Attributions, anticipations, and plagiarism in chess compositions

Composing chess problems is a creative art that takes skill, time, and effort, so it’s only natural that when an original work is published, it’s always accompanied by the author’s name. It’s also normal, when quoting a problem, to identify the composer and provide the source and date of first publication. These standard practices, however, are not always followed – whether by accident or intention – and the resulting difficulties are in various degrees the bane of the chess problem world. Let’s take an informal look at the three main types of incorrect attributions: (1) the lack of attribution at all when presenting a composed problem, (2) anticipation, where a composer unknowingly recreates another problemist’s work, and (3) plagiarism, where a person claims credit for a composition produced by someone else.

Attributions (lack of)

The chess boom of recent years has led to a noticeable increase in the dissemination of chess compositions, especially online. It’s probably unrealistic to expect most players, especially novices, to mention the composer when posting problems in places like chess forums and Reddit. Indeed, many players have yet to learn the distinction between composed problems and tactics puzzles. (In a positive step, Reddit has introduced an excellent AI-tool that automatically displays the source of a posted composition, and even links to its entry in the YACPDB database.) However, when professional content creators like those on YouTube and Twitter present chess compositions, it’s ethically important to give credit to the problemists responsible – but many don’t, unfortunately.

Such online content creators earn money by entertaining their followers with these chess positions, so it’s dubious not to acknowledge their real authors. Furthermore, the mass audiences of these platforms include plenty of newcomers who may falsely assume that the expert or titled player explaining a composition is responsible for making it. In other words, a sort of unintentional plagiarism could be occurring. Take the case, in the analogue world, of Chess: 5334 Problems, Combinations and Games, a popular book that includes hundreds of composed problems (apart from the tactics puzzles) with no indication of their sources. An average player picking up the book could be forgiven for believing that the person named on the cover, Laszlo Polgar, had devised these positions.

I think the most egregious instance of inappropriate attribution in chess compositions is the so-called “Plaskett’s Puzzle”. This is the brilliant endgame study by the Dutch problemist, Gijs van Breukelen (1946-2022), one that has captured the imaginations of chess players and garnered countless reposts and discussions. Absurdly, the study acquired its name simply because the GM James Plaskett posed it as a challenge to the competitors of a top-class tournament in 1987. Imagine a scenario where Plaskett has played a sensational game that ends with a terrific combination, and then van Breukelen showed it to his fellow problemists as a solving task. Does that mean Plaskett’s creation should become known as “van Breukelen’s Game”?

Anticipations

An anticipation occurs when a composer publishes a problem that duplicates another person’s work, without knowing of the precursor. The newer version is said to be anticipated and it becomes void if it wholly or largely reproduces the earlier position. A partially anticipated problem has sufficiently original aspects to remain valid but its value is usually downgraded. Anticipations are more likely to happen in light problems with few pieces that show a clear-cut idea, but this kind of situation can also occur in heavy positions demonstrating more complex themes. Even more surprisingly, there are known cases of composers independently publishing nearly identical problems at the same time; such instances of synchronicity are termed coincidences.

In brief, anticipations arise because problemists gravitate towards the same worthwhile themes and if they hit on similar initial matrices, a certain logic in the construction process can lead them to arrive at practically the same setting. Most established composers probably have been struck by this type of “misfortune” at some points. In this regard I may be typical in producing a few anticipated works in my formative years, and then with experience learned which ideas are well-worn and should be avoided. Nowadays, online problem databases make it easier to check for anticipations before publishing one’s works, though employing such tools doesn’t guarantee originality.

As illustrations, it’s not commonly known that the two most famous compositions by Raymond Smullyan (1919-2017) were anticipated. These are the retro-analytical problems featured on the covers of his books, The Chess Mysteries of Sherlock Holmes (1979) and The Chess Mysteries of the Arabian Knights (1981). Although Smullyan published the latter problem prior in Manchester Guardian 1957, that was still years after the appearance of a very similar work.

In the first case, it’s apparent that Smullyan’s position is merely a reflection (with a colour change) of Mortensen’s earlier problem, thus the anticipation is complete. The original’s solution is as follows. It’s stipulated that Black is to play, so White has just moved and the king came from a2 (not adjacent to the black king). However, on a2 the king would be in an impossible check by the black bishop, which couldn’t have just played to g8 because of the h7-pawn. The only way the position could have arisen is that the white king has just captured a black knight on a1, and prior to that (with the king on a2), Black made a discovered check, …Sb3-a1+. Therefore the last move was Ka2xSa1.

The second problem by Smullyan is substantially anticipated by Pavlovic’s, which features the same retro-play regardless of the different task specified. Here the white bishop on h4 is checking Black, but the blocking g5-rook means that White didn’t move the bishop directly. Rather, White must have checked by discovery with the king coming from g3. If the last move was simply Kg3-f3+, however, the king would've been in an impossible double-check by the black rook and bishop (neither of which could have discovered check). Such a double-check could in fact come about only by means of an en passant capture. White’s last move was Kg3xPf3+, and before that Black delivered the double-check by playing …g4xf3 e.p.+. Retracting further, White played f2-f4 in answer to a bishop check on e5. This gem of a retro problem conceals an e.p. capture in an economical pawnless setting. Smullyan’s version, instead of posing the standard retro task of finding the last move, asks the solver to replace the missing white king on the right square. It's delightful that there’s only one legally viable square (c3, generating the Pavlovic position reflected), so arguably this is a slight improvement on the original.

For another classic example, see the second position in this blog about the shortest possible games to stalemate. Here the best capture-free game is often misattributed to Sam Loyd, when he was anticipated by Charles Wheeler.

Plagiarism

Plagiarism in chess problems, or misappropriating another person’s composition as one’s own, is of course reprehensible behaviour but it’s rare. For one thing, there’s no financial incentive (composing tourneys bestow “prizes” in name, not money), and for another, guilty parties are caught sooner or later. An obvious question here is how to distinguish between cases of anticipations and plagiarism. The short answer is that it’s a judgement call best left to experts. The clearest instances of fraudulent activities by known problemists are when their “anticipated” works exceed in numbers what’s reasonable by coincidence. It can also be apparent if a novice tries to copy a published problem, the quality of which makes it virtually impossible for a newcomer to reproduce.

Many years ago I came across a Chess.com user who posted a dozen endgame studies on the Forum and claimed to have created them. It took me just a few minutes to find most of these studies in online databases, which provided illustrious names like Rinck and Troitsky as their real authors. When I pointed this out on the Forum, he promptly closed his account!

The most infamous plagiarist in chess problem history is Percy Wenman (1891-1972). He published numerous problem collections, some of which have proper attributions, but in others where he claimed authorship of all the cited positions, large portions of them were in fact lifted from other composers. Researchers who have exposed his offences include the late problemist Cyril Kipping (when Wenman was still active), and more recently the chess writer, Edward Winter. Refer to the latter’s Chess Notes 5641, P. Wenman and plagiarism. To finish on a lighter note, I’m reminded of a jocular literary competition run by British chess problemists, to write a two-line verse about someone in the composition field, starting with “I am the ghost of [insert name]:”. The winning entry by Michael McDowell was:

I am the ghost of Wenman P.:
Some of my problems were made by me.