Emoticons—such as :-)—and emoji—such as —are ubiquitous in online and mobile communications; according to one study, 74 percent of Americans use emoticons, emoji and similar images on a regular basis.

Given their popularity, it comes as no surprise that courts are increasingly being called upon to evaluate the meaning of emoticons and emoji that are included in material entered into evidence, an exercise that has highlighted just how subjective—and fact-specific—interpretations of these symbols can be.

For example, in a opinion last year dismissing a male law school student’s suit against local police and a female classmate for having the male student formally investigated based, at least in part, on text messages that he had sent, a federal district court in Michigan held that the male student’s text messages showed that he may have had an intent to harass the female classmate despite “the inclusion of the emoticon, “a ‘-D,’ which appears to be a wide open-mouth smile.” The court held that the emoticon “does not materially alter the meaning of the text message,” in which the male student otherwise wrote that he wanted to do “just enough to make [the female student] feel crappy.”

On the other hand, in a separate case, also arising in Michigan, the Michigan Court of Appeals held that the ‘:P’ emoticon accompanying a comment allegedly accusing a city worker of corruption made it “patently clear that the commenter was making a joke.”

Here are some other notable instances in which emoticons and emoji were among the evidence courts were asked to evaluate:

  • In a sexual harassment case brought by the female co-CEO of a Delaware corporation against her partner, the Delaware Chancery Court held in an opinion issued last summer that a “smiley-face emoticon at the end of [the defendant’s] text message suggests he was amused by yet another opportunity to harass the plaintiff.”
  • Earlier this year, during the trial of a California man accused of operating a black market called Silk Road over the Internet, the judge instructed jurors that they should take into account the emoji included in the social media posts and other electronic communications submitted into evidence, stating that the emoji are “part of the evidence of the document.” (The defendant, Ross Ulbricht, was ultimately convicted of all seven of the counts he faced; according to one news report, the government’s evidence that he ran “Silk Road’s billion-dollar marketplace under the pseudonym the Dread Pirate Roberts was practically overwhelming.”)
  • In a petition for certiorari by Anthony Elonis, a Pennsylvania man whose conviction for posting threatening status updates to Facebook was ultimately overturned by the U.S. Supreme Court last year, Elonis cited his inclusion of the emoticon “:-P” several times as part of his arguments that: (1) he lacked the intent required for conviction and (2) his posts were easily misunderstood and communications that are subject to misunderstandings shouldn’t be criminalized. Holding that “Elonis’s conviction was premised solely on how his posts would be viewed by a reasonable person, a standard feature of civil liability in tort law inconsistent with the conventional criminal conduct requirement of ‘awareness of some wrongdoing,’” the Supreme Court’s opinion didn’t mention emoticons at all, however.

These decisions highlight that the impact of a party’s use of an emoticon or emoji will turn on the specific facts of each case; context clearly matters. The larger point is that, in the words of WIRED’s Julia Greenberg, “When the digital symbol for a gun, a smile, or a face with stuck-out tongue comes up in court, they aren’t being derided or ignored. Emoji matter.”

And interpreting them will continue to be a challenge for courts. As the emerging case law illustrates, emoticons and emoji are often ambiguous, sometimes supporting the accompanying text, sometimes undermining it. Does that smiley face emoji with a tongue sticking out that accompanies a message purporting to confirm a deal signal the sender’s happiness that the deal has been concluded? Or does it indicate that the sender’s purported confirmation is intended to be a joke?

It’s been said that one shouldn’t send a message that one wouldn’t want to see on the front page of the New York Times. Is it time to adopt a corollary rule: Don’t include an emoticon or emoji in a message if it would inject ambiguity into an otherwise unambiguous communication?