Back to the Online Metaphor Center.

Abstract

As I have just recently started on my project, I am not yet fully oriented in the expansive neurological/cognitive landscape; thus, I am grateful for any comments on the following presentation, as well as for any references that might be relevant.

"To the one object we attach a variety of modes of consciousness, acts or act-noemata. [...]
But separated acts, as, for instance, two perceptions or a perception and a memory, can likewise close together in a "harmonious" unity, and by means of the unique nature of this closing together, which is clearly not foreign to the essential being of the acts thus linked together, the Something of the initially separated nuclei, which at one time may be determined thus or thus and at another otherwise determined, is now consciously grasped as the same Something, or as in common accord the same "object".
...this "modal form" is to be understood as that precisely which the act at the time prescribes, and as such therefore it really belongs to its noema. The "meaning" (Sinn) [...] is this noematic "object in its modal setting" ("Gegenstand im Wie")..." (Husserl: 366-367)

"Whatever the skill employed in thought - that of logic, mathematics, language, spatial or musical symbols - we must not forget that it is driven by the Jamesian processes (see note 3), undergoes flights and perchings, is susceptible to great variations in attention, and in general is fueled by metaphorical and metonymic processes. It is only when the results of many parallel, fluctuating, temporal processes of perception, concept formation, memory, and attentional states are "stored" in a symbolic object - a sequence of logical propositions, a book, a work of art, a musical work - that we have the impression that thought is pure." (Edelman: 174)

I wish to stress that it is not always possible to make clear distinctions between different types of musical meaning, as our categorizations are often subjectively grounded (cf. f.ex. the many divergent views on what to place within the category of intramusical meaning). Furthermore, music cognition involves an inconceivable amount of heterogeneous associations that will contribute to the import of meaning no matter which aspect of cognition a listener chooses to focalize. This latter claim touches upon the difficult problem of attention and conscious vs. unconscious processing (see note 4); a problem that I have chosen not to elaborate upon here. However, I do wish to point at the fact that although we pay attention to only one or a few things at a time, our understanding of what we are paying attention to is partly determined by aspects of cognition that we are not aware of at the time (cf. the way film music can create an atmosphere even when one is not consciously aware of the music). Among those who emphasize the importance of unconscious cognitive processing, is the American neurologist Antonio Damasio, who speaks of priming - "activating a representation incompletely, or activating it but not attending to it" (Damasio: 106) - in his book Descartes' Error.

It is also interesting to note that the areas of the brain that are believed to play an important role in the direction of attention, the thalamus and the basal ganglia (Guyton: 244, Edelman: 142ff), are tightly connected with areas of the brain that are essential to our emotional life; namely the so-called "limbic system" and the frontal cortices - a part of which belong to the limbic system. Edelman claims that attention is directed on the basis of both conscious and unconscious value judgments (Ibid.: 143), and Damasio has presented a somatic-marker hypothesis which can explain the role of bodily affect both in the direction of attention and in the conscious and unconscious processes underlying decision-making (Damasio: 165ff). They both believe that we have a tendency to direct our attention towards that which means the most to us personally; a view that can throw interesting light upon the above-mentioned discussions of musical meaning.

To conclude my comment on musical meaning, I wish to cite two passages from Descartes' Error. In the first passage Damasio discusses the problem of decision-making, stressing that he regards "abstract-symbolic operations under which we can find artistic and scientific reasoning" as a phylogenetically recent decision-making device (Ibid.: 191):

"What dominates the mind landscape once you are faced with a decision [or in our case, a musical composition, which the brain automatically sets out to analyze] is the rich, broad display of knowledge about the situation that is being generated by its consideration. Images corresponding to myriad options for action and myriad possible outcomes [or in our case, myriad associations evoked by the music] are activated and keep being brought into focus. The language counterpart of those entities and scenes, the words and sentences that narrate what your mind sees and hears, is there too, vying for the spotlight (see note 5). This process is based on a continuous creation of combinations of entities and events, resulting in a richly diverse juxtaposition of images which accords with previously categorized knowledge." (Ibid.: 196; bracketed notes are mine)

"...when we see, or hear, or touch or taste or smell, body proper and brain participate in the interaction with the environment.
Think of viewing a favorite landscape. [...] Sooner or later, the viscera are made to react to the images you are seeing, and to the images your memory is generating internally, relative to what you see. Eventually, when a memory of the seen landscape is formed, that memory will be a neural record of many of the organismic changes just described, some of which happen in the brain itself (the image constructed for the outside world, together with the images constituted from memory) and some of which happen in the body proper." (Ibid.: 224-225)

I believe that the meaning of a musical composition is the network of images it evokes in each individual listener - images being understood as the mental "pictures" we know ourselves, our bodies, our feelings, our thoughts, and the world by, regardless of which cognitive modality they involve, and regardless of whether they consitute objects or processes. As many of these images are personal and episodic, I regard musical meaning as a subjective and singular phenomenon; although I do not wish to contest the fact that many of the images evoked by a musical composition will be intersubjective and recurrent, due to shared biological and cultural dispositions among the listeners. I do not, either, wish to contest the fact that the auditory representations of music that are created in response to intramusical structures, are essential to the meaning of music. My goal is rather to argue that auditory representations are only one of many aspects of musical meaning.

Before considering Edelman's work, one should note that Edelman himself takes pains to stress that his TNGS is a theory in need of verification through experiments (cf. Edelman: 71+208). Thus, the reader is warned to take the TNGS not as a proven fact, but as a set of hypotheses that have yet to be tested empirically.

Now turning to the concerns of my paper, I wish to emphasize that music cognition, like all cognition, is based on recognition, which Edelman describes as "adaptive matching"; making explicit references to evolutionary and immunological processes (Ibid.: 81ff). Adaptive matchings underlie the perceptual and conceptual categorizations which enable the brain to "recognize" (i.e. understand) new objects and events - both externally and internally derived. They are the result of selectional processes which occur in reference to the value they represent to the individual (cf. Edelman's term "neural Darwinism"). According to Edelman, "Such value criteria do not determine specific categorizations but they constrain the domains in which they occur." (Ibid.: 90) He believes that the basis for the human brain's inherent value system is set by evolutionary selection; a selection which has ultimately led human beings to consider also social and cultural values (cf. Ibid.: 174). I wish to add personal and purely emotional values to the list, as these are fundamental to our aesthetic experiences.

To reach an understanding of how perceptual categorization is possible, I wish to start out with a comment on the cerebral cortex, which is paramount to our intellectual functions (In this paper I have chosen to concentrate on what happens to music after it reaches the cortex; it should, however, be noted that the categorization processes start already on the subcortical level). The number of neurons (nerve cells) in the cortex alone has been estimated to 1010-1011 (Guyton: 227) - which in itself can give a picture of the complexity of cognition. The cortex is divided into many different functional areas, and I will now present a igure of these areas, rendered from Arthur C. Guyton's Basic Neuroscience:

fig. 2-5 (Sorry! This was a transparency...)

Experiments have shown that the brain's neurons are rarely stimulated individually, but rather in smaller or larger groups (cf. f.ex. Edelman: 89) which are highly specialized, but aslo flexible. The neuroscientists have come especially far in their understanding of how visual stimuli are processed: The neurons in the visual cortex are organized in groups (columns) corresponding to specific areas in the retina. Each of the cells within the groups react exclusively to certain contrast directions, contours, or other specifically defined properties (Guyton: 302). The auditory cortex is believed to be just as specialized, and one has discovered several tonotopic maps (brain "maps" where there is a correspondence between the tones' frequencies and their situation in the map) in the primary auditory cortex and the auditory association areas (see note 7). Each of these maps are believed to be responsible for the analysis of certain sound features; for example the detection of pitch and sound direction (Ibid.: 317-318) - and through the correlation of many such maps a perceptual image arises.

Returning now to Edelman's work, I wish to begin with the three tenets of his TNGS, which describe three different principles of neuronal grouping (Edelman: 83ff):

1) Developmental selection of neural networks which are unique to every individual.
2) Experiential selection of functional circuits, where synapses (contact points between neurons) are strengthened as our experiences are recorded, enabling all of the neuronal groups constituting a memory to be reactivated simultaneously.
3) A process Edelman calls "reentry", where maps formed during the processes described in 1) and 2) interact:

"These maps are connected by massively parallel and reciprocal connections. [...] Reentrant signaling occurs along these connections. This means that, as groups of neurons are selected in a map, other groups in reentrantly connected but different maps may also be selected at the same time. Correlation and coordination of such selection events are achieved by reentrant signaling and by the strengthening of interconnections between the maps within a segment of time." (Ibid.: 85)

"The signals are delivered to the early sensory cortices. [...] Note again that this is a collection of areas rather than one center. The areas that are part of the collection are individually complex and the mesh of connections they form is even more so. The topographically organized representations result from the concerted interaction of these areas, not from one of them only." (Damasio: 99)

fig. 9-2 (from Edelman 1992, omitted in html version)

During the categorization process each brain map receives independent signals from the external world or from other maps, subsequently connecting with new maps (Edelman: 87). Maps pertaining to different sensory modalities can also interact; and there even exist maps that are specialized at combining other maps - for example, one has localized a map that is activated by the combination of music listening and score reading, but not by listening or score reading alone (Sergent et al.: 108). Moreover, information from sensory maps is relayed both to the frontal lobes, which are, among other things, important for the linking of information to our personal experiences; to the lower, non-mapped parts of the brain, which are, among other things, essential to our emotional life; and to motor maps, which enable us to choose a motor behavior appropriate to the sensory information we are receiving (for example, to let our eyes follow the notes in a score as the music is sounding, or to run if we meet a wild boar in a Berlin park...) Edelman calls this vast network of maps a global mapping, stressing that these are dynamic structures which are being modified continuously by our changing external and internal environment. Although he uses animals' motor behavior as an example, the following quote can also illustrate how the brain reacts to the temporal art of music:

"...a global mapping allows selectional events ocurring in its local maps [...] to be connected to the animal's motor behavior, to new sensory samplings of the world, and to further successive reentry events.
Such a global mapping ensures the creation of a dynamic loop that continually matches an animal's gestures and posture to the independent sampling of several kinds of sensory signals. Selection of neuronal groups within the local maps of a global mapping then results in particular categorical responses." (Edelman: 89)

By introducing musical structure into the discussion, I have entered the domain of conceptual categorization, which according to Edelman involves the interaction of many different perceptual categorizations. He defines conceptual capability as "the ability to categorize in terms of general or abstract relations" (Edelman: 101) and stresses that contrary to what is commonly believed, he regards conceptualization as a phylogenetically prelinguistic phenomenon (see note 8). Edelman believes that recognition is the fundamental principle underlying both perceptual and conceptual categorization; however, contrary to perceptual recognition, conceptual recognition is relational:

"It must be able to connect one perceptual categorization to another, apparently unrelated one, even in the absence of the stimuli that triggered those categorizations. The relations that are captured must allow responses to general properties - "object", "up-down", "inside", and so on. Unlike elements of speech, however, concepts are not conventional or arbitrary, do not require linkage to a speech community to develop, and do not depend on sequential presentation." (Ibid.: 108)

"Conceptual categorizations are enormously heterogeneous and general. Concepts involve mixtures of relations concerning the real world, memories, and past behavior. Unlike the brain areas mediating perceptions, those mediating concepts must be able to operate without immediate input." (Loc. cit.)

According to Edelman, these conceptual categories are a prerequisite of consciousness. He postulates two different levels of consciousness, both dependent on conceptualization: primary consciousness, consisting of perceptual scenes that are based on "the correlation by a conceptual memory of a set of ongoing perceptual categorizations" (Ibid.: 120); and the more complex higher-order consciousness, which is capable of linking concepts of the self, the past, and the future to primary consciousness (Ibid.: 132; see note 9).

Edelman sees linguistic competence ("semantic bootstrapping") as a prerequisite of the specifically human higher-order consciousness. However, because language is a faculty that is shared by all humans with normal mental equipment, and because it is virtually impossible to investigate human consciousness without the use of language, his claim can be neither verified nor falsified. This is an aspect of the TNGS which I find problematic, as I am studying musical meaning; an attribute of human consciousness that is not essentially linguistic. Therefore, I have decided to leave Edelman at the level of concepts to follow Damasio's work through the remaining steps of my discussion.

"...the neural activity that is most closely related to the images we experience occurs in early sensory cortices and not in the other regions. The activity in the early sensory cortices, whether it is engaged by perception or by recall of memories, is a result, so to speak, of complex processes operating behind the scenes, in numerous regions of the cerebral cortex and of neuron nuclei beneath the cortex, in basal ganglia, brain stem, and elsewhere. In short: Images are based directly on those neural representations, and only those, which are organized topographically and which occur in early sensory cortices. But they are formed either under the control of sensory receptors oriented to the brain's outside (e.g., a retina), or under the control of dispositional representations (dispositions) contained inside the brain, in cortical regions and subcortical nuclei." (Ibid.: 97-98)

Of special interest in this paper are the dispositions underlying the generation of what I have chosen to call creative images; but before I turn to this problem, I wish to make a brief comment on selfhood, which Damasio sees as a prerequisite of consciousness - and thus as a prerequisite of the conscious experience of images. He believes that the topographically organized representations formed by the early sensory cortices are necessary but not sufficient for images to occur in consciousness, and claims that "Subjectivity, a key feature of consciousness, would be missing from such a design. [...] In essence those neural representations must be correlated with those which, moment by moment, constitute the neural basis for the self." (Ibid.: 99) He uses the perception of a face as an example of how such correlations take place; however, his discussion can apply equally well to the perception of music:

"As images corresponding to a newly perceived entity (e.g., a face) are formed in early sensory cortices, the brain reacts to those images. This happens because signals arising in those images are relayed to several subcortical nuclei (e.g., the amygdala, the thalamus) and multiple cortical regions; and because those nuclei and cortical regions conatin dispositions for response to certain classes of signals. The end result is that dispositional representations in nuclei and cortical regions are activated and, as a consequence, induce some collection of changes in the state of the organism. In turn, those changes alter the body image momentarily, and thus perturb the current instantiation of the concept of self." (Ibid.: 240-241)

Now that we have seen how Damasio explains the experience of a conscious, perceiving self, I wish to return to the problem of how creative images can arise in consciousness - beginning with a quote from Descartes' Error on the function of dispositional representations:

"When dispositional representations are activated, they can have various results. They can fire other dispositional representations to which they are strongly related by circuit design (dispositional representations in the temporal cortex, for example, could fire dispositional representations in the occipital cortex which are part of the same strengthened systems). Or they can generate a topographically organized representation, by firing back to early sensory cortices directly, or by activating other dispositional representations in the same strengthened system. Or they can generate a movement by activating a motor cortex or nucleus such as the basal ganglia." (Ibid.: 105)

What, then, are the cerebral processes underlying such associations? Although this question is not addressed directly in Descartes' Error, there are traces in the book which point towards a possible answer. On several occasions Damasio stresses the importance of the prefrontal cortices, employing Jean-Pierre Changeux's term "generator of diversity" to designate their role in the cognitive processes (Ibid.: 196). According to Damasio, the prefrontal structures lead to "a continuous creation of combinations of entities and events, resulting in a richly diverse juxtaposition of images which accords with previously categorized knowledge" (Loc. cit.; cf. p. 3). This is possible because the prefrontal cortices receive signals from all over the brain, including the early sensory cortices which form the images constituting our thoughts. In addition, the different parts of the prefrontal cortices themselves (the ventromedial and dorsolateral sectors, to which I will return shortly) are mutually interconnected (Ibid. 180-181). Thus, as Damasio writes, it is in the prefrontal cortices that "upstairs and downstairs come together harmoniously" (Ibid.:183; see note 10):

"The prefrontal sectors are indeed in a privileged position among other brain systems. Their cortices receive signals about existing and incoming factual knowledge related to the external world; about innate biological regulatory preferences; and about previous and current body states as continuously modified by that knowledge and those preferences. [...]
...the prefrontal cortices themselves represent categorizations of the situations in which the organism has been involved, classifications of the contingencies of our real-life experience. What this means is that the prefrontal networks establish dispositional representations for certain combinations of things and events, in one's individual experience, according to the personal relevance of those things and events." (Ibid.: 181)

"...the bioregulatory and social domain seem to have an affinity for the systems in the ventromedial sector, while systems in the dorsolateral sector appear to align themselves with domains which subsume knowledge of the external world (entities such as objects and people, their actions in space-time; language; mathematics, music)." (Ibid.: 183)

To conclude, I will repeat some of the prefrontal cortices' most important functions: They receive information from the entire brain, they appear to be engaged in the direction of attention as well as in the subsumption of knowledge of our external and internal environment, and they function as a "generator of diversity". Thus, it seems likely that the prefrontal cortices are of central importance to the creative associations we make when we listen to music. It is also interesting to note that the activity of the prefrontal cortices is emotionally constrained and based on each individual's real-life experiences; an observation which can throw interesting light upon the meaning of music.

My next step is to study the cognitive principles underlying creative associations, namely metaphor and metonymy - but that will have to be the subject of another paper...

2) I will later discuss different views on the nature of concepts.

3) The psychologist William James (1842-1910) claimed that consciousness was a personal and intentional process and not a substance, as had been believed earlier (Edelman: 6).

4) It should be noted that it is common to distinguish between unconscious and non-conscious cognitive processes; i.e. processes that are never available to consciousness (cf. f.ex. Edelman: 143).

5) I believe that language is a mode of consciousness that not only "narrates" what the mind sees and hears, but also constitutes a new form of meaning with specific linguistic constraints.

6) I have chosen to put little emphasis on the images of visceral sensations that are elicited by music listening, due to the limitations on the length of this paper. This is, however, an important aspect of musical meaning; an aspect to which I plan to return in my dr. art. thesis.

7) The "main route" for auditory information is from the peripheral nervous system (PNS) via several subcortical nuclei (groups of neurons) to the primary auditory cortices, which subsequently excite the auditory association areas.

8) It is interesting to note that the folk notion of concept has been criticized also within cognitive linguistics; among others by Gilles Fauconnier and Mark Turner, who seem to share Edelman's view on concepts: "We have no concept house, but we do have a word, "house", and being able to use that word [...] requires the ability to construct, link, and activate the appropriate space configurations, frames, and cognitive models. [...] Any single use of the word "house" for any particular purpose will involve construction of meaning as an operation of selective integration over these various distributed spaces." (Fauconnier+Turner: 33; cf. also Turner 1996)

9) In this respect it is interesting to compare Edelman with Nietzsche, who believed that the essential difference between humans and animals lay in the animals' experience of an eternal "now" without a past to correlate the experiences from one moment to the next (Nietzsche: 211).

10) Damasio is here referring to the cortical and subcortical regions of the brain. The subcortical regions play an important role in the processes of bodily homeostasis, emotion, and memory.

11) ventromedial = towards the front and the middle; dorsolateral = towards the back and the sides.

Edelman, Gerald (1992): Bright Air, Brilliant Fire: On the Matter of the Mind, London.

Fauconnier, Gilles and Turner, Mark (1994): "Conceptual Projection and Middle Spaces", UCSD: Department of Cognitive Science Technical Report 9401.

Føllesdal, Dagfinn (1989): "Fenomenologien: En tilnærming til det subjektive" in Liv Bliksrud and Asbjørn Aarnes (eds.): Spor etter mennesket: Essays til minne om A. H. Winsnes, Oslo.

Guyton, Arthur (1987): Basic Neuroscience: Anatomy and Physiology, Philadelphia.

Husserl, Edmund (1976): Ideas: General Introduction to Pure Phenomenology, London (transl. W. R. Boyce Gibson).

Johnson, Mark (1987): The Body in the Mind: The Bodily Basis of Meaning, Imagination, and Reason, Chicago.

Lübcke, Poul (1988): Filosofilexikonet, Stockholm (transl. Jan Hartman).

Meyer, Leonard (1956): Emotion and Meaning in Music, Chicago.

Nietzsche, Friedrich (1954): "Vom Nutzen und Nachteil der Historie für das Leben", Unzeitgemässe Betrachtung II in Karl Schlechta (ed.): Werke I, München.

Sergent, Justine et al. (1992): "Distributed Neural Network Underlying Musical Sight-Reading and Keyboard Performance" in Science, vol. 257, July 3.

Smoliar, Stephen (1993): "Parsing, Structure, Memory, and Affect: A Critical Review of the Generalized Theory of Tonal Music (GTTM)" in Jouko Laaksamo and Jukka Louhivuori (eds.): Proceedings of the First Internationsl Conference on Cognitive Musicology, Jyväskylä.

Turner, Mark (forthcoming): The Literary Mind, New York: Oxford University Press.

Hallgjerd Aksnes
Institutt for musikk og teater
Universitetet i Oslo
Postboks 1152, Blindern
0317 Oslo, Norway
Tlf: +47 22 85 47 64 Pr.: +47 67 54 01 05

Comments are welcome to the author at hallgjerd.aksnes@imt.uio.no

Back to the Online Metaphor Center.