Abstract
It has now been over fifteen years since Gilles Fauconnier and Mark Turner (2002) introduced Conceptual Blending Theory (CBT). In what Fauconnier and Turner describe as “blending”, elements from diverse, but structurally related, mental spaces become conceptually enriched through their integration or give rise to altogether new concepts.
The theory has found considerable application in music analysis, particularly in attempts to unpack complex music-word or music-image “blends” in song, opera, music video, and other multimodal genres (Zbikowski, 2002, 2008; Cook, 2001; Bauer, 2004; Moore, 2012). Other applications of CBT have centered on the exploration of intra-musical conceptual integration (e.g. Spitzer, 2004; Antović, 2011) and addressing challenges to the theory’s applicability, most notably due to its breadth of scope (Bache, 2005; Ox, 2014). Fauconnier and Turner’s theory provides useful accounts of why certain existent concepts can be considered as the results of blending (Turner, 2006, 2014). However, it does not specify in detail how one could go about formally producing new conceptual blends (Schorlemmer et al., 2014; Stefanou & Cambouropoulos, 2015). Ultimately, the theory’s open-endedness raises a series of broader, related questions: How do we define creative blends in music; where do we look for them, and how?
The idea for this collection first took shape in October 2015 during a workshop on Creative Conceptual Blending in Music as part of the Concept Invention Theory (COINVENT) project. 1 The workshop helped reveal several questions of shared interest. What constitutes a concept (and, by extension, a blended concept) in music? How can novel music-based blends be created and evaluated? How does structural blending relate to musical meaning? How is temporality compressed and unpacked in music-based blends?
Beginning as reflections and elaborations on such questions, the eight articles that comprise this collection end up fostering broader, more pertinent questions regarding musical creativity and conceptualization. Conceptual blending is critically revisited here from a variety of perspectives, including computational creativity, cognitive musicology, historical musicology, music analysis and semiotics. On more than one occasion, discussion and feedback between authors from different disciplines enabled individual contributions to ultimately take shape as collaborative projects. The majority of the research presented here is directly or more loosely associated with the COINVENT project; the articles by Lawrence Zbikowski and Michael Spitzer were specifically prepared for this issue.
Lawrence Zbikowski’s contribution offers an accessible summary of CBT and its applications in music, and goes on to identify two of the most pervasive challenges presented by the theory: how creativity is defined, and how musical utterances are conceptualized. Zbikowski employs examples from Purcell’s Dido and Aeneas, re-imagined in a variety of performance scenarios, to argue for significant qualitative differences between musicological and linguistic perspectives on conceptualization.
Michael Spitzer notes the distinct modalities in operation between music and language blends in the domain of musical emotion. Focusing particularly on the case of anger in examples by Vivaldi and Haydn, Spitzer addresses music as a kind of “metaphorical body” where emotions are expressed in varying stages of compression and unpacked in real time.
Costas Tsougras and Danae Stefanou deal with hermeneutics of structure through a detailed look at two piano pieces from Mussorgsky’s Pictures at an Exhibition. Conceptual integration networks are employed to examine not only cross-domain but also “intra-musical” blending. The study reveals complex, multi-scope embedded blends in Musorgsky’s composition where contrastive temporal, ontological and experiential states become creatively integrated.
The “intra-” vs. “extra-musical” division is challenged in Mihailo Antović’s contribution, which also problematizes the universalism vs. relativism divide in linguistics. Antović discusses data from several sets of experiments with children and adults to review the notions of the “generic space” and the “grounding box” in CBT and their applicability for musical conceptualization, particularly in examples of blended image-schemata, like the idea of musical motion.
The idea of the grounding box is further explored in George Athanasopoulos’ article, which has been co-authored with Antović. Here the authors are concerned with how abstract musical concepts become associated with particular visual shapes, discussing findings from a cross-cultural empirical study on musical visualization. This leads them to the formulation of a model of “perceptual modalities” which highlights the implication of physiological, perceptual and sociocultural factors in the visual representation of a musical stimulus.
Petros Vouvaris and Dimitris Tasoudis examine temporality and identity compressions and decompressions in Virgil Widrich’s short film Copy Shop (2001). Their study combines data from multimedia analysis, published interviews and direct communication with the film’s creators to argue for a more dialectical division between how blends are created and analyzed. By examining the shifting interrelations between mental spaces at different time points in the film, the authors highlight music’s contribution to the creative narrativization of the blending process.
Similarly, Danae Stefanou’s contribution calls for greater emphasis on the process of conceptual disintegration and highlights the importance of performance in musical conceptualization. Stefanou discusses open-form compositions and verbal/graphic scores by Earle Brown, Pauline Oliveros and Jennifer Walshe to suggest that blending theory could benefit from application to different kinds of repertoire.
In the concluding article, Asteris Zacharakis, Maximos Kaliakatsos-Papakostas, Costas Tsougras and Emilios Cambouropoulos present CHAMELEON, a newly developed automatic harmonization assistant which creates novel blends both within and across different harmonic idioms. CHAMELEON is based on the COINVENT formal model of computational creative blending and illustrates practically the potential of CBT in a specific domain of musical creativity. The article attempts to evaluate the output of the system by means of an experimental study with human subjects who were asked to classify and evaluate the automatically generated blends in terms of idiom, preference and expectancy.
The various perspectives presented in this special issue suggest that if we are to move beyond a post-hoc explication of existent musical structures and onto the invention of new blends, more emphasis should be placed on studying the processes of musical conceptualization and creativity in specific contexts (either re-imagined in existing music, re-cast in creative performance situations or examined via cognitive/computational modeling in specific domains). In addition to explicating the outcomes of musical creativity, it is useful to look more closely into the creative processes that inform musical understanding, performance and listening. This dual (or multiple) focus may shed light on the kinds of constraints, mapping strategies, grounding principles, saliences and optimality criteria that allow the alignment and combination of diverse intra- or cross-domain concepts into original musically “meaningful” blends.