In this study of music and madness as literary topoi in writers from Denis Diderot to E. T. A. Hoffmann, John Hamilton has offered us one of the finest accounts of the late eighteenth-century idea of music as an ineffable and immediate art of tones, opposed to the referential and reflective language of words and thoughts. The account stands shoulder to shoulder with classic studies by Carl Dahlhaus (Die Idee der absoluten Musik (Kassel: Bärenreiter, 1978)) and John Neubauer (The Emancipation of Music from Language: Departure from Mimesis in Eighteenth-Century Aesthetics (New Haven: Yale University Press, 1986)).

Among Hamilton's original contributions to this history is, first, his subtle account of the complex fraternity of music and madness and their common opposition to language. Where language allows protagonists in literary works to constitute themselves as stable subjects, using memory, reflection and concepts to achieve identity with themselves over time, music and madness ‘unwork’ language, thought and identity, causing these individuals to confront their own instability as individuals and making boundaries between the fictional worlds (or works) that they inhabit and the historical worlds of their authors more porous. Yet, as Hamilton emphasizes, music and madness are not identical twins. They operate on language from above and below. Music allows individuals to touch the absolute or to identify with a sympathetic community, while madness corrodes processes of individuation with more ambiguous payoffs. By blurring the lines between music and madness and by allowing the two to permeate their texts in varied ways, the writers studied by Hamilton not only investigated the hazy differences and hidden commonalities between the metaphysical and the nonsensical, but often also managed to make their very opposition to language less clear.

Second, Hamilton places the late eighteenth- and early nineteenth-century obsession with music and language in a well-defined disciplinary context. As he notes, the ‘project is about music and madness in the strictest terms: not as tonal art and mental states as such but rather as specialized metaphorical strategies deployed in or constituting works of literature’ (9). In other words, as they invoked music and madness in their poetic, fictional, autobiographical, epistolary, journalistic and philosophical texts (to list the broad range of text types that appear in the book), writers were at least as much (if not more) concerned with their own discipline and art – language and writing – as they were interested in music or psychology per se. Although the concerns varied from writer to writer, Hamilton makes the striking general claim ‘that, above all, the nonsemantic, fascinating voice of romanticism's mad music is the voice of the author or rather the voice of the living person who is to become an author, who is about to ascribe his or her voice to a system that will work it into sense and thereby work it off’ (12). In many of the texts under consideration the writers created protagonists with unmistakable autobiographical traits (such as Diderot's Moi in Le neveu de Rameau, Wackenroder's Berglinger and Hoffmann's Kreisler), yet it is precisely because these texts eschew straight autobiography, because they aim rather at fictional alternative worlds (that is, works) but use invocations of music and madness to fracture the walls of the work, that they are able to roll back the dead letter of literature: ‘as metaphors of nonrepresentability, music and madness could introduce into a text the nonrepresentability of the self’ (14). As a result, texts may lose some conceptual precision and internal self-sufficiency – that is, emphatic work character – but they contain a little more of their authors.

The book consists of an Introduction that clearly identifies the subject matter, its inherent challenges and the central claims of the book; six chapters, chronologically organized, each of which consists primarily in close readings of a few selected authors but envelopes them in a broad historical and philosophical context; and, framing the whole, two short texts ‘Hors d'œuvre’ that both prepare and question the monograph's stated focus on literary strategies. These consider two writers interested in music who did actually go mad – Hölderlin and Nietzsche. It is a virtue of the study that Hamilton is aware of the methodological problems involved in writing a scholarly book about things resistant to sense (see pages 8–11), but he does not give in to the possible seductions of the topic. He writes both sensibly and with great aplomb. Striking metaphors illuminate his interpretations and never degenerate into pure plays on words. The book is a pleasure to read. Both author and Columbia University Press also deserve a tip of the hat for careful proofreading and beautiful production.

Denis Diderot's ‘Le neveu de Rameau’ is the focus of the first two chapters (‘Hearing Voices’ and ‘Unequal Song’). In the first, Hamilton analyses the ‘double articulation’ implicit in the narrator's account of his conversation with the nephew. While the nephew (Lui) articulates the immediacy of his self through disjointed thoughts and musical outbursts, the narrator (Moi) attempts to make articulate sense of Lui's utterances (his ‘voice’). In the second, the distinctions of the first chapter are used to delineate two approaches to time and consciousness, one in which a person becomes a self by achieving constancy over time through conceptual reflection, and another in which one remains ‘unequal’ to oneself as time reels through the body. The third chapter (‘Resounding Sense’) focuses on Hegel's famous reading of ‘Le neveu de Rameau’ in his Phänomenologie des Geistes and on Rousseau's Rêveries d'un promeneur solitaire to show the varied ways that writers criticized the Enlightenment culture of ready-made reason by finding sense in sound and irrationality. The fourth chapter (‘The Most Violent of the Arts’) deals with the sublimity of sound and music in a variety of writers, from Longinus and Burke through to Kant, Herder and Wackenroder. The fifth and sixth chapters (‘With Arts Unknown Before’ and ‘Before and After Language’) study Kleist and E. T. A. Hoffmann respectively. Both writers felt the twin tugs of music and madness in their own personal lives and turned them into central literary topoi. But whereas Kleist never managed to resolve their assaults on his own identity and eventually committed suicide, Hoffmann found ways to question rationality and fixed forms without being ‘unworked’ himself.

These brief summaries do little justice to Hamilton's subtlety of argument or to his wealth of insight into slight details, for his strength lies in imaginative close readings. An example: Hamilton uses two small details of ‘Le neveu de Rameau’ – the narrator designates the nephew a ‘Silenus’ and Diderot called the dialogue a ‘satyre’ (he does not use the etymologically distinct term ‘satire’) – to relate the text to Greek origins: a foundational myth of music (Apollo and Marsyas) and Plato's Apology and Symposium (in which an intoxicated Alcibiades compares Socrates to both Silenus and Marsyas). (Greek texts make frequent cameo appearances in the book.) Apollo and Marsyas represent an opposition between two species of madness that represent the upper and lower bounds of embodied thought – the madness of Apollonian reason that subjugates all experience to timeless concepts and beautiful forms (represented by the Pythagorean tuning of the lyre), and the madness of Marsyan embodiment that never rises above immediate experience (represented by the breathy aulos). As a Marsyas and Silenus, Socrates uses language not to define truths but rather to question, to involve his disputants in a process. Unlike Plato, Socrates mistrusts writing. ‘He unworks each interlocutor's beautiful words’ and his quest ‘is to interrogate what language says by analyzing how language works. His irony shows little respect for all truths formed by language’ (44). When Hamilton returns to Diderot from the excursion into Greek literature, he uses these insights to illuminate small details of Diderot's text (the nephew's line ‘Ah, ah, vous voilà, M. le philosphe’ recalls the howls of Marsyas in Ovid's telling of the myth ‘A! Piget! A! non est … tibia tanti’, both representing the non-semantic cry of uncomprehending immediacy) and emphasizing how language need not aspire to conceptual constancy, both in Diderot's dialogue and in general.

One of the striking aspects of Hamilton's method is that he is sensitive to the many ways that authors themselves strained against the reification of their texts, though he emphasizes close textual analysis of literary ‘works’: ‘the line between life and work that critical interpretation invariably must draw can never be rigorous’ (206). Under the pressure of close reading, Hamilton reveals literary works both as polysemous webs of textual tropes and topoi, each with a broad and long European history, and as specific responses to personal, literary and philosophical problems, each with a specific historical context. Hamilton's inventive close readings are grounded in succinct summaries of philosophical arguments, historical contexts and biography.

Occasionally, the double approach – respectfully contextual and inventively interpretive – shows strains. For example, in the chapter on sublimity, Hamilton first summarizes the standard account of Kant's approach to sublimity: when the imagination finds its limits, a person feels overwhelmed, yet the experience of blockage reveals the existence of a rationality and moral code otherwise unavailable to empirical experience. Kant argues that it is the rational and moral subject who is sublime. Music itself does not enter into Kant's thoughts on sublimity. Hamilton then draws the reader's attention to well-known disparaging remarks about the art: one cannot shut one's ears to music, which in any case may only be agreeable rather than beautiful. As Hamilton reads it, Kant's own inability to shut his ears – to keep music from disturbing his philosophical meditations – is read as an implicit bow to the sublime violence of music. The reading is broadly persuasive with respect to music, for writers both before and after Kant often did bring up music in their discussions of sublimity. However, Kant also noted that strong odours were similarly difficult to withstand. By Hamilton's logic, stench should also be sublime, yet one would be hard pressed to find an eighteenth-century tradition to support this conclusion.

Hamilton offers an implicit rationale for such readings that transcend the conscious intentions of authors. In a discussion of Hegel's productive misreading of Diderot's ‘Le neveu de Rameau’ in the Phänomenologie des Geistes – as Hamilton notes, Hegel's three quotations reconstrue the dichotomies that Diderot built into his dialogue, for example making what was natural for Diderot a product of the fractures of culture – the author appreciatively describes Hegel's method as ‘eminently philosophical’ (92). ‘Like an echo, he cuts and splices snatches of text, which thereby transcend and rework the original. … In this way, the idealist philosopher hopes to correct the insufficiencies of a purely rationalist standpoint’ (94). In the best instances, Hamilton does transcend straight literary and philosophical history to lay bare significances smouldering in the text, and, unlike Hegel, he is too respectful and careful a philologist to dispossess his authors entirely of their texts.

Given Hamilton's careful circumscription of his topic – he is interested in music and madness as literary topoi and metaphors, not as art forms – the fixed positions of music and madness as markers of immediacy, embodiment and all things non-semantic makes sense. For eighteenth- and early nineteenth-century writers struggling with the ideals of the Enlightenment, such was their undoubted attraction. Yet at times Hamilton seems to move from a historically situated investigation of a literary topos towards general statements about the essence of music (to leave madness out of consideration). ‘Taking place in an irreversible flow, one could say that there is no true repetition in music, at least no repetition in the strict sense. Although all aesthetic experience admittedly participates in some kind of temporal unfolding … music's time is perfectly inherent’ (31–32); ‘music is the art of time par excellence, an art of process, change, and alternation’ (61). For musicologists who, as they have investigated issues of musical narrative and meaning in the last thirty years, have either rejected or nuanced accounts of music's ineffability, such statements will seem questionable. While Hamilton is willing to accompany the writers he studies as they critique eighteenth-century prejudices about the inherent rationality of language and its relationship to stable subjectivities, he is less interested in the contemporaneous attempts of musicians and critics to discover the possibilities of making music signify, even without words, or to arrest its dissipation in time by making it into a work. Were literary and philosophical writers of the time so unaware of such explorations? Or if they were aware of them (as E. T. A. Hoffmann surely was), was their commitment to literary issues such that they could bar such explorations from their sight? As a study of a literary obsession, Hamilton's book will remain a key text for those interested in the genesis of the idea of ineffable music, but work remains to be done on the period's madness for music.