The Patron Saint of Superheroes

[Update: SCOTUS ruled on May 18, 2023, which I discuss here.]

Last week I looked at two pixilated versions of the same Miles Davis cover art photograph. I think the first likely violates copyright, and the second would fall under fair use.

Why? Because, unlike the Prince, Obama, and Soglin examples discussed two weeks ago, the differences between the mid-range pixilation image and the original involves more than just simplification. The remaining details are also distorted in relation to the corresponding areas in the original. The size of units in ratio to the size of the object that the units combine to represent may matter. Once the size of the pixel-like squares is greater than the regions from the original that they represent, the squares become the primary element of composition. The resulting effect combines simplification with exaggeration.

The squares in the image that triggered the lawsuit are akin to paint strokes. They are the micro-level units that combine to create the gestalt effect of the macro-level image. I suspect the nature of smaller units doesn’t matter matter legally because resemblance occurs at the macro-level. This image, for example, has no representational relationship to its source material:

But when viewed at a different size/distance and within a larger image context, it resembles and so represents my right eye:

If I were suing myself for copyright infringement of only my right eye, I suspect I would lose.

Micro-level units matter when their qualities become dominant, including the macro-level effects they create in combination. Look at three pixilated Miles Davis hands:

The first is from what I identified as the first likely fair-use version of Baio’s Davis series last week. The middle is from the image that triggered the lawsuit, and the last is from the source photo. All three are made of pixels, but only the first two are considered “pixel art,” which is a misleading term since each large square is made of multiple identically colored pixels arranged in the shape of a significantly larger square to produce the effect of an enlarged pixel.

I also suspect that the top hand would not resemble a hand out of context, while the other two would be more identifiable. That could matter legally. The pixels that combine to suggest the qualities of Prince’s, Obama’s, and Soglin’s faces are essentially identical in the source images and the adapted images.

Fairey’s HOPE does involve some alterations, but the effect does not alter a primary experience of resemblance between the original and the adaption. The more pixilated Davis image moves into different terrain.

It moves into the top right of my four-area self-portrait by combining both simplification and exaggeration. Though there may be cases when an image that only simplifies its source is protected by fair use, I suspect the region generally is in legally dicey waters.

That leaves one area, the top left: exaggerated but not simplified.

And that, not coincidentally, describes my Prince art:

This is also helpful for discussing process, which is often in focus during court cases. And process differentiates actions (performed by an artist) from effects (experienced by viewers). I suspect viewers experience the Prince image as though it were hand-painted by an artist who was looking at the Goldsmith photograph as a visual reference. It wasn’t. The process is similar to Baio’s pixilated Miles Davis series.

I began with a digital version of the photograph, slightly pixilated after enlarging. I then selected a mouse-scribbled jigsaw shape, copied it, and pasted it imperfectly over the original so as to duplicate content along certain edges and obscure other content at opposite edges. I did this multiple times with multiple jigsaw shapes, while periodically saving works-in-progress:

After I settled on a final version in Word Paint, I opened the document in Adobe Illustrator, and saturated the colors, performing a final copy and paste in Word Paint to combine areas.

Although the “raw material” is still entirely the original photo, the final effect is different from Warhol’s adaptation because the placement of the duplicated material exaggerates certain areas, creating caricatural-like facial features. Even though my authorial intent wasn’t initially parody, parody arguably emerged during the process. Regardless, I suspect the resulting macro-level resemblance between my image and its source is sufficiently distant to fall under fair use due to the range of distinguishing exaggerations.

There is still some gestalt resemblance, since the question of infringement wouldn’t come up otherwise, but despite a process that uses the source as digital raw material, the two images bear almost no similarities at the micro level. Looking again at only right eyes reveals fundamental dissimilarities and so presumably no plausible infringement:

Distinguishing between micro-level units (such as paint strokes and pixels) and macro-level effects (experienced only when viewing an entire canvas) is similar to the differences between words and paraphrasable content. Copying words verbatim is a form of plagiarism, but it is still possible to plagiarize without reproducing any words. Ideas are copyrighted, not just the words that constitute them. If I express essentially the same idea (or gestalt image effect) using entirely different words (or paint strokes), I’m still potentially plagiarizing.

Look at the pixilated “Kiss” again. It’s taken from the 2015 film Eadweard about the 19th century photographer Eadweard Muybridge. The film recreates many of Muybridge’s images, including the moment represented in my hyper-pixilated “Kiss.” Since the film uses entirely different actors, the micro-level “raw materials” are unrelated to the actual individuals who appear in Muybridge’ work, but were Muybridge’s photographs still copyrighted, the film recreation would likely be an infringement. My “Kiss” could potentially infringe on both the 2014 film and the original 19th-century images–except that the level of simplification and exaggeration is so extreme, resemblance is minimal. The image is much more about its style (because of the size and shape of micro-level units) than its representational content.

And look at my second self-portrait again. Although the emergent “new expression, meaning or message” seems less prone to parodic effects, I suspect the complete dissimilarity at the micro-level and the partial dissimilarity at the macro-level would place the second image within fair use:

But I could be wrong, since the courts have yet to establish any clear standards, and decisions keep establishing contradictory or ambiguous precedents. Although the 2nd Circuit of Appeals recently warned judges to “not assume the role of art critic,” until the judiciary develops the necessary expertise of art criticism, fair use will remain in legal chaos.

[Spoiler Alert: This somehow evolved into a four-part analysis (one, two, three, four), with an on-going artistic coda starting here.]

[Update: SCOTUS ruled on May 18, 2023, which I discuss here.]

Last week I looked at three court cases where an artist used a copyrighted photograph to create a new work: Warhol used a Goldsmith photograph of Prince to create the Prince Series, Fairey used a Garcia photograph of Obama to create HOPE, and a group used a photograph of a Madison Mayor to create a t-shirt mocking him. A judge ruled Warhol was protected under fair use, but the ruling was overturned on appeal. Fairy settled out of court (after falsifying evidence which led to a separate conviction). And a judge and an appeals court both ruled that the mayor t-shirt was fair use.

Although the Second Circuit warned judges to “not assume the role of art critic,” some grounding in visual analysis might be helpful. To start, here’s a self-portrait I created for my forthcoming book The Comics Form: The Art of Sequenced Images (Bloomsbury 2022).

I analyze the representational relationship of an image to its source according to two poles, exaggeration and simplification, producing four combinations: 1) the bottom left corner is an actual photograph, so it is unsimplified and unexaggerated; 2) the bottom right corner is adapted from the same photo by erasing everything but the minimum lines needed to represent a face, so it is simplified but not exaggerated; 3) the top left corner is adapted from the same photo by variously expanding and rearranging details, so it is unsimplified and exaggerated; and 4) the top right corner is drawn using roughly the same number of lines as the image below it and with roughly the same degree of distortion as the image beside it, so it is simplified and exaggerated. 

I suspect no judge would question the legal legitimacy of a simplified and exaggerated image (top right), and I suspect any judge would find an unsimplified and unexaggerated image to infringe on copyrighted image (bottom left). That wasn’t always the case though. Roy Lichtenstein duplicated a section of page 3 from the 1963 comic Secret Hearts No. 83 drawn by Tony Abruzzo. Since Abruzzo was employed by Arleigh Publishing, which was owned by DC, DC owns the rights to the Abruzzo image, which Lichtenstein used without permission or attribution. I’m not sure what the statutes of limitations are on copyright infringement, but Lichtenstein’s are well past the half-century mark.

Sticking to this century, a court ruled against Richard Prince for using 30 photographs taken by Patrick Cariou in collage works in 2011. But then in 2013, an appeals court ruled in Prince’s favor, finding that 25 of Prince’s transformed images fell under fair use. The two parties settled the remaining five images out of court, including Graduation:

The appeals course was uncertain whether Graduation fell under fair use because it and the other four questionable works “do not sufficiently differ from the photographs of Cariou’s that they incorporate for us confidently to make a determination about their transformative nature as a matter of law. Although the minimal alterations that Prince made in those instances moved the work in a different direction from Cariou’s classical portraiture and landscape photos, we can not say with certainty at this point whether those artworks present a ‘new expression, meaning, or message.’”

Prince’s In the Garden is one of the 25 works that did not infringe:

Setting Prince aside (because how and how much an image or an image part changes meaning contextually is a wide open subject), the legally contestable and analytically confusing areas of my four-mode self-portrait are the remaining two squares (top left and bottom right).

Most of the above examples fall into the second category, simplified but no exaggerated. Warhol (for both the Prince and Marilyn series), Sconnie Nation, and Fairey removed considerable detail from their source photographs, but did not alter (or in the case of Fairey, altered minimally) the remaining details. The images (as far as how they relate to their raw material) are therefore simplified and unexaggerated. The adaptors also added details (mostly non-realistic colors), but those additions combine with but do not directly alter the details that remain from the sources.

The question then in each of those court cases is how much simplification (removal of detail) crosses the threshold of fair use?

Andy Baio asked the same question when he was sued in 2010 for using a photograph of Miles Davis taken by Jay Maisel. The photograph originally appeared on the cover of the 1959 Miles Davis album Kind of Blue, and Baio used a pixelated version of the image for a “retro videogame music” tribute album he produced titled Kind of Bloop.

Baio believed he was protected by fair use: “With regard to the third factor, although the illustration does represent the cover of Kind of Blue, it does so at a dramatically reduced resolution that incorporates few of the photograph’s protectable elements.” The pixilation also matches the purpose of the album: “to engage both artist and viewer in the same exercise — can NES-style pixel art capture the artistic essence of the original album cover, with a fraction of the resolution and color depth of an analog photograph? It reinforced the artistic themes of the project, to convey the feel of an entire album reimagined through an 8-bit lens. Far from being a copy, the cover art comments on it and uses the photo in new ways to send a new message.”

But litigation is expensive even when a defendant wins, and the risks of losing can be financially catastrophic. Baio settled out of court. He also asks on his blog: “Where would you draw the line?”

The question is difficult to answer in part because the law doesn’t consider under what condition an image is viewed. Consider this abstract grid of multi-colored squares:

If you’re looking at a computer screen right now, move a few yards back. If you’re looking at your phone screen, just extend your arm and squint. Or look at the identical image when reduced in size, and you’ll see why I titled it “Kiss.”

A low-resolution image looks low resolution only when viewed at sizes large enough (or distances great enough) to reveal the distortions produced by the low resolution. If you stand far enough away from Banksy’s Cardinal Sin, pixel-like squares on the face of the statue will look a normal face.

Courts often allow thumbnail reproductions (2003 Kelly v. Arriba-Soft, 2006 Graham Archives v. Kindersley), but even though a postage stamp is smaller than a square inch, the 2010 Gaylord v. United States ruled that the postal service needed the sculptor’s permission to reproduce an image of the Korean War veteran’s memorial. That’s presumably because thumbnails are low resolution, and stamps are high resolution—though you may need a magnifying glass to notice.

Pretending there’s a legal standard for “under what conditions” that accounts for image size and viewer distance, I wonder how Baio’s question applies to Goldsmith’s photograph of Prince. Though any digital enlargement is pixilated, the effect can be so mild that it goes mostly unnoticed:

At the other extreme, when the pixilation is so great the face is no longer recognizably Prince, we should be safely in the zone of fair use.

It’s the middle region that’s ambiguous and so legally treacherous.

With Baio’s examples, I think the first four images likely infringe on copyright, and probably the next two as well.

I think the last four likely fall into fair use, and probably the remaining middle two.

Compare that first (according to me) fair use image to the original and to the one that Baio actually used:

Why draw the line there?

I’ll try to explain that next week.

[Spoiler Alert: This somehow evolved into a four-part analysis (one, two, three, four), with an on-going artistic coda starting here.]

[Update: SCOTUS ruled on May 18, 2023, which I discuss here.]

[Update: Since I wrote this series of posts in spring 2021, an appeal went to the Supreme Court in October 2022, which I discuss here and here.]

That’s an image of Prince. I made it. How exactly I made it I’ll talk about later (because process may or may not matter legally), but I “used” a 1981 photograph of Prince taken by Lynn Goldsmith:

Is my use of Goldsmith’s photograph “fair use” or am I violating copyright law? It’s a basic question that lacks a basic answer. As a starting point, the relevant portion of the legal code, 17 U.S. Code § 107, has been law since 1976:

That could be a problem for the Warhol estate, since some of the most famous and expensive Warhols apparently are not protected either.

The raw material of Warhol’s numerous “Marilyn” paintings is a photograph taken by Eugene Korman to be used as publicity for the 1953 film Niagara. Korman was employed by 20th Century Fox, who, based on The Andy Warhol Foundation v. Goldsmith, could sue the Warhol estate for financial compensation (Orange Marilyn sold $17.3 million in 1998) and control of Warhol’s artwork.

For his Obama political poster HOPE, Shepard Fairey used a photograph taken by Mannie Garcia while covering a press conference for the Associated Press. The Associated Press sued Fairey and the two settled out of court in 2011. (Fairey later pled guilty to fabricating evidence that he had used a different photograph.)

Since the very different purpose of HOPE was to help Obama win the 2008 Democratic primary, factor one (purpose) seems especially significant. Looking only at the third factor though, the degree to which the raw material of Garcia’s photograph remains recognizable is similar to the Warhols.

But Warhol Foundation v. Goldsmith is not the only precedent. In the 2014 case Kienitz v. Sconnie Nation and Underground Printing, the circuit judges ruled in the opposite direction. The defendants downloaded a photograph from the Madison, Wisconsin city website to use for a t-shirt design ridiculing Mayor Peter Soglin.

The appeals court upheld the lower court’s ruling that the image was protected by fair use. Though they felt “the most important usually is the fourth (market effect),” the third factor, “the amount taken in relation to the copyrighted work as a whole—has much bite in this litigation.” They write:

“Defendants removed so much of the original that, as with the Cheshire Cat, only the smile remains. Defendants started with a low resolution version posted on the City’s website, so much of the original’s detail never had a chance to reach the copy; the original’s background is gone; its colors and shading are gone; the expression in Soglin’s eyes can no longer be read; after the posterization (and reproduction by silk-screening), the effect of the lighting in the original is almost extinguished. What is left, besides a hint of Soglin’s smile, is the outline of his face, which can’t be copyrighted. Defendants could have achieved the same effect by starting with a snapshot taken on the street.”

The court’s description is odd. Only in a hyperbolic sense does “only the smile remain” from the source photograph. Though it’s technically true that “so much of the original’s detail never had a chance to reach the copy,” the amount of detail in the low resolution version is highly significant. It’s also true that the defendants could have taken a snapshot of Soglin on the street and used it as their source material, but they didn’t, and whatever “effect” that hypothetical image might have had seems irrelevant. More importantly, there is considerably more “left” than “a hint of Soglin’s smile” and “the outline of his face.” And most importantly, it seems highly questionable whether the t-shirt image only contains content “which can’t be copyrighted.”

If outlines and hints of features can’t be copyrighted, then the Picasso estate cannot control rights to many of the artist’s drawings:

Outlines of faces and hints of their features also describe cartoons. Can the following not be copyrighted either?

What about outline-defined logos?

This would seem to relate to the Los Angeles News Service v. KCAL-TV decision which rejected fair use because the duplicated portion of a videotape “was the heart of the work.” In terms of single-image representation, some details always matter more than others, so “the heart” would presumably refer to only those areas that make portrayed faces recognizable. Yet Warhol Foundation v. Goldsmith and Kienitz v. Sconnie Nation draw opposite conclusions. While the amount of detail contained in the Warhol and the t-shirt appears objectively similar, the two panels of judges described them as diametrically different. For the image of Prince, “there can be no reasonable debate that the works are substantially similar,” but for the image of Mayor Soglin, what remains of the original is so scant it “can’t be copyrighted.”

I began this post expecting it to run my usual 1000-word range, but I'm close to doubling, with apparently lots more to go. So more next week ...


[Spoiler Alert: This somehow evolved into a four-part analysis (one, two, three, four), with an on-going artistic coda starting here.]

How many artists have created their own genres? Robert Sikoryak may stand alone in that category, especially for genres within the comics form. He has an eloquently simple concept: combine a set of words with incongruous drawings in the styles of famous comics.

For his first 2009 graphic novel, Masterpiece Comics, Sikoryak retold classic works of literature, such as The Scarlet Letter, Doctor Faustus, and Crime and Punishment, featuring Little Lulu as Hawthorne’s Pearl, Garfield as Marlowe’s Mephistopheles, and Batman as Dostoyevsky’s homicidal protagonist.  In 2017, Terms and Conditions earned Sikoryak greater attention for an even stranger premise: the complete, unabridged iTunes user agreement with Steve Jobs drawn in 94 pages of constantly changing styles. For The Unquotable Trump, released later the same year, he applied his formula to political satire, inserting Donald Trump cartoon images and verbatim quotes into comic book covers, with an appropriate emphasis on supervillains.

Now Sikoryak delves even deeper into American politics by adapting the most central U.S. text. Constitution Illustrated provides the complete, unaltered Articles and Amendments in 114 cartoon vignettes. The book is both Sikoryak’s widest range of comics homages yet and, more oddly, his most practical. Where the iTunes contract was a comically absurd choice because so few people have ever bothered to read it, the Constitution is a keystone of U.S. law and culture. Sikoryak even evokes a pocket-sized edition, that ubiquitous prop used by politicians and pundits in need of something to clench and wave above their heads.

I just used my own copy to check whether the 19th Amendment established the right of women to both vote and hold office or just to vote. The page features a spot-on imitation of H. G. Peter, the first but uncredited Wonder Woman artist. That pairing is a good illustration of Sikoryak’s logic and humor. Though unlike the adaptions in The Unquotable Trump, the page isn’t an exact recreation (like John Romita’s 1975 The Hulk on the Rampage cover), but a formally freer combination of style and subject. (The Amendment was just to vote.)

If you’re a comics aficionado, Constitution Illustrated is also the ultimate pop quiz. I didn’t keep score as I flipped through the first time, but I chuckled when I recognized the logic behind each discordant pairing, especially the superhero motif.

For Article I, Section I describing the division of Congress into the Senate and the House, Sikoryak draws two muscular and oppositely colored patriots sprinting in a mirrored pose cribbed from the 1976 cover of “The Greatest Race of All Time! Superman vs. the Flash.” The two DC heroes are allies on the same team, but they still compete against each other all too often. That antagonism increases when the House’s Super Friends face a row of Senate supervillains, an illustration of the House’s sole power to create tax-raising bills and the Senate’s power to amend them.

Instead of Spider-Man’s antagonists Prowler and J. Jonah Jameson watching Peter Parker fall from a window, Sikoryak draws a presiding Chief Justice and a Senator watching the President in the same pose—an apt illustration for the protocols for trying an impeachment. President Parker bears no resemblance to either Donald Trump or Bill CIinton, but Sikoryak kindly adds tingling spider senses emanata as a helpful clue (something artist John Romita did not include on the original 1969 cover).

A 1943-based colonial Captain America blocks a spray of musket bullets—metaphorically blocking the states’ ability to wage war, a power exclusive to the federal government. Sikoryak leaps to 1992 for Article II, Section 2’s description of the President’s role as Commander in Chief. I admit I didn’t recognize Jim Lee’s Wild C.A.T.s cover, just the decade-defining style which I took for Rob Liefeld. Happily, Sikoryak provides a cheat sheet in the appendixes, listing the source for each of adaption.

The list is dizzyingly eclectic: Alison Bechdel, Garry Trudeau, Roz Chast, Art Spiegelman, Chris Ware, Charles Schultz, Frank Miller, Scott McCloud, Adrian Tomine, and on and on. It’s a master course in comics and cartoon history, featuring some of the earliest creators, like Richard Outcault (The Yellow Kid) and Windsor McCay (Little Nemo), and some of the most recent, like Noelle Stevenson (Lumberjanes) and Bianca Xunise (Six Chix).

Given the current cultural moment of Black Lives Matter (the book was released during protests last summer), I took particular interest in Black artists George Herriman, Jackie Ormes, Matt Baker, Barbara Brandon-Croft, and Aaron McGruder, as well as the presentation of Black characters by non-Black artists. The Jaime and Gilbert Hernandez tribute is the most diverse, implying a hope that the Electoral College which it describes should reflect the same level of diversity.

Less subtlety, Sikoryak draws a chain-breaking Luke Cage to illustrate the slavery-ending 13^th Amendment. His 25^th Amendment depicts a Black vice-president assuming the presidency—just as the Black character John Stewart assumed the role of Green Lantern in DC comics.

My disturbing favorite though is Mandrake the Magician turning his African servant Lothar partially invisible beneath the census directive to count only “three fifths of all other persons,” meaning slaves. The image unites the racism of the Article with the racism of the 1930s characters. It also highlights how any contemporary analysis of the Constitution must address its deep flaws too. Sikoryak’s satirical pairings breathe new and sometimes uncomfortable life into the United States’ most living document.