Lerdahl and Jackendoff’s Generative Theory of Tonal Music (1983) is an attempt to transform music theory into a theory of musical understanding by adopting the formal method and psychological premises of Generative Grammar, along with some Schenkerian elements. It has failed to fulfil its promise mainly because, like Schenker theory, it is compromised by what Leonard Meyer (1967) has called the ‘fallacy of hierarchic uniformity’. Considering the authors’ reluctance to compare music and language, it is paradoxical that its basic premise, the conception of a theory of music as a grammar, implies a stronger linguistic analogy than should be conceded.