WHY NEUROAESTHETICS?
Tudor Balinisteanu
University of Suceava (primary), and Goldsmiths, University of London (PT)
1. The Neuroaesthetics Discipline: Why Should You Study It?
Neuroaesthetics is a fairly new discipline studying the neural, neurophysiological, and physiological mechanisms involved in the perception and creation of all forms of art. Neuroaesthetics researchers use both quantitative methods (studying ‘aesthetics from below’) and qualitative methods (studying ‘aesthetics from above’). The two phrases between inverted commas anchor the empirical approach developed by the German experimental psychologist and philosopher, Theodor Fechner, in 1876: ‘aesthetics from below’ requires the study of a phenomenon using empirical and scientific analysis methods, whereas ‘aesthetics from above’ refers to the birds-eye view provided by philosophers and to the assessment of a phenomenon through deep introspection, offered by philosophers of art, as well as the artists themselves, through their work. Neuroaesthetics therefore stands at the intersection of science and the humanities. About thirty years old, neuroaesthetics has been steadily growing, picking up momentum very recently, with the development of the first masters programme in the world dedicated to this discipline in 2018, at the Psychology Department, Goldsmiths, University of London (entitled MSc Psychology of the Arts, Neuroaesthetics and Creativity, or PANC).
Neuroaesthetics was initially developed as ‘side work’ by major scholars such as David Freedberg (History of Art, Columbia University in the City of New York), Semir Zeki (Wellcome Centre for Human Neuroimaging, University College London), or Anjan Chatterjee (Neurology, Perelman School of Medicine at the University of Pennsylvania). Since the beginning, it constantly attracted followers from a wide range of disciplines from psycholinguistics to computer science. One way or another, researching the interaction between humans and art using both quantitative and qualitative methods has proven fertile, focusing scientists’ attention on features such as colour contrasts or rhythmic stimuli, with broader implications for re-assessing purely scientific research on, for example, attention or speech, respectively. Humanists, on the other hand, may find in neuroaesthetics a means to bring the depth and richness of their insight to bear on very particular instances of humans’ interaction with art, enabling them to reassess questions concerning the distinction between artistic and non-artistic stimuli, the institutionalisation of art, or the complexity of the psychological range involved in aesthetic perception.
Neuroaesthetics research seems to answer scientists’ need for engagement with art, and humanists’ need to further refine critical wisdom against the infinite complexities of the human brain. Industry stakeholders have since long tackled, in limited ways, the question of how potential customers react to colour, taste, musical rhythm, or inflections of the human voice. With the advent of Artificial Intelligence (AI) and Industry 4.0, both industry and academic actors have had to develop ways to integrate humanist wisdom and scientific knowledge. In our own lab, we focus on developing empathic traits for visual identities (faces) of virtual agents representing an AI. We are moving into an age where the integration of the humanities and science will only become stronger, with students from the sciences profiting from access to the insights humanities can provide, and students from the humanities profiting from access to quantitative methods involving statistical analyses and programming.
Hence, neuroaesthetics is a discipline that, par excellence, is poised to attract both humanists and scientists, as the first PANC graduates cohort demonstrates, with about half of the graduates coming from a humanities background (like this author), and the other half coming from a sciences background. We therefore address ourselves to students from a wide range of programmes in most disciplines of the humanities and sciences who are interested in taking a month-long module in Neuroaesthetics, within the student mobility programmes between the University of Suceava, the University of Bergen, and NTNU. The exchange programme is suitable for students at any level, from undergraduate level 2 (second year undergraduates) to masters level students without prior training in neuroaesthetics. Whatever the student’s level, no previous advanced knowledge of science methods or humanities critical analysis methods is required, beyond that gained from general education, which is to say that we address our manual to the educated reader irrespective of their previous or ongoing specialisation.
To get an idea of the range of disciplinary backgrounds of students who might choose to study a neuroaesthetics module, I am providing a list of disciplines from which we recruited students for our pilot programme (for details of this pilot, please access the following link, and navigate to Norway Student Mobilities for USV Students: https://neuroaestheticslab.usv.ro/neuroarts/announces/): the students’ majors were Food Quality, Psychology, International Studies, English Language and Literature, Romanian Language and Literature, Medical Systems and Equipment (Bioengineering), German Language and Literature, French Language and Literature, Spanish Language and Literature, Business Administration, Primary Education, Energy Systems Management, Tourism, and Communication and Public Relations. The students were from all study years (year 2 to year 4) and from masters programmes. Our own Romanian-Norwegian research team includes scholars from Psychology, English Literature, Bioengineering, History of Art, Computational Linguistics, Medical Psychology, Cognitive Neuroscience, Art and Media Studies, and Applied Engineering.
Given that each of these fields is richly grounded in paradigms established, in some cases, by several hundreds of years of research, why should one leave the comfort of established frameworks, methods, and research designs, not to mention the security provided by a community of like-minded scholars, in order to venture to study neuroaesthetics, a highly interdisciplinary field that requires, more often than not, higher level qualifications in at least two disciplines? The following two sections aim to provide a possible answer by briefly tackling the question of what is understood by ‘model’ in the humanities and the sciences. It is hoped that this will illustrate how scholars from each of the broad fields of the humanities and the sciences may benefit from stronger interdisciplinary communication. The final section will offer a short excursus on a topic of general human interest, namely, consciousness, that has captured the imagination of humanists and scientists alike. Far from aiming to clarify what consciousness is, or to provide definitive answers to the question of whether subjectivity and objectivity are commensurable, I will merely attempt to offer glimpses into exciting possibilities of interdisciplinary dialogue across the humanities and the sciences, dialogue focussed on the Big Questions about what it means to be human.
2. MODELS and REPRESENTATIONS of the World, Experience, and Reality in the Humanities and the Sciences
2.1.MODELS and REPRESENTATIONS in the Humanities. Social Constructionism and Science
Again, why study neuroaesthetics? Well, one might wish to simply enrich oneself by contact with a wider range of disciplines. This effort can be extremely fulfilling both personally and professionally. Especially when it comes to art, introspective aesthetics from above can seem like pure rhetoric, disconnected from the actual reality of measurable fact. One knows in one’s heart that things are in a certain way, and reaches out for the tools of philosophy, art, or critical interpretation to create a model of the world, or a phenomenon, or an experience (or a range of experiences). But without recourse to science, that model is never tested. One might say that it is not necessary to test it if it is convincing. But what does ‘convincing’ mean in a humanities research framework?
Humanities scholars find arguments convincing when broader implications of general human interest are reasoned out based on a sequence of ‘text’, to use a favourite structuralist and poststructuralist term, with ‘text’ referring to any form of representation, whether written text made of words, or art text made of images, or musical text composed of sounds, architectural text composed of shapes, dance text composed of dance phrases, or cinematic text composed of frame rhythms and moving images. By ‘reasoning out’ I mean ‘building a succession of logical deductions’. The argument does rest on empirical observations but what emerges from the concatenation of the observations and glosses accompanying these is a REPRESENTATION of the studied phenomenon mediated by subjective thought, and not an unmediated, direct presentation of it. We examine how other individuals understood an experience and represented it to their fellow men and women. We then use that unique and deep insight to perhaps shift our (and our readers’) worldview, so that our feeling of the world (really, how we feel the world) becomes more complex, richer, fairer, so that we may judge others, their actions, their motives, in more complex, deeper, and just ways. What enables this better, deeper, and more profound judgement and feeling is, at the end of the day, a MODEL of the world, or, again, a REPRESENTATION of the world. Am I being reductive in abstracting hundreds of years of humanist research into the statement that, ultimately, ‘the aim of the humanities is to present the world anew (thus re-present it) for the betterment of our societies’? That may be so, for no one single sentence can do justice to centuries of research in the humanities. However, I believe that saying that humanities research ultimately creates yet another representational model of the world based on other representations of the world, using introspection and reasoned argument, is neither simplistic nor reductive.
The model can become ever more complex, and shaping it through using arguments based on empirical observation and building on previous humanist insights requires experience in scholarship as well as life experience, by which I principally mean a wide psychological range. One might say that this is more important for writers, painters, architects, film directors, or dancers themselves, than for the logical humanist scholar, the literary critic, the historian of art, the film critic, or the performance studies analyst. But then: if the scholar is to build a MODEL (call it a new, politically important, or simply more satisfying worldview), and so a REPRESENTATION of the world, from other REPRESENTATIONS of the world, themselves based on particular life experiences, how can s/he do so without possessing a wide psychological range, that is, the ability to sense nuances, shifting colours of thought, emotions that words can barely capture, waves upon waves of thought, feeling, lightness and darkness, that create the stormy sea of the living stream (look up the phrase in W. B. Yeats’s poetry) of life, pounding at times, quietly rhythmical at times, never just one thing, never black and white, always changing colour as it reflects a changing landscape (mindscape?)?
In other words, the depth provided by humanist introspection, a light taste of which may perhaps be glimpsed from the metaphors I used in the previous, rather long, sentence, ultimately offers a subjective representation (or model) of the world. That somewhat poetic sentence, offering a textual representation of what it means to me, subjectively, to have psychological range, implicitly also offers a model of the world, a worldview. In that worldview, more sensitive people, whose taste has been formed by years of reading or viewing art, are valuable in our societies because they show us the manifold ways of reflecting our and others’ experiences. Yet, for all the richness and improvement in quality of life that art and commentary on art brings to our living together socially (the more sensitive we become through aesthetic experience, the more sensible citizens we will be), what we get is a model of the world that is objectively valid for only one person: the artist who created the model, or the critic who unveiled the model by adding critical wisdom to the artist’s world model. For anyone else, the model or worldview is relatively valid, according to how it tunes in with a person’s own subjective perspective or worldview. Blending one’s own subjective worldview into an externally received worldview makes a new worldview that, objectively valid for its author, will touch the subjectivities of others, and so on. Hence, creating models in the humanities is eminently a subjective process. A subject, an ‘I’, is at the centre and origin of the worldview he or she sets forth in argumentative expression. At the same time, no man is an island. A weaving of subjective representations of the world, created by consensus among critics/artists, each bringing a fairly unique perspective on the same experience complex, becomes so dense, that it acquires a certain quality of weight, of materiality. In other words the density of overlapping, interweaving, entangling tensely here, loosely there, subjective models (representations) of the world, arrived at through, mainly, introspection, makes the model seem equivalent to reality.
And so is opened a Pandora’s box (one, of many). For, as many scholars who adopt a radical social constructionist frame of thought propose, if that were true, reality gets constructed through such commonwealth of subjective thought. For such scholars, science is but another set of subjectively created forms of expressing, or representing, the world. Scientific truth would be a dense network of points of view that are as many offshoots branching out of the trunk of a professed objectivity, offshoots that interweave to create a texture that, having gained materiality and weight through consensus among science scholars, may be thought to hold truth, because it is thought to be an adequate REPRESENTATION of the reality of the world. But can such radical social constructionist perspectives adequately represent the endeavour of scientific research? I propose that they cannot. The main reason is that the core theory underlying radical social constructionist models cannot be tested. One cannot devise an experiment for testing the hypothesis that each human lives only in their individual subjective world of thought. One may test whether humans agree or disagree on values of an experience, but that means that the experience is mediated by a phenomenon that is either shared by several humans, and so it has a biologically universal, and therefore objective, dimension, or is simply a phenomenon that occurs outside one’s individual subjective realm of experience, meaning that the world is larger than one person’s mind, and that whatever is out there can be measured, being reflected as variance in individual experience.
But it is not my aim here to mount a full critique of social constructionism. What I do wish to highlight, by using the extreme example of radical social constructionism, is the difference between subjective and objective models and representations of the world. What I aim at is to clarify a distinction between the ways in which the humanities and the sciences work with models and representations, a distinction that might help us to understand the complementarity of humanities and science research. In fewer words, the distinction rests on finding out whether models (and representations) of the world can, and should be, tested. Perhaps some models of the world cannot be tested. In other cases, attempting to test models of the world created within the humanities would not significantly increase our knowledge of a phenomenon, or further deepen our understanding of the experience of that phenomenon. Hence, let us say that humanities research works mainly with models of the world that either cannot be tested, or testing those would not significantly add to the sum of our knowledge (meaning that the richness and depth of those models is sufficient). The models developed in humanities research are convincing because of a consensus arising among perceivers or critics of art, but their objective reality has often not been tested. Scientists do indeed create testable models based on the inexhaustible depth of rich humanist enquiry, but, ultimately, a critic’s or artist’s model is one person’s subjective experience of the world, one data point, so to speak. Data is available from only one participant and cannot be generalised with the means of science. No scientific model can be created based on data collected from only one participant. Nevertheless, again, the critic’s or artist’s model can be generalised by critical consensus, or consensus of feeling and emotion. Which is closer to reality, or to truth? Deeply felt, reasoned, conviction and shared feeling leading to consensus, or statistically validated measurement? And do we have to choose between these two? Let us postpone the answer to this question, and focus for a while on the meaning of models and representations of the world in scientific research.
2.2 MODELS and REPRESENTATIONS in the Sciences. Statistical Models and the Humanities
A scientific model, or representation of a phenomenon, is one that can be tested. Note that, having said that, it becomes nearly impossible to speak of a comprehensive model of the world, or a worldview. Unlike humanities research, science does not aim to cope, in a single effort (that is, an article or monograph), with the manifold complexity of a given experience as a whole, never mind the entirety of the human experience of, say, love, or justice, or spirituality. This is not to say that science is in some way poorer than the introspection characteristic of humanities research. Rather, science works patiently and painstakingly to create an objective representation of the world, or more appropriately put, of reality, that should emerge, in time, from a coherent corpus of validated hypotheses, each an ‘atom’ of truth, yet together the ‘body’ of reality. Against the view held by many scholars in the past, fewer in the present, science does not pretend to offer a disembodied perspective of reality. A once widely debated topic in the humanities, rooted in the seminal book, The View from Nowhere, by the American philosopher Thomas Nagel, was the thought that science claims to offer, on account of its claims to objectivity, a ‘view from nowhere’ (Nagel, 1986). I would propose that many scientists did not engage with the debate, not only because its echoes may have seldom been heard in science labs, but also because the phrase does not seem to adequately capture the ways in which objectivity is ensured in scientific research.
Indeed, no proper scientist can ever claim a purely objective standpoint, nor would they wish to do so. In fact, scientists have developed complex and sophisticated means to account for bias, that is, the variance that comes from errors attributable to the scientists undertaking an experiment, or from other external factors (please become used henceforth to read ‘factors’ as also meaning ‘variables’). However, a scientist, as ‘also’ a subjective human being, is not directly amenable to the analysis of bias. Rather, bias appears because of the ways in which a scientist’s statistical model is fit to the data. In discussing bias, I am following Andy Field’s exceptionally accessible book (highly recommended to all students undertaking this introductory course in neuroaesthetics), Discovering Statistics Using IBM SPSS Statistics (Field, 2020). It is our ambition, as lecturers on the module, to make scientific terminology accessible to humanities students who have never studied statistics, so please bear with me while I am attempting to explain scientific terms to a non-specialist audience. If you are a student with training in statistics, you may safely skip to the third section.
To understand where bias comes from in scientific research, we need to emphatically note that scientific analyses are based on samples. Hence, undertaking scientific analyses, we will no longer rely solely on the insight of the humanist isolated in his/her ivory tower, sunk in deep introspection, but on data that is associated with our sample. The sample is taken from the material world, and not from the thought-world that acquired materiality through the dense network of humanist thinking. You may remember that I am tempted to describe the latter world as engendering a sense of reality, rather than being objectively real.
One might pause here a little to think about what ‘a sense of reality’ means, for the phrase can be equally employed, with justice, in both humanist and scientific frames of thought. Let us tentatively say that ‘a sense of reality’ means a subjective understanding of the world, or of a phenomenon/ experience imbued intuitively with the quality of truth. This tentative definition, reductive as it is, would capture the subjective dimension of reality as studied in humanities research. The definition may arguably hold even when the means of introspection, at least in what regards mental processes of subjective and predicative expression, is, say, analytic philosophy. We will return to this topic tangentially in the final section. For now, suffice it to say that, for the purposes of the present section, the word ‘intuition’ is used to evoke the etymological Latin root of the word, intueri, which means ‘to look at, to consider’, a meaning quite different from more modern, even theologically inflected, meanings that spring to mind, such as ‘artistic intuition’. In science, the phrase ‘a sense of reality’ really means a sense of the phenomenon acquired from testing models of behaviour of members or items belonging to the selected sample. The sense of reality in this case will not be purely objective, but neither is it derived from a single data point, or from only one person’s perspective. True, a critical humanist might say that his/her arguments are built from comparing multiple points of view developed by other humanists. But what s/he offers, having considered these other points of view, is ultimately one point of view that ‘sub-sums’ those other. Had the humanist devised a research design based on established criteria in order to measure stated dimensions of those other points of view (read ‘subjective view points’), s/he would have found him/herself in the possession of one or several datasets, that is, in the possession of data extracted from a sample. Having found him/herself in this situation, the humanist could either once more resort to reasoned introspection, in other words, to considered interpretation (what ‘intuition’ evokes etymologically), or resort to scientific analyses, in which case s/he would be ‘forced’ to give up depth (no slighting of science here!) in favour of extensiveness. As neuroaesthetics scholars, you may well find yourselves in this position!
But back to the discussion of the term ‘sample’ and the topic of sample items/ members. In science, a sample is a multitude of items selected from the material world. The sample can be composed of dead matter (a geological sample of rocks, or an archaeological sample of bones, or a climate scientist’s sample of water, and so forth). However, in neuroaesthetics, as in psychology, we are concerned with a sample of individual subjective experiences, and therefore our sample will contain people (as well as other animals, in separate studies). In this case, we refer to our sample as a population sample more properly than other scientists who, for example, refer to their multitude of rocks as also a ‘population sample’ (because the rocks can be seen as a population). Note that we are not interested in the richness of the many points of view the people in our sample may cherish (as is the case with humanities scholars). Instead, we are interested in a limited number of dimensions of a specific experience that is lived by all the individuals in our sample population. What we are seeking is not to construct a subjectively considered point of view (a model of a worldview), but to determine (or model, here used as a verb) an objective outline of that experience. We are interested to find out how likely it is that a certain experience will occur, given a number of individuals with given characteristics (those characteristics that, based on a theory that may well have originated in the humanities, have informed the inclusion and exclusion criteria we have used in our sample selection). To further confuse students, one might say that we want to extract from each person’s point of view on an experience the exact same features of the experience, given that we control what the participants will experience (we control what the individuals will experience in our experiment via our experimental design). We will then use this data to create a model of the parts (features) of the experience that should apply to all individuals who display characteristics that are similar to the characteristics of the individuals in our sample. This model experience sits ‘above’ these individuals, and no embodied person will behave, in what concerns the features of the experience selected for scientific study, exactly according to the model, but the differences should be minimal. So much so that we should be able to say with sufficient certainty that, objectively speaking, those features of the experience that we study will be part of the experience of any individual that conforms to our sample selection criteria, and who has found him/herself subjected to conditions that are highly similar to those we have enacted in our experiment.
Perhaps we can now understand the difference between subjective introspection and objective assessment better. There is a middle ground, for there is bias in subjective introspection as there is bias in objective assessment. One might be philosophically inclined to say that there is chance in our Universe, and that we cannot master it. Perhaps chance is what makes our lives fulfilling, for we have evolved to cope with chance, and there is reward in mastering chance, when, and if we can do it. But, for fear of slipping into familiar humanists habits, let us go back to Andy Field and his wonderfully clear assessment of bias in statistical analysis. For the data collected based on any scientific hypothesis must, after the hypothesis has been operationalised in an experimental research design, stand the test of statistical significance.
We are hopefully enthusiastic at this point about creating a model that sits ‘above’ the individuals in our sample, yet a model that is, however distortedly, embodied in all of our participants. Suppose that we have measured a dimension of these people’s experience (yes, experiences can be said to have dimensions, and psychological science has various instruments of measuring these, such as questionnaires where each answer is assigned a figure, or indeed, physical instruments that collect mathematically expressible data, such as Electroencephalographs measuring brain waves experienced by a participant). To devise a scientific model of how an experience is lived by all our participants, we would first calculate the mean of measurements (values) that correspond to a dimension of experience as it has been lived by each individual in our sample. However, this mean may be biased if one individual in our sample happens to have an extreme version of the experience by comparison to most other individuals in our sample. We therefore need to see how representative our measurement is relative to the mean, or, how much each data point (one per individual) deviates from the mean (graphically, how far that data point is from the mean). The distance of each data point from the mean can be easily calculated by subtracting the mean value from each data point’s value. But what we want in order to test our model (the one that sits ‘above’ our participants) is a representative measure of how differently individuals behave, and simply having each data point’s distance from the mean doesn’t help: we would be tempted to sum up the differences as this seems to be a good way to get a figure that is representative for the entire sample, but the result would always be zero, for any sample. To ensure that we get a value that represents the variance among our data (mere figures abstracted from individuals), and to make a long story short, we square each distance and calculate the average of the resulting values. It will not be zero, and it will express the variance in our data set. This allows us to calculate units of variance, and we do this by taking the square of variance. The resulting value expresses the standard deviation. We can now go back and look at each data point to see how far each point is from the mean, but this time we have a measure of dispersion, and the unit of this measure is the standard deviation value. We would know, at this point, with mathematical certainty, whether individuals in our sample behaved in more or less similar ways with respect to a certain dimension of their experience (e.g. a specific emotional reaction). In other words, we would know whether our model does sit well with our participants who together make up our, hopefully, representative population sample. In fact, we would thus know whether the model we propose, or rather, how we model the phenomenon statistically, can or cannot be representative of the data.
There are other sources of bias, including, for example, our participants’ belonging to similar population samples, which we can assess by looking at confidence intervals. The values for a 95% confidence interval would tell us that, given 100 samples of participants from a population, the values measuring the dimension of experience of interest to us would fall within those confidence interval values 95 times. Knowing this would increase our, well, confidence, that our model is representative of how individuals generally live the experience we are interested in studying. If we are justified in saying that the model is a good fit to the data, we are justified in saying that our hypotheses are confirmed. It may be that our hypotheses are, indeed, derived from humanist research. We have taken in the depth of insight humanists have provided, and we have employed a mathematical model to check whether that insight passes the test of science. We may not have gotten here without the humanist insight! Scientists are known to be less imaginative, if not less creative, than humanists! (Do take that with a pinch of salt.)
How wonderful is the world of science!’, might an undergraduate humanist now exclaim. While future versions of this introduction will be further expanded and more relevant detail will be added, let us move on to an insight that Andy Field made central to his book on statistics. Before moving on, let us take forward the idea that a scientific model is amenable to statistical analysis, and that whatever conclusions we derive from ascertaining that our model is a good mathematical expression of the experience for which we have collected data, those conclusions are representative of a population composed of real, tangible, individuals (or things, but in neuroaesthetics we work with people). These individuals do likely embody the model in concrete ways as they fit within its slightly distorting embrace. We thank the humanist for providing the richness of insight that led to our hypotheses, but we cannot help noticing that that insight was highly subjective before we lent it a degree of certainty through using science, even as we are aware, and perhaps deplore, that our scientific modelling tools have much more limited scope than the imagination of the humanist, or, indeed, the artist.
But, to the promised insight, masterfully and playfully revealed by Andy Field: OUTCOME = MODEL + ERROR. ‘Outcome’ here refers to a prediction, and the equation also answers the perennially humanist question, ‘Why do we even need scientific/ statistical analyses?’. The answer, as you may have already inferred, is ‘In order to make mathematically accurate predictions about how a phenomenon will unfold’. You may replace ‘phenomenon’ with ‘dimension of experience’, ‘dimension of emotion’, etc. A validated (that is, statistically estimated as having high probability value) scientific model, in its difference from a humanist model, is capable of telling us whether a worldview is supported by data. The data is empirically collected, that is, it is objective, drawn from the reality of the living individuals who compose our population sample. The sources of bias have been eliminated to such extent that we may confidently argue that our model is generalisable. There will always be some error in our prediction. But it is tamed. We no longer depend only on prediction that results from subjective introspection. Indeed, by eliminating bias, we sought to eliminate all sources of error, including error that comes from subjective interference.
Let me offer a popular culture example, while apologising in advance for trivialising humanist research, if only for the sake of clarity (I will do it justice in the following section): a long held popular culture myth is that we use only a small percentage of our brain. How wonderfully this myth opens the gates to humanist, indeed artistic, interpretation! What marvellous infinity may there be hidden in our minds that awaits to be unleashed! What worlds may there lie within us, untapped and mysterious! Perhaps it is a locus of a collective consciousness, or a mental wormhole to brave new worlds, universes with several dimensions, all there within reach, however veiled at present. Sadly, no. Scientific enquiry invalidates this model of a worldview, even lifestyle. Our brains work 100% all of the time. But has the beauty of our speculation about what strange worlds may lie within our brain been washed away? No, it hasn’t. We feel it to be intuitively true, and that is why the myth persists. If that beauty is not hidden in the neural machinery of our brains, perhaps we should look for it elsewhere, in some other depths and complexities of our beings. Myth, and, more often than not, art, along with the craft of literary criticism or philosophy, are ultimately a matter of persuasion. Logos, or science, is a matter of negotiating models of objective reality. Would you be convinced that there is richness in making persuasion as sensual as possible with the aid of science? That strange and marvellous worlds uncovered or created by art are worth examining as concrete realms of experience? I would say that if the answer is ‘yes’, then neuroaesthetics is for you! Welcome to the fold.
3. The Sense of Art is the Emotion of Science
When we set up the Neuroaesthetics Lab we spent some time thinking about a suitable logo or slogan for our teams. It may not come as a surprise that we looked to humanists for inspiration. Thus, we started from the title of a novel by Jane Austen, Sense and Sensibility (1811). The conjunction of these two words has a long history. In a translation of Thomas Aquinas’ Commentary on Aristotle’s On sense and what is sensed, a treatise written about 1260 a. c., Kevin White points out that Austen’s title:
| which might be paraphrased as ‘Good Sense versus Sentimental Sensitivity’, suggests a confrontation between reasonableness and romanticism, or, as Aquinas might have said, between reason and the passions of sense-appetite; [J. L. Austin’s later work, Sense and Sensibilia (1962)] criticizes earlier twentieth-century British philosophers for the view that we never directly perceive or sense material objects, but only sense-data of our ideas, a criticism with which Aristotle and Aquinas would be in sympathy [even though they] are primarily concerned with perceptual consciousness of things, not sense-data […]. (White, 2005, p. 6, brackets added) |
In devising our logo, ‘The sense of arts is the emotion of science’, we played a little bit with the notions mentioned in the quotation above, in order to capture the idea that neuroaesthetics, as we understand it, is premised on a nature-culture integration, rather than on separating these. To explain: perceptual consciousness refers to how brain states translate into individual perceptions. How are our individually lived experiences determined by neural mechanisms? (For an excellent introduction to perceptual consciousness research see Fazekas & Overgaard, 2018.) Answering this question is harder than it might seem because we learn about the experiences of others, and, indeed, our own, because they are reported. Others tell us what they have lived, or we tell ourselves the story of our own experiences that we may also share, in turn, with others. This means that we might define consciousness as the totality of elements of a lived experience about which we can give an account in narrative form (although the narrative need not be chronologically linear, and need not be in words only – indeed, artists can develop arresting ways of experimenting with narrative in order to convey accounts of lived experiences to an audience). In this definition, accounts of lived experiences of which we have been conscious are always subjective. If so they are, they are not amenable to objective measurement. Sense-data, on the other hand, refers to lived experiences as they unfolded prior to our being able to use cognition to access whatever information exchange occurred between an individual and the world. Sense-data is data that our bodies ‘collected’ while immersed in a lived experience, but that has not been accessed cognitively. These data sets can be measured objectively.
This apparent division mirrors the nature-culture dualism we have inherited from centuries of scientific as well as humanist research as a methodological arch-premise. However, just as in the humanities many scholars have developed concepts that reflect nature-culture hybridity (e.g. Latour, 1993; Latour, 2004), innovative research in neuroscience has developed the idea that neural processes determine states of knowledge that can be reported, as if culture, where language is located, existed separately from nature, and states of knowledge that are not reported, and that point to a space between sense-data and perceptual consciousness, a space of embodied propositional possibilities. Embodied propositional possibilities comprise experiences that have not made it into cultural, conscious expression, but that depend on natural perception such as we have evolved to have as pro-social mammals. They are propositional because they are proposed in the relations between our organism and the affordances it encounters among physical objects in the material world. Using these affordances, the organism potentially positions itself in certain ways in material environments (Gibson, 2015). It is helpful to think that this is the ‘position’ part of ‘proposition’ (Latour, 2004). Could it be that consciousness comprises experiences expressed in some narrative account because a decision was made to report them, but the decision-making mechanism is a neural one that functions according to objective principles, indeed, outside consciousness, and heavily influenced by the need for survival and adaptation? Taking in recent neuroscience advances in the study of consciousness, Shadlen & Kandel (2021) argue in favour of the hypothesis that consciousness comprises the set of reports an organism has decided that it is worth reporting. Note that not an ‘I’ makes that decision. ‘I’, between inverted commas because it expresses the subject of a story, is an account I give of my existence, expressed through using the first person pronoun. It does have propositional valences, as a number of verbs can be predicated about the ‘I’, but not all that the organism, to which I attribute an ‘I’, has done, or can do, has been, or will be, reported. This may mean that the ‘I’ didn’t do it, but it does not mean that I didn’t do it, so to speak. I am not only what the ‘I’ does. At the same time, I am more than I without the ‘I’. I am in a state of nature-culture hybridity.
Let us elaborate a little within a neuroscientific framework. Within the human, and, generally, the mammalian brain, the basal ganglia have evolved to select a behaviour option from several competing options (should you wish to learn more, refer to Chapter 38 of the excellent Principles of Neural Science, edited by Kandel et al.) (Redgrave & Costa, 2021). The basal ganglia receives all kinds of excitatory input, conveyed through neural pathways, from all parts of the cortex and via the thalamus. These signals are qualitatively different, as they pertain to various aspects of mammal biological life including affective states, degrees of cognitive control, and/ or sensorimotor states conditioned by external stimuli from the natural environment. Within the complex architecture of the basal ganglia, these excitatory signals arrive at different functional territories of the striatum, where they are filtered according to magnitude. Magnitude, as a quantifier of salience, brings all these qualitatively different excitatory inputs within a common ground. The activity of neurons conveying weaker signals is suppressed whereas stronger signals are passed on for motor expression. There is a constant competition between magnitudes of qualitatively different excitatory input, but the criterion of selection for motor expression is not a qualitative one. Nor are ‘losing’ signals suppressed completely. In fact, a lot of input that might be taken to represent propositional possibilities, that is, actions one might perform, is kept under inhibitory control, but not fully suppressed. This input may or may not become expressed as action.
Furthermore, some of this input is ‘familiar’, that is, it would become expressed as an action that we previously performed, had it not been suppressed momentarily because more salient excitatory input was selected for motor expression. When I say ‘familiar’, I mean input that was deemed salient in the past, and which has been inflected, within the basal ganglia mechanisms, with positive or negative reward value. In time (indeed, in one’s life time) reward becomes itself a salience factor, within a mechanism termed ‘reinforcement learning’, thereby influencing the basal-ganglia decision making process.
Importantly, none of the processes described above are processes of which we are consciously aware. Selections and inhibitions of excitatory input happen amazingly fast, faster than (conscious) thought, so to speak. In terms of data, one might term these quantifiable processes as based on sense-data, that is, these processes work with elements of experience that precede conscious expression. In the perspective on decision-making and consciousness outlined by Shadlen & Kandel (2021), one might argue that whether experiences become expressed in conscious narrative accounts depends on the salience of a need to account for those experiences. It may be more important for an individual to express an experience to himself/herself in an account (a narrative or story), or to share that experience at a given point in time, than it may be to drink water, for example. However, when the need to drink water is more important for survival than sharing stories, the decision will go in that direction, even though the reward one derives from sharing or expressing an experience may bias the decision-making process in favour of the action of giving a narrative account of some other experience. Needless to say, drinking water will always win in the end in extreme situations. This example of what may be seen as a competition between more spiritual needs and bodily needs illustrates the point that the mechanisms of selection and inhibition within the basal ganglia have evolved over millions of years, being conserved evolutionarily, which means that they are encountered in mammals that are not endowed with conscious expression capacity. Therefore, while mammals, including humans, make decisions, making decisions does not always require consciousness. It is, of course, possible for an expressed (conscious) narrative account to loop back to the basal ganglia in the form of excitatory input that will compete with other salient input for motor expression, for example when we ask a question that is sufficiently imperative to receive an answer form a collocutor (or from ourselves) (provided we, and our collocutor, are not dying of thirst within sight of water).
In short, human consciousness may be more properly defined as systematic biological functioning that has as effect accounts (narratives, stories) of lived experiences. This may hopefully also resolve the social constructionist question of reality being purely subjective, that is, fully created in consciousness. It is not. Biological mechanisms functioning outside awareness and consciousness create symmetries, asymmetrical spaces, colour, shades, rhythm, and so forth, and assemblages of these compete for expression in conscious thought, or manage to become sufficiently salient stimuli in order to trigger responsive motor expression. There is a subjective reality, and mining it through introspection is itself rewarding. It enriches our experiences, whether we are giving reasoned accounts, or accounts influenced by sense-appetites. These accounts do give depth and psychological range to what we perceive as reasonable or attractively un-reasonable. Neural processes can be characterised, somewhat metaphorically, as infinitely complex, which would warrant the humanist statement that realms of emotion and imagination are infinite. Our biological needs may or may not overpower, extend, force us to push the limits of these realms, or defend them, or restrict them. But the criterion by which biological mechanisms allow for any or several of these acts is salience in terms of adaptation and survival.
Herein lies a possible answer to the question, ‘Why neuroaesthetics?’. To me, the most important reason for studying the perception and creation of art using the methods of neuroaesthetics is that such research makes aesthetic experiences more concrete. There is a concreteness in aesthetic experience that comes from the immediacy of sensation, whether we speak of somatic effects of certain words, rhythm, symmetry, colour, and so forth. But there is another kind of concreteness that points to survival and adaptation, a concreteness that comes from the immediacy of our animal existence. This latter kind of concreteness is probably seldom associated with the former kind of concreteness. That is, we sense beauty, or richness, and the concreteness of that experience is sufficient in the moment of aesthetic experience, or upon recollection (be that in tranquillity, as the romantic poet, William Wordsworth, put it). But there is more terrible beauty (as another Romantic poet, W. B. Yeats, put it) that we can experience when we understand the physical limits of consciousness, or of our sense organs, that reveal the sound and fury of life that struggles to continue and adapts to adverse or friendly material environments in order to flourish.
Humanists will argue that such limits have been probed beautifully by art. Indeed they have been. But that depth experienced subjectively is only one side of the coin. That depth has what might be termed subjective density, that is, a quality of realness that thrills because it evokes previous aesthetic accounts of the sound and fury of life (to use William Shakespeare’s philosophical-poetic words in this context). Nevertheless, when matched to scientific knowledge about how that sound and fury are experienced viscerally, the humanist and artistic accounts become more complete. The concreteness of measurable physicality is added to the concreteness of somatic immediacy that does not require a measured mind (Yeats again, as humanists might have figured!). The model of the world disclosed aesthetically, or through reasoned introspection, agreed upon through humanist consensus, acquires further concreteness when verified experimentally, that is, when, if only partially, data sampled from material life fits the model. Metaphorically speaking, humanities research gives depth and density of texture when reflecting a phenomenon, while science gives physical limits, and therefore concreteness of shape and dimension, to the same phenomenon. If you wish to utterly understand the phenomenon, you will need qualitative as well as quantitative means of assessing it, and the discipline of neuroaesthetics provides a space for working with both.
References:
Fazekas, P. & Overgaard, M. (2018) Perceptual consciousness and cognitive access: an introduction, Phil. Trans. R. Soc. B3732017034020170340 http://doi.org/10.1098/rstb.2017.0340
Field, A. (2020) Discovering Statistics Using IBM SPSS Statistics, Fifth Edition. London: Sage.
Gibson, J. J. (2015) The Ecological Approach to Visual Perception. New York: Psychology Press.
Latour, B. (1993) We Have Never Been Modern, trans. C. Porter. Cambridge, MA: Harvard University Press.
Latour, B. (2004) How to talk about the body? The normative dimension of science studies,Body and Society,10(2-3).
Nagel, T. (1986) The View from Nowhere. Oxford: Oxford University Press.
Redgrave, P. & Costa, R. M. (2021) The basal ganglia. In E. R. Kandel, J. D. Koester, S. H. Mack, & S. A. Siegelbaum (Eds.) Principles of Neural Science (Sixth Edition). New York: McGraw Hill, 932–952.
Shadlen, M. N. & Kandel, E. R. (2021) Decision-making and consciousness. In E. R. Kandel, J. D. Koester, S. H. Mack, & S. A. Siegelbaum (Eds.) Principles of Neural Science (Sixth Edition). New York: McGraw Hill, 1392–1416.
White, K., trans. (2005) St Thomas Aquinas, Commentary on Aristotle’s On sense and what is sensed. Washington D. C.: The Catholic University of America Press.