Even when emoji are used to substitute for words, we still understand the sentence. But how does that work? Do we interpret an emoji primarily as an image or as a word? To find out, a research team asked volunteers to read texts with emoji and measured the reading time precisely. It turns out that it takes a little longer to comprehend a sentence that includes emoji than one that doesn’t. If the emoji does not directly represent the intended meaning, but another word with the same pronunciation, we need yet a little more time. Based on these results, the team concludes that emoji comprehension consists of two steps: first the image is interpreted, then the word is inferred.