The theoretic model of the “unreliable narrative” in fiction took flight in the early 1960s; it has since become a key concept in narratology, and an indispensable one. Simply put, first-person unreliable narrators are ones about whom we as readers, in collusion with the author, learn more than they know about themselves. Romantic precursors of modernist experiments in fiction—incipient cases of narrative unreliability—arise in the works of, among others, Jean Paul Richter and Heinrich Heine, two of Robert Schumann's favorite writers. In his early solo piano cycle, Papillons , op. 2, Schumann draws inspiration from Jean Paul's novel Flegeljahre , surely capturing something of the author's unreliably quirky literary style, in part through the strategy of tonal pairing . Whereas Schumann ultimately played down the programmatic elements of Papillons that trace back to the unpredictable Jean Paul, a genuine instance of the unreliable narrator is Heine's troubled poet-persona in Schumann's Dichterliebe . Here the composer invites us to perceive a second persona through the voice of the piano—one that understands the poet better than he does, and whose music reveals from the outset that rejection in love lies ahead. The emergence of narrative unreliability in fiction may have served as an influence that drove experimentation not only for Schumann but also for some of his contemporaries and successors. Debates about musical narrativity might profit from considering the recent literary concept of a “feedback loop,” in which the author, the narrator (text), and the narratee (reader)—in our case, the composer, the performer, and the listener (including analysts, performers, and composers, who are also intensive listeners)—continually and recursively interact.

Poetry and music have in common various ways of structuring sound. In both, one can speak of rhythm, meter, loudness (e.g., accent), and pitch. Beyond sound in the narrow sense, one can also speak of syntax in that both language and Western tonal music create expectations that are satisfied or not in a variety of ways. These material aspects of poetry and music can be the basis for exploring how poetry and music fit together in vocal music in general and in individual works. The study of poetry in these terms came to prominence in the work of Roman Jakobson and others beginning in 1960 and has more recently been taken up by Marjorie Perloff and colleagues. The study of music with words has in general not considered the materiality that they share, especially not in the analysis of individual works. Writers on music have generally placed emphasis on expression of the semantic content of texts instead, privileging texts that can be read in relation to Romantic lyric theory. This has led to the search for word painting of one kind or another that has shaped the understanding of whole periods in the history of music and that is very much with us still, though this semantic domain cannot ultimately be separated from the material aspect of language. This article analyzes Songs 1, 2, 4, and 6 of Schumann's Dichterliebe in terms of this materiality with a view to showing how closely congruent poetry and music can be in their own terms in individual works without dwelling first on what they might be thought to express. This is to speak not at first about meaning but rather about the means by which meaning is created. Analysis of this kind in no way precludes hermeneutics. It is but one starting point and one that bears directly on the act of listening and perhaps performance.