Toil and Trouble: The Curse of Macbeth

by Brian Dunning
September 7, 2010

They simply call it "The Scottish Play", because even to utter the title of Shakespeare's Macbeth is to invite bad luck. The very same bad luck, in fact, that has plagued performances throughout its history, according to theater lore. From tragedies onstage to deaths and riots surrounding performances, the curse of Macbeth is one of the most enduring superstitions of the stage, and seems to be taken quite seriously.

The basic claim is that performing Macbeth, or even speaking its title in a theater, invokes an ancient curse as old as the play. This curse strikes actors or other people associated with the performance, sometimes killing or maiming them. This curse, so goes the tale, has its roots in the play's occult storyline of witchcraft, murder, and ghosts. The most often cited reason for the curse is a belief at the time that Shakespeare had used real witches' incantations in the famous scene where the three witches chant:

Double, double toil and trouble;
Fire burn, and cauldron bubble.

The legend is that Shakespeare wanted to throw something special into the play to please King James, who had written the 1597 book Daemonologie which discussed witchcraft and warned against its use. So Shakespeare used some of King James' documented incantations in the scene, probably hoping to ingratiate himself with the King. No good evidence exists for this, but the story maintains that some practicing witches saw the play and took great offense at this misuse of their sacred craft, and placed a curse upon any who might perform Macbeth. Now, whenever the play is given, the three witches whose spells were appropriated are awoken and it is they who cause the disasters onstage.

If you search the Internet for examples of such disasters, you'll find plenty. I could rattle on for 25 minutes just listing some of the many that have been reported: Actors being killed or injured during the stage fights when real weapons were used by mistake, natural disasters happening during performances, accidents and illnesses striking the crew before and after shows; the list goes on and on until you're bored to death and would be glad to count yourself among the casualties. One such episode, however, deserves special mention for its extraordinarily high body count.

Often cited as the most dramatic evidence of the curse is a riot that erupted at the Astor Place Theater in New York in 1849. The National Guard was called and fired on the crowd, killing at least 25 people and injuring some 120, all due to rival support for two different actors playing Macbeth on the same night at two different theaters. At least, that's usually how it's framed by fans of the curse. In fact, the Astor Place Riot had everything to do with class struggles in New York City, and little to do with Macbeth. Unrest had been growing for years between the working class, which included many Irish immigrants, and the Anglophile upper class. The discontent was coming to a head, and the National Guard was already in place some days before the actual riot on Macbeth's opening night. Irish and American workers planned to express themselves by crashing the opening night of the upper class's favorite British actor, William Macready. They stoned and tried to burn down the theater, people started shooting guns, and by morning the cobblestones were awash with blood. The next night an angry mob demanded an explanation from the authorities, and more violence ensued, this time resulting in the death of a young boy.

The Astor Place Theater had been built, apparently, largely as a way for the well-heeled to have somewhere to go other than the Bowery Theater, traditionally the principal theater in town, but which catered to all classes. The rising American star with a blue collar image, Edwin Forrest, planned his opening on the same night as Macready's largely as a slap in the face to this rising elitist sentiment. So although the riots were technically touched off by performances of Macbeth, the play itself had nothing to do with them. The riots were due, and would have happened whether Macbeth existed or not. If the three witches had chosen this particular performance to cast spells and cause trouble, they would have been well advised to hide under a table.

Many people have tried to put forth rational explanations for the events attributed to the curse. Often cited is that Macbeth has a lot of dim lighting and fight scenes using stage weapons. Such weapons are still dangerous, just not very sharp; and you're bound to have statistically more injuries in any play that has weapon fights than in plays that don't. Statistically, we should also expect more falls and other onstage accidents in plays with dim lighting. Even in brightly lit plays, it's hard enough to see what you're doing onstage because of the stage lights shining in your eyes; in a dimly lit scene, you could easily be practically blind. I don't really buy the dim lighting explanation. Granted my own stage experience is fairly limited, but when I'm brightly lit is when I have the hardest time seeing. Dim lighting, even no lighting, lets my eyes adjust and I can see my way around much better than when I'm blinded by spotlights. Other stage performers' experiences may vary.

But let's stop here and think back to the skeptical process. One of our fundamental rules is that before trying to explain a strange event, you must first establish whether that strange event ever actually happened. In this case, we need not bother looking into the validity of the curse, or any other such thing, unless and until we've established that there is in fact a history of mysterious accidents associated with the performance of Macbeth that deviates beyond the range of what typically happens in plays.

I was inspired by an earlier success I had when researching the curse of King Tut, when I discovered that a doctor had performed a retrospective cohort study on the people who were alleged to have fallen victim to that curse. He discovered that, when analyzed properly, the curse (if it existed) was not a terribly effective one. The lifespans of those who were exposed to the curse did not significantly differ from those who were not exposed. Encouraged by the publication of this study (it was in the British Medical Journal), I turned to all the scholarly sources to see if anyone had performed a proper statistical analysis of theater accidents, ideally involving Macbeth. I even assigned this task to my backup research team, a Google Groups list to whom I'll throw a question or two on occasion when I have trouble tracking something down. It's a heck of a list; hundreds of scientists and researchers in virtually every discipline, but even this mighty team came up short. We couldn't find any such research published anywhere. There is no end to scholarly articles discussing the curse: Lists of tragic events, Shakespeare's history with King James and the witches, and how to combat the curse (leave the theater, spin around, spout some profanity); but not a whisper inquiring into the proving curse's existence.

So, college students, there's a research project for you. This would not be easy. First you'd have to eliminate things like natural disasters that can't reasonably be attributed to the performance, and things like accidents striking people weeks after the play. To include these would require you to also correlate any other plays the victim may have attended, since it makes just as much sense to blame the accident on a different play he may have also seen in the same time period. Theaters may have records of accidents occurring on their premises, and those dates could certainly be matched with whatever was being performed. The number of cast and crew required would likely impact the chances of an accident for any given show, as would the use of dangerous equipment like lights, trap doors, flying harnesses, trickery scenery, and stage weapons. All of these things would need to be taken into account. It's little wonder that we couldn't find a record of such a study being performed.

Even the longest of the published lists of tragedies associated with Macbeth performances does not seem surprising, considering that we're talking about one of the world's most popular plays that has been performed constantly worldwide for more than 400 years. But since it appears that no analysis has been done, we can't conclude for certain that Macbeth is any more or less dangerous to perform than any other play.

The longest lists I've seen include perhaps twenty or thirty tragedies. Considering that there have been unknown tens of thousands (possibly hundreds of thousands) of Macbeth performances, the curse (if it exists at all) appears to be decidedly impotent. An actor probably stands a greater chance of having a car accident on his way to the theater than he has of being struck by the curse of Macbeth — that's another research project for a student.

Therefore, since we can't establish that the curse exists, we don't yet have a confirmed phenomenon to explain. We have a tall tale, told and retold over the centuries, and sufficient reason to suspect that wizened veteran actors may enjoy having a little fun at the expense of the newbies, perpetuating the story of a curse regardless of whether the actual number of associated accidents deviates from the norm.

Let's also keep in mind that the legend of the curse began almost right away when Shakespeare originally opened the play. Shakespeare was probably not a fool and knew that there's no such thing as bad publicity. What show's ticket sales would suffer if word got out that one of the actors might be accidentally killed during the performance? A little curse never hurt anybody (well, maybe a few).

So the curse of Macbeth: fact or fiction? I'm going to remain unconvinced that there's anything extraordinary going on, but will eagerly take a look at any good research that emerges. A list of anecdotes on the Internet is insufficient to prove that a supernatural force must be in effect.

By Brian Dunning
Follow @BrianDunning