Musical semantics: Dimensions, processes, and neural correlates 99 ated for words that are preceded by a semantically congruous context, compared to when preceded by a semantically incongruous context (Kellenbach et al., 2000). That is, when a word is preceded by a semantic context, the amplitude of the N400 is in-versely related to the degree of semantic t between the word and its preceding se-mantic context (for an example see upper panel of Figure 1).There is ample evidence that the N400 elicited by words is sensitive to concep-tual analysis, and it is assumed that the N400 refects that readers and listeners im-mediately relate the word to a semantic representation of the preceding contextual information. That is, the N400 elicited by words is particularly sensitive to the pro-cessing of meaning information, both in prime-target and in sentential contexts (Kutas & Federmeier, 2000; Friederici, 1999). More generally, an N400 effect can be elicited by the processing of almost any type of semantically meaningful informa-tion, such as faces and pictures (e. g. Lau et al., 2008), environmental sounds (Van Petten & Rheinfelder, 1995; Cummings et al., 2006. Orgs et al., 2006, 2007; Aramaki et al. 2010), and odours (Grigor et al., 1999); importantly, the N400 can also be eli-cited by music, and the N400 elicited by a word can be modulated by the meaning of musical information preceding that word.An initial study investigating this issue (Koelsch et al., 2004) used a classical se-mantic priming paradigm in which target words (presented visually) were preceded by (auditorily presented) prime stimuli that were either sentences or musical ex-cerpts (Figure 1). The prime stimuli were semantically either related or unrelated to the target word. In the language condition, for example, participants heard the sen-tence: “The gaze wandered into the distance” and then saw the target word “wide-ness”, which is semantically more closely related to the prime sentence than to a prime sentence like “The manacles allow only little movement” (Figure 1). Target words (n = 44) were both concrete (e. g., needle, river, staircase, blossom, king, bird, pearls, sun) and abstract words (e. g., wideness, limitedness, devotion, mischief, reality, illusion, arrival, leave).In the ERP experiment, participants made a two-alternative relatedness judge-ment for each target word. In the music condition, participants judged 78% of the related target words as “related”, and 82% of the unrelated target words as “unre-lated”; hit rates in the language condition were higher (93 and 91%, the difference between language and music condition was signi cant). The upper panel of Figure 1 shows the ERPs of the target words following prime sentences to which the target words were semantically either related or unrelated. ERPs elicited by the target words showed the classical N400 priming effect: The N400 was clearly larger for the semantically unrelated compared to the semantically related target words (as to be expected). Importantly, the conditions of primary interest in that study were the conditions in which the target words were preceded by musical excerpts. These ex-cerpts were believed to be meaningfully (or “semantically”) un/related to the target words, and it was of interest whether a musical excerpt could have the same effect on the semantic processing of a target word as a sentence. In the example presented in the lower panel of Figure 1, the target word “wideness” was in one trial preceded