in the HDS-2089, the thicker wall provides a way to accommodate the plumbing for a bath tub, but in the Laurent plan, the tub is turned, so the plumbing for the tub necessarily would not run through the .thicker wall.
True, as the Majority points out, the plans have differences. The Majority notes, among .others, that the door to the laundry room in the garage swings inwards in HDS-2089 and outwards in the Turner plans, and the plans’ respective secondary bathrooms feature different countertops. See id. In Intervest, we held that similar sorts of differences as those identified by the Majority would have precluded a reasonable jury from finding “substantial similarity.” See 554 F.3d at 916-17. Respectfully,’ I disagree.
Instead, I would hold that “an average lay observer” could find these differences between the HDS-2089 and Turner plans to be immaterial in light of the plans’ otherwise “substantial similarity” at the level of protected expression. Indeed, the first trial judge denied Turner’s summary-judgment motion because the judge concluded that “there are ... myriad similarities in areas of protectable expression,” so a reasonable jury could find the works “to be substantially or even strikingly similar.” And, after a five-day jury trial, a jury of average lay observers did just that, concluding that all 165 of the Laurent and Dakota houses at issue are “substantially similar” to HDS-2089. To second guess the jury at this stage based on nothing more than a list of modest dissimilarities — the very sort of list that we have already rejected as “inherently subjective and unreliable,” Herzog v. Castle Rock Entertainment, 193 F.3d 1241, 1257 (11th Cir. 1999) — seems to me to unjustifiably usurp the role of the jury in copyright cases and deny owners of architectural work copyrights the full protection of the Copyright Act.
So while I agree with the majority that our decision in Intervest dictates that we uphold the district court’s decision granting Appellee’s Rule 50 motion, I think it time to revisit Intervest.
. Ironically, it seems more likely that at least some members of a jury would have architectural or design experience or training of some type and therefore actually have a practical understanding of the significance of various protectable design elements in an architectural work.
. HDS relied on the expert-witness testimony of Kevin Samuel Alter, Associate Dean for Graduate Programs and Professor of Architecture at the University of Texas at Austin.