Music, Meaning and the Embodied Mind Towards an Enactive Approach to Music Cognition By Dylan van der Schyff Submitted in partial fulfilment of the MA in Psychology for Musicians, Department of Music, University of Sheffield July 2013
CONTENTS Table of Contents...2 Abstract... 3 Introduction...4 Outline and method...7 One A Dual Orthodoxy: Understanding the Cognitivist-Adapationist Program...10 Cognition as computation...11 Information processing, modularity and the adapted mind...14 Some preliminary concerns...17 Two Music and Meaning: From Information-Processing and the Music as Language Metaphor to Embodied Action and Bio-Cognitive Ecology...26 Music and the computational mind...28 Modularity and the biological origins of music...31 Moving beyond biased, reductive and reified notions of music...33 Human development and the bio-cultural meaning of music...36 The musical brain: a dynamic and interactive perspective...38 Movement and the corporeal origins of musical meaning...42 Three An Enactive Approach to the Musical Mind...50 Cognitive ecology and the idea of affordances...51 Outlining the enactive approach...54 Conclusion: music, consciousness and the experience of self...62 References...68-2 -
Abstract The study of music cognition has been dominated by a largely disembodied conception of the mind. This so-called cognitivist perspective treats mental activity in terms of abstract information-processing where the world is represented in the mind via the computation of sub-personal symbols; and where the mind-brain relationship is explained in terms of a collection of cognitive modules shaped by natural selection. Recent decades have seen an ecologicalembodied paradigm emerge in cognitive science, as well as more plastic and interactive conceptions of the mind-brain and organism-environment relationships. These new perspectives offer a much broader understanding of meaning-making and the mind and are becoming increasingly influential in music cognition studies. The orthodox approach to the mind and its origins is examined; and its influence on music cognition research is discussed. Alternative embodied, developmental, ecological and bio-cultural perspectives on cognition and the musical mind are considered. The enactive approach to embodied cognition is then offered as a theoretical framework that better accommodates these broader and more nuanced ways of understanding musical meaning. To conclude, the relevance of the enactive approach is considered for music education, performance and practice. - 3 -
Introduction The profound influence of the information-processing approach to cognition has tended to promote a disembodied view of musical experience. This perspective often treats musical cognition as if it were an abstract reasoning or problem solving process that proceeds in a hierarchical fashion (Clarke, 2005: 15). It relies heavily on a computational model of the mind-brain relationship where cognition is understood in terms of symbol processing at sub-personal or non-conscious levels : the outputs of lower levels of mental processing feed inputs to higher levels, with increasingly complex representations of a world out there produced at each stage. This conception of cognition is supported by an objectivist analytical philosophical tradition (e.g. Ayer, 1936; Stevenson, 1944) that understands meaning formation largely in terms of iterative (linguistic-logical) processes i.e. the formation of representations, propositions and concepts (Johnson, 2007). By this view, the cognitive function of emotions, feelings, as well as embodied perceptions and activities are largely ignored. This so-called orthodox cognitivist conception of mind (see Dennett, 1978; Hofstadter & Dennett, 1981) is often paired with a modular understanding of the mind-brain relationship grounded in a strict adaptationist approach to biological evolution (Fodor, 1983; Pinker, 2009). As a result, the complexity of human thought and behaviour is often discussed in terms of the evolution of a large array of cognitive modules, each adapted by natural selection to process a specific type of information in ways that contribute to the survival of the individual and its genes (Tooby & Cosmides, 1989, 1992; Pinker, 1997). Because environmental factors (culture, experience and so on) are thought to exert a negligible influence on the genome, this approach seeks firm distinctions between the products of nature (i.e. natural selection) and those of culture in the human phenotype. - 4 -
The cognitivist-adaptationist program has underpinned much theory and interpretation in the study of human cognition, often to the exclusion of other viable approaches. As I will discuss below, this has resulted in problematic reductions and reifications, dichotomous conceptions of nature and culture (i.e. the nature vs. nurture debate), as well as disassociated notions of mind, body and world. Furthermore, the highly objectivising view this approach demands has contributed to a problematic epistemological divide between scientific inquiry and phenomenal experience that has marginalized the study of consciousness in Western scientific and intellectual circles. However, while this orthodox approach remains highly influential in Anglo- American philosophy and psychology, developments in cognitive science and philosophy of mind have begun to open up alternative possibilities. Increasingly, cognition is understood to be grounded, first and foremost, in interactive and embodied experience with the physical and socio-cultural environment (Changizi, 2011; Tomasello, 1999; 2005) where cross-modal perceptions, sensory-motor activity, emotions and basic metabolic processes play a central role (Damasio, 1994, 1999, 2003; Gallagher, 2005; Johnson, 2007; Lakoff & Johnson, 1999, 2003; Ramachadran, 2011). Related research has questioned the dominance of the modular conception of mind, especially as it is understood within the strict adaptationist context of evolutionary psychology (Besson & Schön, 2012; Uttal, 2001; Panskepp, 2009; Lickliter & Honeycutt, 2003). There is growing neurobiological evidence that complex human behaviours, such as those involved with social cognition, depend more on organism environment interactions than on innate dispositions; the plasticity of the human neocortex and the formation of selforganizing cognitive structures through experience and epigenetic effects are increasingly emphasized over a large suite of genetically determined modules (Gregory, 1987; Maturana & Varela, 1987; Lewontin, 1983; Pansepp, 2009; Sur & Learney, 2001). In brief, the cognitivist and adaptationist orthodoxies are being - 5 -
challenged by more nuanced, integrated and interactive understandings of the nature and origin of human cognition. Because the functions and meanings of music span such a wide range of human experience this work clearly holds great promise for the study of musical cognition, which is increasingly understood as a bio-cognitive and bio-cultural phenomenon. A number of scholars have begun to consider the enmeshed biological, cognitive and cultural origins of music in human evolution and ontogenesis (e.g. Cross, 1999; Tobert, 2001; Trevarthen, 1998, 1999). While others, such as Leman (2008) have employed embodied theories of music cognition to discuss possibilities for more effective mediation technologies. Crossmodal approaches to music perception and cognition have been investigated by Eitan (e.g. Eitan & Granot, 2006; Eitan & Timmers, 2010) and Johnson (2007) among others. And Gibson s (1966) ecological approach to perception an early alternative to representational cognitivist models continues to be influential; it has been developed in a musical context by Clarke (2005) and others (e.g. Kreuger, 2011a, 2011b; Windsor, 2000, 2004) in order to discuss musical cognition in terms of specification and affordance (as opposed to cognitivist principles of codification and representation). Indeed, recent research has prompted some scholars to reexamine a number of earlier theories that fell by the wayside during the period of preoccupation with computational models of mind that arose in the 1950 s and 60 s. The work of James and Dewey, Merleau-Ponty, as well as the early pioneers of cybernetics, is beginning to reassert an influence among a new generation of psychological musicologists (e.g. Reybrouk, 2005) and musically inclined philosophers (e.g. Johnson, 2007). With this in mind, this paper aims to better understand the so-called enactive approach to cognition within the context of human musicality. This approach is most clearly articulated in The Embodied Mind: Cognitive Science and Human Experience (1993), a joint effort by cognitive scientist Franscisco Varela, - 6 -
philosopher Evan Thompson and psychologist Eleanor Rosch. It has also found some more recent advocates, such as philosopher Mark Johnson (2007), among others. I suggest that the enactive approach may provide a valuable addition to the study of musical experience. It presents a compelling critique of the dual orthodoxies of adaptationism and cognitivism, potentially allowing for deeper, biocognitive and bio-cultural approaches to questions of music and meaning. It also offers a critical addition to Gibson s ecological theory of perception and by extension the work of Clarke (2005) in music by placing a greater emphasis on cognition as perceptually guided action that both drives and creates the historical context of structural coupling between organism and environment (i.e. organismenvironment codetermination). Perhaps most interesting is the potential this approach holds for dealing with the seemingly irreconcilable gap between scientific aspirations for objectivity and the reality of direct personal experience. This last concern seems especially relevant in the context of music psychology, which seeks to understand the experience of music, in all its diversity, largely through scientific methods. As I will discuss further, the enactive approach offers useful tools for the analysis of conscious experience most notably the mindful-awareness techniques associated with the empirical/pragmatic strands of Buddhist philosophy (Murti, 1955; Varela et al., 1993; Kalupahana, 1987; Lowe, 2011; Biswas, 2011). It is suggested that this may enable a new level of dialogue between scientific inquiry and subjective experience, as well as a systematic (and radically empirical) means by which performers, teachers, students, and listeners may analyze and gain a deeper appreciation of their musical experiences. Outline and method I am attempting to draw out a broader and more inclusive conception of musical meaning than is often entertained in mainstream music cognition studies. As a result, I will necessarily be bringing together a broad range of ideas and concerns, - 7 -
some of which may at first appear to be related only tangentially. I will do my best to contextualize things as I go. The following overview should give the reader some idea of where the discussion is headed as well as the key areas to be considered. Section One offers an overview of the dominant cognitivist-adaptationist understanding of the human mind and its origins. I discuss the history and influence of both approaches and consider how they reinforce one another. I conclude by discussing some concerns that will be developed in later sections in musical contexts i.e. the problematic dichotomies and reductions mentioned above, as well as difficult issues concerning consciousness and phenomenal experience. Because the cognitivist-adaptationist approach underpins much research in music cognition, but is rarely discussed critically in a musical context, it will be necessary to articulate its core principles as clearly as possible. Therefore, for the sake of clarity, musical concerns will not be addressed here, but rather will be taken up in the following section. Section Two begins with a brief look at Eric Clarke s (2005) illuminating critique of the dominant cognitivist view of music cognition. Clarke s work is worth considering not only because he articulates and critiques the underlying assumptions of the standard approach so clearly, but also because he is one of the few established figures in systematic musicology to do so. Following Clarke s lead, I attempt to demonstrate how the deep influence of the cognitivist-adaptationist program has contributed to a standard view whereby music cognition is understood to proceed in accordance with the computational model of mind outlined in Section One i.e. abstract symbolic operations carried out by an adapted modular brain. As I will discuss, because such operations are thought to function syntactically, research is generally grounded in the assumption that musical cognition is best understood in terms of rule-based and representational processes. This, I argue, has led to a preoccupation with music s relationship to language a valuable area of study, but one that, in the absence of a discussion of other factors - 8 -
involved, provides only an incomplete understanding of musical experience. As I go, I draw on studies of human ontogenesis; more dynamic and interactive conceptions of the mind-brain and organism-environment relationships; as well as bio-cultural, action-based, and cross-modal approaches to human cognition in order to develop a deeper, embodied understanding of the nature and origin of the musical mind. In Section Three I consider this broader conception of musicality in the context of the enactive approach to cognition. In order to distinguish the enactive view from similar theories I begin with a look at Gibson s ecological (1966) approach to cognition. Here I return briefly to the thought of Clarke (2005) and consider his approach to musical meaning (which is largely based on Gibson s theory). At this point I take a somewhat more critical view of Clarke s work and discuss some problematic issues with the Gibsonian perspective in general. I then introduce the enactive approach as a viable alternative to both the cognitivistadaptationist and the Gibsonian points of view one that better encompasses the wide range of concerns and ideas developed in Section Two. I explain key terms and concepts; and I consider how the enactive approach offers a welcome theoretical framework through which we may better understand musical cognition as an evolving, embodied and bio-cultural phenomenon. To conclude, I discuss the enactive approach to consciousness and consider how the insights it offers may afford a means by which the individual may systematically examine his or her own experience, and thereby gain a more nuanced understanding of their direct relationship to the musical activities they engage in. This, I argue, may have important implications for music education, practice and performance, as well as for the scientific study of musical experience. - 9 -
One A Dual Orthodoxy Understanding the Cognitivist-Adaptationist Program Until recently, the understanding of human cognitive capacities and potentials has been dominated by two mutually reinforcing paradigms of thought. Respectively, the so-called cognitivist and adaptationist approaches attempt to explain how mental operations occur, as well as the evolutionary mechanism that produces the brains capable of carrying out such processes. Whether implicitly or explicitly, the theoretical framework imposed by this understanding of mind has underpinned much of the research, interpretation, and theory related to the study of musical cognition (see Clarke, 2005). The study of how and why musical experiences are meaningful is of central importance to systematic musicology; and it draws on an increasingly wide range of disciplines including sociology (Denora, 2000; 2011), the cognitive and biological sciences (Rebuschat et al., 2012; Wallin et al., 2000), education and cultural studies (Small, 1999), archaeology (Mithen, 2005), and the humanities (Johnson, 2007). Indeed, it is becoming more and more evident that musical cognition cannot be properly understood apart from the emotional-physiological responses to music (Juslin & Sloboda, 2010), the role of musical behaviour in human development (Trevarthern, 1998; Trehub, 2003), and the difficult biocultural question of why humans should have universally evolved into musical beings in the first place (Cross, 1999, 2001, 2010, 2012; Fitch, 2006; Patel, 2008, 2010; Tolbert, 2001; Pinker, 2009). But opinions and theories vary greatly and the debate over the hows and whys of musical meaning is far from settled. In fact, the - 10 -
growing interest in music cognition has revealed special problems and opportunities for the study of human cognition in general: because musicality takes so many forms and spans such a wide range of human experience it forces the question of how deep and varied the bio-cultural origins of cognition and meaningmaking really are. Nevertheless, the dominant cognitivist-adaptationist approach continues to focus our investigations and understandings of musical meaning largely towards notions of 'internal' mental representations that are assumed to correspond with pre-given musical structures in the 'external' aesthetic environment. In what remains of this section, I outline the central tenets of the cognitivist-adaptationist approach and consider some implications and problems associated with it. In subsequent sections I will examine the influence of this approach on our understanding of music cognition and discuss alternatives. I should state from the outset that it is not my intention to simply refute the cognitivist and adaptationist programs; they are both based on established research strategies and are as plausible as any other scientific endeavour (Varela et al., 1993). However, I suggest that the often-uncritical acceptance of these approaches as the only viable models reinforces problematic nature/culture and mind/body dualisms, resulting in reductive, reified, and disembodied understandings of the nature and origins of the musical mind (see also, Still & Costall, 1991; Potter, 2000). Cognition as computation Cognitive science as we know it today can be traced back to the cybernetics movement that emerged in the early 1940 s. It was during this period that researchers first introduced the idea that mental processes could be understood in terms of computations carried out by machines (Gardener, 1985; McCulloch, 1965). Such machines (a brain or computer) would consist of many simple threshold devices i.e. neurons, silicon chips, or tubes that function in a binary - 11 -
on/off or active/inactive capacity connected so as to be able to perform logical operations (Heims, 1980; for a thorough account see Pinker, 2009; and Betchel et al., 1998). While the theoretical and practical implications of computational cognition were recognized immediately, in the early days of cybernetics voices coming from the social and biological sciences still held considerable influence, and there was much debate whether or not such a strict computational model was sufficient to fully explain the functioning of the human mind. However, the rapidly growing achievements and incredible potential of digital computing outshone alternative theories, and the computational model quickly became the dominant approach by which all cognition was understood (Boden, 2006). By the mid 1950 s social and biological concerns were shunted to the margins and the field of cybernetics became focused on a clear computational hypothesis (Posner, 1989). The central idea is that intelligent behaviour should be grounded in the capacity to represent the world intentionally that is, in the ability to cognize the aboutness of things and situations in ways that allow us to form representations of (about or directed to ) the world out there in certain ways rather than others; [t]o the extent that her representation of a situation is accurate, the agent s behaviour will be successful (all other things being equal) (Varela et al., 1993: 40). While there is something rather commonsensical about this view of things, the cognitivist account of the causal nature of intentionality is not quite so intuitive. The central problem, for the cognitivist, is to explain how intentional states are physically instantiated and how these states result in intelligent behaviour i.e. to explain what goes on between raw sensory input and intelligent behavioural output. The cognitivist solution to this question of how internal ideas and processes are related to external situations and actions in the world is symbolic computation : because symbols may be instantiated physically and may be - 12 -
ascribed semantic value, they are able to be subject to computational operations that function syntactically according to the language of the system (Haugleland, 1981). Or to put it another way, when semantic distinctions are encoded into the rules of syntax, abstract symbolic representations become possible; these representations are then able to be logically manipulated (i.e. computed) by the system in order to produce further representations and intelligent outputs (Betchel et al., 1998; Pinker, 2009). To be a cognizer is to possess a system of syntactically structured symbols-in-thehead (mind/brain), which undergo processing that is sensitive to that structure. Cognition, in all its forms, from the simplest perception of a stimulus to the most complex judgment concerning the grammaticality of an utterance consists of manipulating symbols in the head in accord with that syntax. The system of primitive, innate symbols-in-the-head and their syntactic, sentence-like structures is sometimes called mentalese. (Betchel et al., 1998: 63-64; also quoted in Johnson, 2007: 116) Of course, a computer carries out operations only on the physical form of the symbols available to it in accordance with the rules of syntax programmed into it by human beings. It possesses no knowledge of semantic values; the computer has no access to what a symbol, or group of symbols, is understood to represent, and therefore has no way of inferring the meanings of the computational processes themselves beyond the rules of its programmed syntax (see the Chinese room argument, Searle, 1990; Leman 2008; Dreyfus, 1979). Nevertheless, the computer has provided the dominant model or metaphor (Costall, 1991; Lakoff & Johnson, 1999) for the mechanics, grammar, or language of thought. - 13 -
Information processing, modularity and the adapted mind As I have just discussed, for most of its existence cognitive science has focused almost exclusively on the idea of representations produced by syntactic operations on physical symbol systems as the best way to understand the mind as an information-processing machine. By this view, mental operations are understood to proceed in a hierarchical process where increasingly complex representations of the world out there are generated at each stage according to the syntactic rules of mentalese (Fodor, 1983). Outputs (representations) from lower levels of processing (mental abstractions of raw sensory input) feed inputs to higher levels, where representations of form, structure and, eventually, reference and meaning are generated (Pinker, 2009; see also Fodor, 2003). In order to account for the incredible processing capabilities of the human mind, many theorists have enhanced the information processing approach with a modular conception of the mind/brain relationship. Fodor s (1983) initial theory proposed that functionally specified cognitive systems (i.e. modules) exist only in localized lower levels of processing. Two of the principle features of such modules, according to Fodor, are domain-specificity and information encapsulation meaning that each module works on a specific type of information and that the processing in a given module cannot be affected by information in other areas of the brain not directly associated with its input/output path. The modular approach has been greatly expanded by the field of evolutionary psychology, which, contra Fodor, understands the brain to be massively modular across all levels of functioning (Tooby & Cosmides, 1989, 1992; Pinker, 2009). This view attempts to explain the diversity of human thought, behaviour, and culture in terms of the evolution of a large array of such modules, each adapted by natural selection to serve a specific function related to survival: - 14 -
The mind is a system of organs of computation, designed by natural selection to solve the kinds of problems our ancestors faced in their foraging way of life, in particular, understanding and outmaneuvering objects, animals, plants and other people [...]. (Pinker, 2009) Indeed, evolutionary psychology understands the vast majority of human psychological activity in terms of adaptations that occurred early in the evolutionary development of human beings, when the species as we know it today was forming through the process of natural selection. Thus, by this view, an essentially stone-age mind is at the core of our modern day thoughts and activities. The influence of evolutionary psychology has been substantial. Much of its appeal, as Pinker s words clearly demonstrate, is due to the fact that it sets up a clear parallel between the cognitivist and adaptationist programs, where the origin and structure of human cognition are explained in terms of adaptation by natural selection. Like cognitivism, the central tenets of the adaptationist (or neo- Darwinian) approach can be stated fairly clearly. At the core of Darwin s original theory is the idea that biological evolution occurs through a process of modification by descent i.e. through mutation and the recombination of hereditary material through reproduction. The mechanism responsible for this process is known as natural selection, which chooses the phenotypes that function most effectively within a given environment; organisms are selected on the basis of how optimally they fit the environment at hand hence the famous phrase, survival of the fittest (Sober, 1984). The neo-darwinian program emerged in the 1930 s in order to synthesize new knowledge from the developing field of genetics with the concerns of classical Darwinism (Dawkins, 1976; Pinker 2009; see also Hecht & Hoffman, 1986; Ho & Saunders, 1984). This led to a focus on changes in the fitness of genes as the quantitative basis for understanding the adaptive traits organisms exhibit in relation to the environments they inhabit where the fitness of a given gene, and its - 15 -
associated phenotypic trait, is understood in terms of abundance (optimization of surplus offspring and the growth of an interbreeding population) and/or persistence (optimization of reproductive permanence; long term survival) (Dawkins, 1976). [ ] the dominant orthodoxy in evolutionary thinking over the last few decades saw evolution as a field of forces. Selective pressures [...] act on the genetic variety of a population producing changes over time according to an optimization of the fitness potential. The adaptationist or neo-darwinian stance comes from taking this process of natural selection as the main factor in organic evolution. In other words, orthodox evolutionary theory does not deny that there are a number of other factors operating in evolution; it simply downplays their importance and seeks to account for observed phenomena mainly on the basis of optimizing fitness. (Varela et al., 1992: 187) Thus the cognitive capacities of the human phenotype are understood to have emerged from adaptive processes associated with fitness optimization that occurred over an evolutionary timescale. This drives evolutionary psychology s central claims that the human mind evolved towards fitness optimization i.e. towards the capacity to create representations that optimally correspond with a stone-age hunter-gatherer environment and that many of the perceptions, thoughts, behaviours, and desires associated with modern life (a life we are supposedly not biologically adapted for) are largely parasitic, invasive or otherwise dependent on mental (computational) processes and structures (modules) that developed deep in human prehistory (Sperber, 1996; Sperber & Hirschfield, 2004). It is one of the central projects of evolutionary psychology to discern just what human activities and thoughts can be understood as properly adaptive from those that are biologically irrelevant (see Pinker, 2009). - 16 -
Some preliminary concerns I have offered here only a brief outline of the cognitivist and adaptationist programs. There is, of course, much more to be said about both. Nevertheless, I believe I have clearly outlined the central tenets of these two approaches and have demonstrated the clear theoretical connection they share in explaining the origins and functioning of the human mind. In Section Two I will consider their influence on music cognition research more closely. Before I go on, however, I should introduce a few critical concerns associated with the cognitivist-adaptationist point of view. Most importantly, I would like to draw attention to research that demonstrates a more dynamic and interactive relationship between genome and environment; and to discuss problematic issues surrounding perception, knowledge and consciousness associated with the cognitivist approach. These matters will be developed in the context of the embodied, bio-cultural, and enactive approaches to music cognition I discuss further on. While the neo-darwinian approach remains influential in psychological circles, other scientific fields appear to be moving away from such a strict adaptationist understanding of biological evolution (Gould, 1982; Gould & Lewontin, 1979; Sober, 1984; Ho and Sunders, 1984). An increasing number of evolutionary biologists have emphasized that the processes by which evolution proceeds are multiple and must be subject to levels of explanation (Sober, 1993; Oyama, 1985 Oyama et al., 2003). Indeed, it has been argued that the strict adaptationist approach does not properly take into account epigenetic factors and the important role of ontogenetic and environmental processes (Jablonka & Lamb, 2005; Goodwin et al., 1983). For example, one might consider here the discovery of polygenic traits, as well as the phenomenon of epistasis where the regulation and expression of a given gene is dependent on, and contributes to, the activity of other genes in the intra/inter-cellular environment via epigenetic processes; - 17 -
environmental and bio-chemical factors (e.g. hormones) play an important role (Lambert et al., 1986; Ridley, 2003). These insights and observations, among others, have led to a more dynamic understanding of how genes and environment interact (Lewontin, 1983; Oyama et al., 2003; Bateson & Mameli, 2007). Rather than focusing on the gene (Dawkins, 1976) or the individual as the fundamental unit of selection, researchers are turning their attention to how phenotypes develop via complex interactions and couplings across a range of units: DNA short sequences, genes, whole gene families, the cell itself, the species genome, the individual, inclusive groups of genes carried by different individuals, the social group, the actually interbreeding population, the entire species [...], the ecosystem of actually interacting species, and the global bio-sphere. (Varela et al. 1992: 192; see also (Meaney, 2001; Eldridge & Salthe, 1984) This substantially expands the orthodox Mendelian understanding of genetic inheritance, which posits an additive one directional schema (genes cells environment phenotype). By the classical view, genes trigger protein production, this guides the functioning of cells, which, with some influence from the environment, produce identifiable traits (Moore, 2003). This older approach works well when explaining so called single-gene disorders like Huntington s disease, or certain elementary physical features like eye colour, especially as they develop in relatively static and homogeneous environmental contexts (e.g. Mendel s pea plants). And there are also certain basic biological features that may still be understood in terms of a neo-darwinian comparative fitness scale (e.g. oxygen consumption; see Varela et al., 1993). However, a growing number of biologists find classical theories of genetics and evolution lacking in the context of more complex physical, behavioural, and psychological attributes such as personality or - 18 -
cognitive, athletic and musical ability, which increasingly appear to be heavily influenced by environment, motivation, activity and experience (Bateson, 2003; Bateson & Mameli, 2007; Meaney, 2001; Ericsson, et al. 2006; Sternberg, 2005). Furthermore, evolutionary psychology's conception of the massively modular mind has come under fire in recent years, most notably from Fodor himself (2001; see also Besson & Schön, 2012). This has resulted in more plastic and self-organizing conceptions of both the genome/phenotype and the mind/brain relationships (Maturana & Varela, 1987; Lewontin, 1983; Lickliter & Honeycutt, 2003; Pigliucci, 2001, Uttal, 2001). I should make it clear that it is not the idea of a genealogy of species that is in question. Rather, it is the mechanism by which this process occurs that is contentious. Darwin himself did not believe that adaptation through natural selection should be the sole force driving evolution. And indeed, it has been argued that natural selection (i.e. the constraints of reproduction and survival) may not be sufficient to shape genomes and organisms towards optimal fitness; and that survival of the fittest may not be the goal of evolutionary processes after all (Fodor & Piattelli-Palmarini, 2010; Gould & Lewontin, 1979; Lewontin, 1983; Varela, et al., 1993; Sober, 1984, 1993; Ho & Saunders, 1984). In brief, the dynamic-interactive approach that has developed over the last few decades (genome cells environment phenotype) is dedicated to better understanding the complex ways in which genes, proteins, and environmental factors including behaviour and experience interact with each other to guide the functioning of cells and the formation of phenotypes (Lewontin, 1983; Bateson & Martin, 2001). This view seeks to dismantle the classic nature/nurture dichotomy, preferring instead to examine the interaction between genes and environment as a dialectical phenomenon (Pigliucci, 2001) where no single unit or mechanism is sufficient to explain all processes. - 19 -
In Section Three I will make use of this dynamic understanding in order to give a biological grounding to the concept of structural coupling between organism and environment that is so central to the enactive approach to cognition (and that distinguishes it from other ecologically inclined theories, e.g. Gibson s, 1966). For the moment, however, it should be noted that despite the developments in evolutionary biology I have sketched above, evolutionary psychology and the cognitivist approach in general remains committed to the orthodox adaptationist conception of biological evolution (see Pinker, 2009). For the strict cognitivist there must be a means of optimizing representational correspondence between inner mental processes and a pre-given environment out there. This capacity is found in the modular computational mind provided by the selective constraints associated with survival and reproduction. Evolution is often invoked as an explanation for the kind of cognition that we or other animals presently have. This idea makes reference to the adaptive value of knowledge, and is usually framed along neo-darwinian lines. [ ] Evolution is often used as a source of concepts and metaphors in building cognitive theories. This tendency is clearly visible in the proposal of so-called selective theories of brain function and learning. (Varela et al. 1992: 193) Both the cognitivist and adaptationist programs depend on the notion of optimal fit (or correspondence in cognitive terms) with a pre-given environment. The key issue here is the notion of optimization between the otherwise autonomous categories of inner (genes, mental processes) and outer (environment). This division between mind and environment is, of course, nothing new. It is one of the central problems of modern philosophy, which often understands the mind as the mirror of nature (Rorty, 1979). As can be seen most famously in the mountain of critique surrounding the work of Descartes and Kant, such a dualistic perspective not only draws the ontological relationship between mind and body - 20 -
and into question, it also introduces serious epistemological issues regarding how, and to what degree, true objective knowledge of the world outside of our minds is possible. The thought of Descartes and Kant are often understood as precursors to the cognitivist philosophy of mind and I will have more to say about both of them below. Initially, cognitivism may seem to bypass many of the traditional philosophical problems associated with consciousness and knowledge of the world (Block et al., 1997). As I have discussed above, cognitivism adopts an a posteriori conception of knowledge as the result of symbolic representations that are physically instantiated in the brain through causal processes beginning with raw sensory input. Cognitivism s naturalizing approach is unconcerned with a priori representations and thus appears to avoid the metaphysical antinomies and transcendentalism, as well as the solipsism and skepticism that emerge in traditional debates. This has led strong advocates like Pinker (2009) to triumphantly claim not only that the cognitivist approach is wholly empirical and objective, but that it has neatly solved the mind/body problem as well. But while the classic substance dualism associated with Descartes is essentially a non-starter in current debates, a modern version of it is indeed at the core of the cognitivist approach to mind. As Damasio points out, the dominant idea is that mind and brain are related but only in the sense that the mind is the software run in a piece of computer hardware called the brain; or that the brain and body are related but only in the sense that the former cannot survive without the life support of the latter (1994: 247-48). This ingrained notion of the (rational-cognitive) mind as a disembodied and autonomous category (Leman, 2008) implies a number of other potentially troubling concerns. For example, a central aspect of the cognitivist model of mind is that the operations it describes must be played out at the sub-personal level (Dennett, 1978; Pinker, 2009). This means that not only are we not aware of such - 21 -
processes, but that we can never be aware of them. The understanding is that because these processes must occur rapidly there is no time for them to be parsed consciously (lest our ancestor fall prey to the lion that is about to spring from the bushes). This has prompted the obvious question of just how representational outputs of proposed innate cognitive modules are meaningfully recognized by the system beyond the mechanics of syntax, leading to homunculus metaphors and philosophical problems of infinite regress (Searle, 1990; Dreyfus, 1979; Clarke, 2005; Still & Costall, 1991; Potter, 2000). The cognitivist response follows that the characterization of these sub-personal systems in fanciful homunculus metaphors is only provisional, for eventually all such metaphors are discharged they are traded in for the storm of activity among such selfless processes as neural networks or AI data structures (Varela et al., 1992: 50; see also Dennett, 1978; and Pinker, 1997: 79). It has been argued, however, that this response (i.e. retreating into the complexity of mental activity) does not properly explain consciousness i.e. how the computational world of symbols and representations emerges into the daylight of phenomenal experience (Jackendoff, 1987). As cognition is clearly directed towards the world as we experience it (Varela et al., 1992: 52; see also Johnson, 2007: 4-6) it would seem that conscious awareness should be accounted for by any empirically based theory of mind. However, the issue of consciousness is often sidestepped because, for the strict cognitivist, consciousness and cognition are not synonymous: all that cognition requires is the ability to produce representations and intentional states; conscious awareness is not a prerequisite for cognition to occur. Thus the cognitivist program is generally not concerned with accounting for phenomenal experience. Rather it discusses notions of access-consciousness and executive functions, with only vague suggestions of how this might correlate with consciousness as sentience (see Pinker, 2009: 131-148). - 22 -
This said, some supporters of computational-representational cognition have refused to let the problem of consciousness slip away altogether. As Jackendoff (1987) has argued, explaining cognition must involve more than describing the relationship between a brain and a computational mind that is inaccessible to consciousness (the mind-brain problem). One must also account for the relationship between what he terms the computational mind and the phenomenological mind the mind-mind problem (1987: 20). Jackendoff attempts to deal with this issue by developing a theory of intermediate-level representations that are understood to support or project conscious awareness. An important implication of this approach is that it puts phenomenological constraints on computational models: The empirical force of this hypothesis is to bring phenomenological evidence to bear on the computational theory. [ ] Thus, if there is a phenomenological distinction that is not yet expressed by current computational theory, the theory must be enriched or revised. (Jackendoff, 1987: 25) This insight appears to highlight the open-ended approach necessary in including conscious awareness in the study of human cognition where the structural analysis of our minds and the development of cognitive theories are continually enriched by a disciplined examination of phenomenological distinctions. Another important and challenging insight offered by Jackendoff concerns the disunity of conscious experience. Indeed, our awareness of the world is modal we have distinct forms of awareness that correspond to our sensory capacities: visual, auditory, tactile and so on. His theory attempts to account for this by claiming, each modality of awareness comes from a different level or set of levels of representation. By this view, the disunity of experience arises from the fact that each of the relevant levels involves its own special repertoire of distinctions. (1987: 52). What is notable here is that instead of beginning with the notion that - 23 -
consciousness is unified and ultimately traceable to some unique locus, Jackendoff suggests that consciousness is fundamentally not unified and that one should seek multiple sources (1987: 52). This insight into the fundamental disunity of the cognizing subject is cause for a good deal of tension as it goes against the common assumption that there should be such a thing as a stable unchanging I at the centre of experience. Furthermore, the recognition that a proper theory of cognition (in its full blown consciously aware state) requires an ongoing discourse with direct experience also poses a challenge to modern science since the demise of the psychological movement known as 'introspectionism', the study of experience has been essentially ignored in psychological circles in favour of an objective approach to theory and research (Varela et al., 1993). This return to experience is also an issue for the dominant objectivist trends in analytic Anglo-American philosophy, which, after Frege (1970), understands rule-based logical propositions to be at the foundation of all meaningful thought and cognition (see Johnson, 1987, 2007; Lakoff and Johnson, 1999). Unfortunately, in the end Jackendoff glosses over the kind of disciplined and open-ended examination of conscious experience his theory appears to demand. As Varela et al. point out, Jackendoff assumes that everyday largely mindless experience provides access to all relevant phenomenological evidence and that the phenomenological quest is limited to just that largely mindless state. He considers neither the possibility that conscious awareness can be progressively developed beyond its everyday form (a strange omission given his interest in musical cognition) nor that such development can be used to provide direct insight into the structure and constitution of experience. (1993: 54) - 24 -
Jackendoff is a committed cognitivist. And while his assertion that the computational theory of mind must account for experience is a clear advancement, this does not mean he understands consciousness as having any causal influence over sub-personal computational processes. This leads him to the rather unpleasant conclusion that consciousness may not be good for anything (1987: 56). Neverthless, the observations Jackendoff offers are important and deserve further investigation, problematic though they may be for prevailing approaches. As I will discuss in Section Three, the issue of the non-unified subject and the (seemingly vicious) circularity between scientific objectivity and subjective experience are not problematic from the enactive point of view. To the contrary, these insights lie at the very heart of this approach, which also offers useful responses and alternatives to the bio-cognitive issues I have discussed here. To conclude I will return to the issues of phenomenal experience introduced above and consider what the enactive approach and its relationship to the Buddhist mindful-awareness tradition may offer in terms of understanding music's relationship to human consciousness and the experience of self. For now, however, I go on to consider the influences and implications of cogntitivism and adaptationism on the understanding of music cognition and meaning, and to explore alternative embodied and bio-cultural approaches. - 25 -
Two Music and Meaning From Information-Processing and the Music as Language Metaphor to Embodied Action and Bio-Cognitive Ecology In the opening pages of Ways of Listening: An Ecological Approach to the Perception of Musical Meaning (2005), Eric Clarke discusses the influence of information-processing theory on how we understand musical meaning. He demonstrates how the notion of cognition-as-representation is often accepted as a given in music cognition studies; and he offers an alternative ecological approach drawn from the work of Gibson (1966; see Section Three). Clarke identifies an ubiquitous view whereby music cognition is understood to proceed in a hierarchical and rule-based fashion from more basic to more complex levels of information-processing i.e. from the processing of basic sound attributes such as timbre, pitch and rhythm; to more explicitly cognitive functions such as the perception of form, tonality, meter, and melody; to advanced levels that deal with aesthetic value, and reference. According to Clarke, this perspective suffers from the same problems associated with the strict cognitivist approach to mind I discussed in Section One i.e. the question of the actual existence of sub-personal representations beyond theory; the homunculus problem; and the fact that the information-processing account of musical cognition tends to be disembodied (2005: 11-16). As Clarke argues, the cognitivist approach reduces musical experience to a kind of abstract "reasoning or problem-solving process" where "perception is treated as a kind of disinterested contemplation with no connection to action - 26 -
which bears little relationship to the essentially exploratory function of perception in the life of an organism" (2005:15). He further questions the validity of this approach as it appears to contradict direct experience. Indeed, we tend to understand music first in terms of its meaning in our lives and how it makes us feel (Langer, 1953, 1957; Johnson, 2007), only teasing apart its constituent structural elements post facto an activity that often requires difficult (and conscious) analysis if not sustained training. Such concerns notwithstanding, the idea that musical cognition should be, most fundamentally, a hierarchical and rule-based process remains a central assumption in music cognition studies. And perhaps not surprisingly, this has prompted much research into the relationship between music and language as cognitive systems. As I discussed in the last section, the computational approach to mind sees cognition proceeding according to the non-conscious language (i.e. syntax) of thought or mentalese (Fodor, 1983). This process is thought to have a conscious correlate in spoken and written language, which is generally understood to function via the organization of symbolic representations into hierarchical structures according to syntactic rules (Chomsky, 1975, 1980; Pinker, 1994, 2009). Music is discussed in terms of its relationship to language at structural, perceptual, and neuro-biological levels (Patel, 2008; Rebuschat et al., 2012); and a music as language metaphor often pervades common understandings of musical meaning (Johnson, 2007). Furthermore, because language is understood to be the adaptation par excellence of the human species, the evolutionary origin of the musical mind is examined largely in terms of comparisons to language in a neo-darwinian context (Pinker, 2009; Patel, 2008). As a result, debates over music s biological meaning and origin are often carried out within a dichotomous adaptationist nature or culture framework where, in comparison to language, musicality is often understood to be irrelevant from an evolutionary perspective (Pinker, 2009). - 27 -