Toward a Textualist Paradigm for Interpreting Emoticons

By: 

John S. Ehrett, J.D. Candidate, Yale Law School

Authored on: 
Tuesday, October 27, 2015

Introduction

As online communications have proliferated, discursive norms unique to the medium have emerged. The use of emoticons and, more recently, emojis—pictograms often conveying multiple layers of semantic meaning—has figured prominently in this process.[i] In the normal course of online interaction, Internet users routinely parse the symbolic significance of various emoticons and emojis.

Courts, in seeking to ascertain the intent behind, and implications of, statements made online by criminal defendants and other parties, have been forced to grapple with the interpretive nuances of this communication style. While the question is legally salient and has been the subject of analysis in the popular press,[ii] proposed constructive mechanisms for actually overcoming this problem have been in short supply. Here, I evaluate the dimensions of courts’ current interpretive dilemma, and subsequently sketch a possible framework for extending traditional statutory interpretation principles into this new domain: throughout the following analysis, I describe the process of attaching cognizable linguistic referents to emoticons and emojis throughout as symbolic reification, and propose a normative way forward for those tasked with deriving meaning from emoji-laden communications.

Basic Principles of Emoticons and Emojis

Emoticons and emojis, while treated similarly for this analysis, are technically distinct: whereas an emoticon is “a digital icon or a sequence of keyboard symbols that serves to represent a facial expression, as :‐) for a smiling face,”[iii] emojis are “one-character, non-alphabetic symbols.”[iv] From a technical standpoint, emoticons—such as “:)”—are formed from the combination of ASCII (traditional keyboard) characters,[v] whereas emojis are single-character Unicode pictograms. I use these terms semi-interchangeably in this analysis—with a preference for “emoji” due to the wider diversity of symbols to which one has access—since both emoticons and emojis are pictogram-style glyphs lacking direct verbal elements.[vi] The general principles used for interpreting emoji—which, as I subsequently argue, can be meaningfully linked to a cognizable external referent—should be understood as also explicable to earlier emoticons.

The Interpretive Conundrum

It is, as yet, an open question whether or not emoticons and emojis carry independent interpretive significance for legal purposes.[vii] This issue has even been raised at the highest levels of the judicial system: in Elonis v. United States, court filings discussed the question of whether or not a potential “threat” might be mitigated by the presence of an emoticon.[viii] An answer in the affirmative—that emoticons and emojis do have an intrinsic significance or latent meaning—necessarily forces courts to develop a paradigm for ascertaining said meaning.[ix] One significant interpretive challenge, for example, involves the difficulty in identifying the specific actor to whom the emoticon or emoji is meant to refer. The image of an “angry face” might reflect the speaker’s own state of mind, the state of mind of the addressee, or may reflect a statement being made about a third party.[x] In and of themselves, emojis do not contain an overt structure for expressing personal and possessive pronouns.

Outlining a Reification Pathway

Diverse articles in the popular press have explored the problem of court interpretation of emoticons and emojis, but have largely not ventured beyond identification of the intrinsic ambiguities involved. I suggest in this section a specific methodology that courts ought to adopt as a way of beginning to resolve such ambiguities.

Symbolic reification requires some bounding principle in order to allow for reasonable consistency in interpretation: attempting to nail down evanescent cyberspace norms about what a given emoji “inferentially communicates” muddies the interpretive waters. Accordingly, I propose that interpretation begin by treating emojis (and the emoticons from which they are derived) as simple proxies for words. Given this, one must then consider what the proper dictionary-oriented analytical principles may be.[xi] I do so by proceeding along the four-step sequence for dictionary usage proposed by Jeffrey Kirchmeier and Samuel Thumma.[xii]

Selecting a Word: Each individual emoji in question ought to be understood as a single glyph whose meaning must be ascertained (though the meaning of an emoji may require more than one word to explain).

Selecting a Type of Dictionary: Only one suitable emoji dictionary exists.[xiii]

Selecting a Specific Dictionary: An appropriate reference point for this emojis-as-word-proxies analysis is the official Unicode dictionary,[xiv] since Unicode serves as the framing system within which new emojis are introduced to the public. The Unicode dictionary offers the most consistent and defensible starting point for transliteration, the first-order interpretive act that precedes the consideration of context clues.

Selecting the Appropriate Definition: The official Unicode dictionary only associates one descriptive phrase with each emoji; thus, one need not sift through multiple potential “official” definitions.[xv]

Crucially, my suggestion that analysis begin with the Unicode dictionary does not mean that the interpretive task is limited to a crudely substitutionary literalism.[xvi] Just as words may convey meaning through irony, sarcasm, simile, metaphor, and a host of other constructions, so too may emojis and emoticons.[xvii] Just like traditional forms of verbal communication, emojis can express metaphorical, or otherwise nonliteral, propositions. A gun emoji pointed toward a face emoji, for instance, may express a proposition similar to the colloquial, sarcastic expression of frustration, “kill me now.” Immediately construing this as an actual threat to a third party would likely be inappropriate. Consider another example: under the paradigm I have outlined, courts should avoid automatically interpreting an eggplant emoji (often used as a phallic symbol[xviii]) as indisputably such a sexually charged expression.[xix] In a case involving sexual harassment, canons of statutory interpretation[xx] suggest that the court’s interpretative work should occur in view of the rule of lenity.[xxi] This rule ought to be extended to the emoji context: sincere ambiguities in the meaning of a given emoji, in criminal cases, ought to be resolved in favor of the defendant. In other words, an eggplant may be simply an eggplant.

Additionally, the pronoun dilemma I have discussed above is not unique to the emoji/emoticon context, though in these cases it may appear to be a particularly problematic issue. I suggest that the problem here is similar to a speaker’s ambiguous use of the second-person pronoun “you” (which may refer to a specific listener, a group of specific listeners, or an undefined general audience).

Where linguistic ambiguities exist, courts have access to a range of interpretive canons that may be employed to parse context and deduce a coherent meaning. Where emojis are concerned, the same rules may be applied—although an initial transliteration step must first occur. Ultimately, assigning word values to emojis, by way of reference to the official Unicode dictionary, constitutes a rational first step in the task of interpretive construction: it brings emoji interpretation in line with other, preexisting interpretive norms.

Emoticons and Emojis: Meaningful Differences?

Emoticons, like emojis, may be said to have a cognizable referent, given that recent editions of the Oxford English Dictionary contain an appendix categorizing common emoticons.[xxii] I suggest that interpreters of ASCII emoticons parallel the interpretive procedures for emoji outlined above—which have their genesis in the use of a particular, identifiable dictionary.

The emoji/emoticon distinction may pose a unique difficulty for the interpreter. When one types “ :) ” into a particular application (such as Apple’s iMessage), the application may transform the typed emoticon into a single emoji character. This clearly has interpretive significance: if the meaning of the ASCII emoticon is not identical to that of the Unicode emoji, there may be a disjunction between the speaker’s actual intended communication and the communication expressed onscreen.

Such a problem, however, is not insurmountable within the interpretive paradigm I propose here. When an individual speaks or writes anything whatsoever, an interpreter has epistemic access only to the communication as expressed, not the sentiment as intended (though courtroom testimony may shed light on relevant mental states). The use of dictionaries in this context does not preclude investigation into the question of what was meant by the use of a given emoji: for the purposes of the court’s first pass at interpretation, an emoji in a given court record may be interpreted as an emoji, without a need to “reverse-engineer” the ASCII emoticon which may have, hypothetically, been initially typed.

Conclusion

In addressing symbolic reification in the context of emoticons and emojis, courts face an indisputably difficult task. Evolutions in communication technologies, and changing innuendos associated with given pictograms, render it increasingly difficult to decipher emoji-littered communications.

Here, I have sketched the initial contours of a methodology for interpreting such symbols: the interpreter begins by treating emoticons and emojis as reified word-groupings, to which traditional methods of interpretive construction may subsequently be applied. This paradigm allows for the use of context clues as interpretive tools, without sacrificing an initial grounding in objective meaning. The Unicode dictionary constitutes an ostensibly authoritative source of meaning, as well as a conceptual pathway for codifying semi-permanent linguistic referents. As court cases continue to hinge on the interpretation of online communications, deriving (and employing) such a source of meaning will become increasingly vital.

[i] See Adam Sternbergh, Smile, You’re Speaking Emoji, N.Y. Mag. (Nov. 16, 2014), http://nymag.com/daily/intelligencer/2014/11/emojis-rapid-evolution.html [http://perma.cc/Z24K-U98C] (“These seemingly infantile cartoons are instantly recognizable, which makes them understandable even across linguistic barriers. Yet the implications of emoji—their secret meanings—are constantly in flux.”).

[ii] See, e.g., Amanda Hess, Exhibit A: ;-), Slate (Oct. 26, 2015), http://www.slate.com/articles/technology/users/2015/10/emoticons_and_emojis_as_evidence_in_court.html[http://perma.cc/SCC4-YKLG] (“Every case that includes a digital record presents a high-stakes venue for arguing over what emoticons really mean.”).

[iii] See Emoticon Definition, Dictionary.com, http://dictionary.reference.com/browse/emoticon[http://perma.cc/79AZ-QUAF](last visited Oct. 26, 2015).

[iv] See Eli Hager, Is an Emoji Worth 1,000 Words?, The Marshall Project (Feb. 2, 2015), https://www.themarshallproject.org/2015/02/02/is-an-emoji-worth-1-000-words [https://perma.cc/Z9UB-L6PL].

[v] See Memorandum from R. Shirey, Internet Engineering Task Force (IETF), Internet Security Glossary, Version 2 (Aug. 2007), http://tools.ietf.org/search/rfc4949.

[vi] Some emojis do contain verbal content (e.g. the actual emoji is a box with the letters “OK” inside). These require no independent linguistic-construction analysis beyond that applied to the verbal content already contained within the emoji; the discussion here centers on how courts may apply principles of construction to nonverbal emojis.

[vii] See Shirley Li, What Do Emojis Mean?, The Atlantic (Feb. 4, 2015),  http://www.theatlantic.com/technology/archive/2015/02/what-do-emojis-mean/385127/  [http://perma.cc/MD98-N2V6] (“The question at the moment, therefore, isn’t how emojis are defined; it’s whether they’re vital to a case in the first place.”).

[viii] Petition for Writ of Certiorari at 8, Elonis v. United States, 135 S. Ct. 2001 (2015) (No. 13-983).

[ix] See Lauri Stevens, Using Emojis as Evidence, Law Officer (Apr. 7, 2015), http://www.lawofficer.com/articles/print/volume-11/issue-4/columns-departments/social-media/social-media.html[http://perma.cc/8K5K-82M8] (“In January 2015, prosecuting attorneys in the Silk Road case insisted that the inclusion of a smiley face emoticon by defendant Ross Ulbricht was significant.”); Debra Cassens Weiss, Emoticons Matter, Judge Rules in Silk Road Trial, ABA Journal (Jan. 30, 2015), http://www.abajournal.com/news/article/emoticons_matter_judge_rules_in_silk_road_trial/[http://perma.cc/JTH3-BKTC] (quoting U.S. District Judge Katherine Forrest as saying that “The jury should note the punctuation and emoticons”).

[x] See Julia Greenberg, That ;) You Type Can and Will Be Used Against You in a Court of Law, Wired (Feb. 12, 2015), http://www.wired.com/2015/02/emoji-in-court-cases/ [http://perma.cc/L3J3-BJSA](“Without the context of who is using the symbol, who received it, and an understanding of how those two people—or the people in their community—typically use it, the intent may not immediately be clear.”).

[xi] Cf. John Calhoun, Measuring the Fortress: Explaining Trends in Supreme Court and Circuit Court Dictionary Use, 124 Yale L.J. 484, 517 (2014) (“Different areas of law are more or less ripe for dictionary analysis.”).

[xii] See Jeffrey L. Kirchmeier & Samuel A. Thumma, Scaling the Lexicon Fortress: The United States Supreme Court’s Use of Dictionaries in the Twenty-First Century, 94 Marq. L. Rev. 77, 101-02 (2010).

[xiii] Full Emoji Data, Unicode.org, http://www.unicode.org/emoji/charts/full-emoji-list.html (last visited Oct. 27, 2015)

[xiv] Id.

[xv] Id.

[xvi] See, e.g., Cara Rose DeFabio, Violent Emoji Are Starting to Get People in Trouble With the Law, Fusion (Feb. 9, 2015), http://fusion.net/story/42623/emoji-violence/ [http://perma.cc/M6DQ-EATP](“Emoji lend emotional context to otherwise expressionless texts, but they also require context in order to be read. When you strip them of their context your only option is to interpret them literally. While nothing stunts the imagination more than taking an emoji at face value, there are times when the literal interpretation cannot be disregarded.”).

[xvii] See Tyler Schnoebelen, Humans Can Barely Understand Emojis. Will Machines Do Any Better?, Qualcomm (Sept. 18, 2015), https://www.qualcomm.com/news/spark/2015/09/18/humans-can-barely-understand-emojis-will-machines-do-any-better [https://perma.cc/M7GE-3ADC].

[xviii] See Kaleigh Rogers, The Eggplant Emoji Means Exactly What You Think It Means, Motherboard (Apr. 21, 2015), http://motherboard.vice.com/read/its-a-dick [http://perma.cc/B38Q-QCJT].

[xix] See also Russia “May Ban Gay Emojis” Under Propaganda Law, BBC News (July 31, 2015), http://www.bbc.com/news/world-europe-33738770 [http://perma.cc/6NS6-884K](observing that Russian authorities have begun to categorize emoji depicting two same-gender figures as “gay propaganda”).

[xx] For a fuller discussion of these canons and their interaction, see Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts (2012).

[xxi] Cf. United States v. Bass, 404 U.S. 336, 347 (1971).

[xxii] See E-Words, Emoticons Invade Latest Dictionary, USA Today (July 12, 2001), http://usatoday30.usatoday.com/tech/news/2001-07-11-ewords.htm [http://perma.cc/EKW6-DKNZ].