Published in Organised Sound 4(3): 151-152 1999 Cambridge University Press.

Music as Mirror of Mind

by Laurie Spiegel


Hierarchical and other historically dominant models of perceived musical organisation are increasingly inapplicable to new musical processes and repetoire. The next major paradigm shift for music models, structures and concepts, perhaps comparable in importance to that in which polyphony gave way to homophony, may be a shift of emphasis from means of acoustic production and the nature of sound per se to new musical models based on psychoacoustics, cognitive studies and subjective auditory experience.


We may be drawing to the end of the period during which hierarchy is the dominant model. Although hierarchy will always be a useful and perhaps essential construct for conceptualizing certain aspects of music, it may soon be relegated to a position underlying rather than dominating our thinking, much as the concept of sequence has by now long been a fundamental but relatively low level component of our musical thought, taken for granted as essential but not often the center of analytic attention or terribly useful as a generative construct.

When very few entities exist, they may easily be individually perceived, each being an object of attention in itself. The first horns or shells, making one awesome sound at a time, must have been overwhelming with no multiplicity needed at all to increase the effect. A group of such sounds inevitably organized into a monophonic series will find attention focused on each sound of the sequence in its turn. Complexity takes a quantum leap when more than one such series is played at a time, forming parallel streams that may be heterophonic, contrapuntal or harmonic. Beyond just a few such simultaneous streams and successive sequential stages, the mind merges and groups them to reduce their number (Miller 1978). We tend to perceive, create, or impose a perception of groups within groups, structures nested within similar structures: hierarchies.

Eventually though, if density continues to increase, even hierarchy cannot accommodate the degree of multiplicity, density and variety with which our ears are confronted. Then what? What structural paradigm comes after point, line, plane made of parallel lines, and nested n-dimensional nodes? We either focus individually here and there within the mass, filtering out the weaker signals (Shannon 1948, Pierce 1961, Spiegel 1997), or our minds step out one step more removed, and group the entire mass into a comprehensible textural whole (Moles 1968, Miller 1978).

When first applied to music, the concept of hierarchy must have been experienced as a major breakthrough solving many problems. It provided a new and radically more effective way to deal with the simultaneity of multiple, multidimensional and multipart sonic streams. Eventually the modern orchestral score evolved, an ultimate culmination of hierarchical modeling, with hundreds of individual sound sources and their sonic parameters organised onto parallel staves grouped in turn into parallel systems of staves. This model allowed the accommodation of what all had in common (pitch, rhythm, tempo, etc.) as well as of what was unique to each (articulation, dynamics, fingerings, etc.). Carried within this complex multidimensional hierarchic sound source was a music built of nested levels of rhythm, both dynamic and harmonic, within multipart multilevel large scale forms, and common in nested figure-ground formats from concerto to accompanied song.

A great shift culminated in the era of Bach. Before the Baroque, perception was sequence-dominated (i.e. contrapuntal lines), and afterward, perception focused on simultaneity (chords). We can and do commonly perceive both of these dimensions at once, but not always with equal focus. In addition, the dimension of orchestration has been awesomely expanded from Bach's time to our own. To meaningfully ingest all of these perspectives, we may switch between these sonic views as though rotating a cube to see it from each side in turn, focusing attention on lines here, harmonic progressions there, and large scale transitions of amplitude or timbre at other times. We flit between multiple perceptual modes, but our minds will also tend to merge and group the sounds, perceiving a totality on some single central inner hearing ear. This Cyclopean (Julesz 1971) ear is the central single "ear" in which we hear with the most profound feeling.

Much of the new music we hear today is not organized along hierarchical lines at all. It may be built by progressive layering of unrelated elements, or extrapolation or permutation. There is often neither dominant melodic line nor supportive accompaniment, let alone one built of hierarchical subgroups such as orchestral sections, nested levels of harmonic rhythm or groups of tracks on tape. Many current works make little use of perceptually separable elements, such as could be identified by listeners as distinct voices, parts, channels, chords, motives, movements or dimensions. The likes of such standard multilevel forms as sonata and rondo are no longer used nor are comparable new forms being invented.

New standard processes are, however. It is already common that musical movement is built within a single complex sound or texture, by variation of amplitudes or harmonic spectra. Use of the concept of the "note" is now an option, rather than a necessity, for music making. Starting with the 1960s advent of voltage-controlled analog filters, we've seen a shift from additive to subtractive techniques, from accretion to filtration, from building up to carving down. If accumulation is the thesis and removal the antithesis, then the synthesis is synthesis in both senses of the word: the making of exactly what is desired or imagined, neither more nor less, and this is where technology seems to be taking us now.

In a music not built of many small units but sculpted from within or generated as a large sonic whole, is there still any place for structures such as hierarchy that evolved for the purpose of resolving otherwise unworkable multitudes of components into coherent unitary structures?

This move away from hierarchy is especially strong among electroacoustic works because they need not rely on production techniques involving multiple disparate sound sources, and because they use technologies that permit visual representation to be based on aspects of music other than the means of sound generation. Visual representation in electronic media often functions solely as a conceptual aid to the creator during the designing of a work, rather than, as in the past, providing detailed instructions to other individuals as to how and when to manufacture each of the work's acoustic components.

Possibly foremost in attractiveness, yet still relatively unexplored among non-traditional approaches, would be musical models based on the way human perceptual and cognitive mechanisms actually receive and interpret sounds (Deutsch 1981). Such musical representations would be designed by working backwards from the nature of human hearing, instead of forward from the source of sound . As more is learned about the way we ingest, parse, and experience sounds, the technical rubric on which our creative concepts take form would be designed for optimal plug-in compatibility with our music’s intended destination, the receiving processor, the ear and mind. Such representations and models might have less to do with any previous concept of musical structure than with the knowledge of aspects of self still sketchily understood at best. Music itself would push such research forward, propelled by the need to better understand its destination, instead of being focused on problems of sonic source.

This path could be paved by the exploration of models couched in neglected vocabularies of subjective experience, enlightened now by psychoacoustic and cognitive research. This would be music optimized for the receiver, the decoding end of the chain, substantially freed from the types of bias and constraint assumed to be inherent in all earlier means of origination.

What kinds of interfaces to sound might we want for this model? What can we make? Will we be able to move our hands through the space around us, shaping the sounds as we listen, roughing them up here and smoothing them there, and pushing and pulling areas of sonic fabric up, down, toward and away, together and apart? What might be the variables with which we interact? In what dimensions might we move?

To the extent that the auditory cortex and its feeder mechanisms and post-processing buffers do in fact turn out to operate as hierarchically arranged structures, such hierarchical models would of course remain in our educations, analyses, visual representations, computer models and concepts of musical structure. Conversely, to the extent that we can free ourselves from such presupposed legacy structures to discern without bias our own processes of response, new sonic vistas and new depths and realms of musical meaning may await.

As calculus introduced to mathematics in the 17th century a previously unimagined way to study motion and change, music deserves, and may finally obtain, a representation and conceptual model designed for the structure of how music is experienced, not just how its components are manufactured.

[Post script added March 19, 2007: This means "Pay close attention to what you feel as you experience music, honestly and with an open mind, and to what's going on in the music you're feeling. Disregard how the music is produced or constructed."


Deutsch, D. (ed.) 1982. The Psychology of Music. Orlando: Academic Press.

Julesz, B. 1971. Foundations of Cyclopean Perception. Chicago, IL: The University of Chicago Press.

Miller, J. G. 1978. Living Systems. pp. 103-5, 121-65. New York: McGraw-Hill.

Moles, A. 1968. Information Theory and Aesthetic Perception. pp. 74-5 (trans. J. E. Cohen). Urbana: University of Illinois Press.

Pierce, J. R. 1961. Symbols, Signals and Noise: The Nature and Process of Communication. New York: Harper & Brothers. Reprinted unabridged and revised as An Introduction to Information Theory: Symbols, Signals and Noise. New York: Dover Publications, 1980.

Shannon, C. E. 1948. A Mathematical Theory of Communication. Bell System Technical Journal. Reprinted in Shannon and Weaver, Mathematical Theory of Communication, University of Illinois Press, 1949.

Spiegel, L. 1997. An Information-Theory-Based Compositional Model. Leonardo Music Journal 7: 89-90.

- Laurie Spiegel
New York City, NY

Copyright © 2000 by Laurie Spiegel. All rights reserved.

Laurie Spiegel's writings page | Laurie Spiegel's home page