An analogy can be formed between sentence reading (language) and arithmetic expression processing (math), which might suggest shared underlying cognitive processes. Sentences are made up of subparts (words) that are systemically combined to convey a potentially coherent message and, similarly, arithmetic equations are made up of subparts (numbers) with combinatorial symbols (+, -, ×, ÷) that can be sensibly completed (e.g., 4×5=20) or not (4×5=15). At the same time, there are important differences in the subjective goals people have in mind when encountering them, namely that sentences are read for meaning whereas arithmetic is processed for seeking a correct solution. Across a series of studies, Danielle Dickson will provide evidence from ERPs that long-term experience with arithmetic problems influences the way adults read them. Whereas children appear to process symbolic arithmetic as if they are reading a sentence with words, building meaning-level expectations, adults shift away from reading arithmetic for meaning and instead interpret arithmetic as a simple categorization task. It seems that when it comes to arithmetic, the brains of children and adults disagree about whether it is a kind of language.