Authors: Elsi Kaiser and Patrick Georg Grosz
Contact Email(s): emkaiser@usc.edu, p.g.grosz@iln.uio.no
4 thoughts on “Anaphoric dependencies in the digital age: On the relation between emoji and text”
Leave a Comment or Question Below
You must be logged in to post a comment.
Hi Elsi, great talk! My comment/question wasn’t selected during the talk, but in response to the last comment, I was wondering if you considered whether the action emoji added any information to the text message or not. It seems to me that in the example of the cake-emoji, the piece of cake emoji echoes known information (dessert-cake), but in the example of the drum emoji, it seems to add new information to the sentence (annoyed-drum). In the medal-emoji example, it seems more between those two in terms of new information. On the other hand, the facial emoji in all three conditions, adds similar amounts of information (sender attitude or a birds eye perspective commentary on the situation)
Hi Valeria, This is a really intriguing point! We did not directly control for this. I’m right now brainstorming different ways to measure this for the items 🙂 thanks for the suggestion to look into this. In the face condition, there is probably also variation in how informative/(un)expected the face emoji is (given the range of face emoji and the fact that we avoided repetition) — or in other words, how predictable the sender’s (or whoever’s) attitude is, given the preceding linguistic text. I think this also relates to one of the question about how emoji are used in the ‘wild’ (something that we are also looking into with corpus work). That being said, I think one of the motivators for using face emoji is the lack of facial expressions, prosody etc in texting, so I fully agree with your point that face emoji often strive to convey information that language users otherwise feel to be missing when communicating via text.
That was a fun and interesting talk, Elsi! A quick thought about the transfer results and aspect manipulation. In the original 2006 experiment with Hannah and Jeff, we used ToP verbs because it’s plausible that the two animate event participants have varying levels of prominence within different components of event structure — the agent/source is more prominent at the initial state and the ongoing development of the event, and the goal is more prominent at the end state, by virtue of having just received the object of transfer. That’s why personal pronouns have a stronger agent/source bias in the imperfective , which focuses on the ongoing development of the event, than in the perfective, which focuses on the end state.
So I wonder if action emoji — by virtue of characterizing the action itself, or perhaps the object of transfer — are naturally taken to comment on the whole event rather than the end state, thereby bringing the ongoing development of the event into focus, regardless of the aspect used. If so, it may not be so surprising that you get a subject bias that is akin to the one we saw for imperfectives across the board. Just a thought, and hope that makes sense.
Hi Andy,
Thanks for your comment. Yes, I think our thinking is very much in line with what you suggest! In our transfer items, the action emoji depicted the transferred object — ie neither the source nor the goal character — so that’s a key difference relative to your earlier work. So it’s not an apples-to-apples comparison. That being said, we were curious whether participants *might* associate the transferred object more with the source vs goal depending on aspect (since the object is arguably more likely to be in the possession of the goal in the perfective condition compared to the imperfective). Nevertheless, I really like your point that the action emoji could well be taken as commenting on the event itself (which would yield the same pattern as you found for imperfectives). Part of this may also stem from the task/how we asked the question, so that’s something we should think more about.