"Fold up your PFA [protection from abuse order] and put it in your pocket," Anthony Elonis wrote on his Facebook. "Is it thick enough to stop a bullet?"

When Elonis started posting what he claimed were "rap lyrics" on Facebook, the people who knew him didn’t take the posts to be mere homages to Eminem. They — his coworkers, his boss, his soon-to-be ex-wife — saw something sinister in his words, something dangerous enough that law enforcement ought to get involved.

After the FBI showed up to question Elonis at home, he responded by posting this:

Elonis was convicted "of transmitting in interstate commerce a ‘threat to injure the person of another.’" He appealed, claiming that prosecutors should have shown that he had intended to threaten. After losing at the Court of Appeals, he again appealed to the Supreme Court. On June 1st, the Supreme Court sided with Elonis — meaning he may be entitled to a new trial.

What Elonis wants — and what he’ll get, now that the Supreme Court has ruled in his favor— is a "heightened mental state requirement" in the law, meaning that prosecutors can’t just show that "a reasonable person would have known" that the communications were threatening. Instead, they have to show a subjective mental state on Elonis’ part. It doesn’t mean they have to read his mind, or his diary, or elicit a confession from him. A subjective state of mind can be inferred from the words themselves, from the context, from a person’s outwardly perceived behavior.

In theory, a "heightened mental state requirement" means an additional burden on prosecutors, meaning it’s harder to go after people like Elonis. But in practice, it’s not even obvious that prosecutors had to bring any additional evidence to Elonis’ trial. If they had just slightly tweaked the wording of the jury instruction, the entire trial and subsequent conviction could have played out exactly the same way, and Elonis wouldn’t have been able to appeal it, either. "You can infer what a person's state of mind is from the circumstances of how and what was said in words, correct?" Justice Sonia Sotomayor asked Elonis’ attorney during oral arguments in December. "So if that’s the case, isn't the jury acting like a reasonable person in looking at the words and the circumstances and saying, did he intend this or didn't he? I mean, I don't know what the difference between the standard given and the one [you want would be]."

Even though the Supreme Court sided with Elonis, it’s not "good news for abusive trolls," as some outlets have phrased it. Elonis could have been a landmark First Amendment case if it had revisited the definition of a "true threat," a legal concept that is in dire need of elaboration, but what ended up happening is that the court didn’t really do much of anything at all. Elonis stands for the proposition that the government can’t convict someone of a crime under 18 USC §875(c) using just a "negligence standard," also known as an objective or "reasonable person" standard.

This isn’t interesting or new. In general, negligence isn’t enough to uphold any kind of criminal conviction. "This is the standard with virtually every other crime," Lee Rowland, an attorney with the American Civil Liberties Union, said to The Verge back in November. "If you really think that there’s a fire in a theater, and you yell fire, you haven’t committed a crime, even if there’s no fire."

So the burning question on everyone’s mind was: if not negligence, what standard is required to convict someone under 18 USC §875(c) — or more importantly, what’s the standard for true threats in general? Do you have to purposefully intend a threat to be threatening? Is it enough to just know it will be threatening? Or can you be convicted simply for recklessly disregarding the risk that it will be threatening to the recipient?

As of this decision, the official answer is: "Who knows?" The majority opinion says that it’s definitely more than negligence, but explicitly declines to decide what the actual standard is, choosing instead to wait for more cases to filter up through the appeals courts.

In a separate opinion, Justice Samuel Alito blasted the court for refusing to elaborate which standard, saying their indecision was "certain to cause confusion and serious problems."

For Alito, the answer is obvious — they should have settled on the recklessness standard, just one step above negligence. Alito’s part-concurrence, part-dissent takes special care to emphasize the justice’s own personal disgust for Elonis. The subtext of his opinion is incredulity and outrage that anyone could take Elonis at his word, that the man could possibly get away with any of it — an unusually emotional reaction from someone who regularly upholds death sentences for a living.

Elonis is a deeply unsympathetic character. Based on the facts of his case, it’s difficult to see why a heightened mental state requirement is important for prosecuting speech on the internet. But take, for example, the teenaged League of Legends player who "threatened" on Facebook to shoot up a kindergarten and was subsequently jailed. (The postings were supposedly followed by "J/K" and "LOL.") The person who reported the post lived in Canada, had no relation to the teen, and was unfamiliar with how players regularly troll each other. A "reasonable person" standard, even a recklessness standard, could net many convictions for similar outbursts on the internet. Trolling isn’t exactly laudable, but in many cases, it doesn’t merit a prison sentence either.

The problem with using Anthony Douglas Elonis as a test case is that it’s hard to believe anything that comes out of his mouth. He says it was all a joke, and some of his postings even link to the Wikipedia article for the First Amendment. But simply tacking "just kidding" onto a Facebook post doesn’t make it non-threatening, especially when graphic verses are interspersed with ASCII diagrams of his estranged wife’s new home and detailed descriptions of which side of the house he’s going to shoot her from.

What Alito and Thomas (the latter of whom wrote his own dissent, siding with prosecutors) fear is that the "just kidding" might be enough to exonerate Elonis, if juries are required to find some form of heightened criminal intent in Elonis. They don’t think it’s possible that 12 people could read the Facebook posts, hear the testimony of his boss, his wife, the FBI agent, and come to the conclusion that despite the "just kidding," Elonis was still purposefully terrorizing his wife.

Yet they themselves can look at the record and find his behavior troubling. "Imagine the effect on Elonis’ estranged wife when she read this," Alito asks of his audience, when neither we nor he know Elonis’ wife. Elonis knew his wife better than any of us. Is it really so impossible to look at the words, hear the testimony, and come to the conclusion that he knew exactly what he was doing?

The Elonis decision isn’t a free pass for domestic abusers or online harassers. And it’s far from the end of the story. The Supreme Court didn’t make its mind up this time, but as legislators and law enforcement take a greater interest in online threats, similar cases are bound to go up in the future. Perhaps the court’s indecision was for the best. Elonis was always a bad vehicle for a very good question. The cases to come might better demonstrate the benefits, trade-offs, and dangers that inhere in prosecuting online speech.