jwilzewske: April 2009

Wednesday, April 29, 2009

Stochastic Incipits & Continuants: Sonic Ontology in Epstein’s ‘Bloom’

B loom’ is about Time or, rather, the suspension of our conventional sense of musical Time… The solo English horn plays bass notes from which chords bloom and rise up through the ensemble. ‘Bloom’ is a series of short sections, each of which is characterized by a specific texture. The English horn is either the prominent solo voice in the texture, or it becomes the lead voice in shaping the sound of the ensemble… ‘Bloom’ ends with a giant canonic chorale, which gradually reduces texturally until all that’s left are the double-reed instruments. This gradually thins out until the last thing we hear is the solo English horn—an ending that reminds us of—but isn’t exactly like—the opening.”
— Marti Epstein, program notes.

E ven the sound of the word ‘Bloom’ is evocative—there’s an onomatopoietic rightness to it that we eagerly concur with after, say, only a minute or two into the piece. The world premiere of this composition by Marti Epstein for English horn and wind ensemble was performed by the Boston Conservatory Wind Ensemble on Tuesday evening this week, in the small recital hall of Boston Conservatory at 31 Hemenway, with Eric Hewitt conducting.

R obert Sheena, principal English horn with the BSO, was the featured soloist. Besides his BSO duties, Sheena is presently on-faculty with the Boston Conservatory, Boston University, New England Conservatory, and the Longy School of Music.

S heena’s treatment of the blossoming, organic phrases in ‘Bloom’ was expansive yet terse. In a work that is permissive of prolixity by the soloist, Sheena’s phrasing and dynamics were elegant, suggestive more than demonstrative—economical. The polyphonic ‘builds’ that Epstein refers to in her program notes—swelling and shrinking, ebbing and flowing—are a novel kind of canonical pattern. Rather than canonical forms that have more or less crisp, decidable beginnings and endings or cyclical structure, the canonical ‘builds’ in ‘Bloom’ involve non-deterministic propagation of musical thoughts or concepts between the soloist and the various sections in the ensemble.

T ime-lapse photography/cinematography of a bed of flowers opening and closing may have been the nominal motivation that inspired Epstein to compose this piece, but the physical and logical correspondences/parallels do not stop there. For example, there are gene-expression and receptor-signalling processes in biology that have closely analogous spatial features and time-oriented kinetics. So, as I sat enthralled with the performance last night I began to think, not so much about the provenance of the imagery that Epstein gave us, but thought instead of meta-concepts—how this piece is constructed, why these waves of sound propagating from section to section work the magic that they do [on us humans? on birds, dogs, whales, cicadas and other musical species?], and the ontology of the concepts underlying the musical structure and processes of ‘Bloom’.

T he English horn, of course, is fundamental to this ontology or world-view. Yes, Sheena’s realization of the part is economical, eschewing anything more than the minimum necessary, never to ‘comment’ on anything (any concept implied by the score) that should be self-evident or obvious to any thinking performer/listener. His refraining from dilatory interpretations makes Time itself more precious—a feature that is consistent with Epstein’s reverence for Sibelius ... Time made to stand still, Time stilled by Music.

B ut my saying this may seem to give too much emphasis to Sheena’s particular choices or Hewitt’s conducting or other performance details. Rather, the preciousness of Time inheres in the composition that Epstein has devised, and in her choice of English horn for the main character/protagonist.

W hat I mean is, not all instruments’ characters are created ‘equal’. The English horn’s distinctive timbre is a lucky endowment—it has an architecture (and a high double-reed back-pressure) that enables it to be ‘terser’ than most instruments are disposed to be.

A writer [musician] is someone for whom writing [musicking] is more difficult [risky; momentous; tight-rope-like] than for other people.”
— Thomas Mann.

T he ebbing and flowing of the canonical structures—the contagious crescendos and decrescendos among the sections—reminded me of the aurora borealis… its dancing, shimmering colored curtains of light in the northern night sky… Reminded me, too, of the Belousov-Zhabotinsky reaction or other autopoietic/morphogenetic integro-differential Reaction-Diffusion equations in chemistry... Or, if Time is standing still, then maybe it’s like the birth of a nebula—something celestial, galactic. Yes, ‘Bloom’ seemed to me more celestial than botanical…

T he stochastic (non-deterministic) quality of the incipits’ beginnings and the stochastic dynamics of their propagating to other regions of the ensemble as they play—the uncertainty of the stochastics decreases our attending to the incipits themselves and increases our attention to what I would call ‘continuants’. ‘Continuant’ is a fancy term used in philosophy of language and in informatics, but it really signifies something pretty commonsensical: a sub-class or descendant of a higher-level (and prior/earlier) concept or utterance. Continuants inherit most of the attributes (morphology, behaviors, capabilities, tendencies, processes, etc.) of their higher-level precursor incipits [of which the English horn part is emblematic in ‘Bloom’], but they have autonomy of their own and are influenced by their own contexts and histories.

T here are links at the bottom of this CMT post on informatics ontology engineering, for you composers who want to see whether ontology architecture may have something to offer you as a compositional framework. Your intentions and results may be quite different form Epstein’s, of course. But the ontology engineering framework and methods might well be amenable to your aims, or might at least offer you some innovative inspiration...

T he ‘Basic Formal Ontology’ (BFO) in informatics theory consists in a series of sub-ontologies at different levels of granularity. The ontologies are divided into two varieties: SNAP (‘snapshot’) ontologies, comprehending continuant entities such as three-dimensional enduring objects, and SPAN ontologies, comprehending processes conceived as extended through, enduring, and spanning time. BFO incorporates both three-dimensionalist and four-dimensionalist perspectives on reality within a single framework. Interrelations are defined between the two types of ontologies in a way which gives BFO the facility to deal with both static/spatial and dynamic/temporal features of reality. Each SNAP ontology is an inventory of all entities existing at a time. Each SPAN ontology is an inventory of all the processes unfolding through a given interval of time. Both types of ontology serve as basis for a series of sub-ontologies, each of which can be conceived as a window on a certain portion of reality at a given level of granularity.”

I n summary, one of the most novel and beautiful aspects of ‘Bloom’, to me, upon hearing this premiere performance, was these cascades of continuants—the waves upon waves of consequents: more and more [temporally- and spatially-disperse logical ——] consequents, piling up and washing over the antecedents/incipits laid down by Sheena’s English horn. The more loquacious instruments inject their colors/flavors—they could not do otherwise, any more than Sheena could prevent the English horn from being its terse self. And the organism, or flower bed, or aquatic algae bloom, or transcriptome, or receptor-signalling network, or galaxy, or whatever it is—is big enough and generous/tolerant enough to accommodate all of it. Heartwarmingly, beautifully so.

W e listened to Epstein’s new nebula blooming with an IMAX-type wonderment—floored by discovering a wideness of musical ‘peripheral vision’ that we didn’t previously know we had, awed by a deep-space ontology we hadn’t known was there. And then we stepped out of the recital hall and into the warm, starry Boston springtime night…

M arti Epstein has completed commissions for the Munich Bienniale, the CORE Ensemble, the Pro Arte Chamber Orchestra, the Foxborough Music Association, the Fromm Foundation, and Guerilla Opera, among others. She has been a fellow of the Tanglewood Music Center and a composer in residence at the MacDowell Colony. She is Professor of Composition at Berklee College of Music and at the Boston Conservatory.

Tarantula Nebula, blooming; photo © NASA

Tuesday, April 28, 2009

Outer Surfaces of Seemingly-Simple Things: Lansky’s Complexity-Hiding

I ’m no longer comfortable with the ‘network’ model [of music structure, including synth and programmatic/aleatoric elements that are not under total human control]. There was an implicit assumption in my [earlier] topology that technology was acting as an equalizer between people... I [now] think that’s a slippery assumption. While I still think that the topology I described is interesting and suggestive, I’ve become a bit more skeptical … I regard listening to a piece of music or reading a book as an intensely interactive activity, a communication between minds. I’m a little more hesitant these days about elevating the ‘sound-giver’ too much, in that there are a lot of blurry boundaries between that node and the listener. I don’t want to regard the ‘sound-giver’ as a node with equal weight to the performer or composer. ‘Instrument builder’, however, is another matter. The design of an instrument will very often involve compositional decisions.”
— Paul Lansky, interview with Jeff Perry, Perspectives New Music 1995.

T he way to make something seem simple is, often, to hide the complexity that it contains—to make a ‘black box’ out of it. (In my day-job, I am currently working on a healthcare information exchange architecture. As I do so, sometimes at night, naturally I listen to music. Inevitably, odd reveries and correspondences between composing music and writing software percolate up. Whether these are (will be—) spurious or useful, I cannot say. But to preserve the memory of them, I sometimes capture them here, in a CMT post…) [Returns to writing code, listening to Paul Lansky CD.]

F or me, Paul Lansky is the go-to guy… a true master of complexity-hiding. Paul Lansky turns 65 in a couple of months. Once a student of George Perle… student, too, of Milton Babbitt (with whom I myself attended lectures back in the early 1980s) and Edward Cone, Paul is currently professor of composition at Princeton. He has for forty years been prominent in electroacoustic and computer music, including language development for algorithmic composition (Real-Time Cmix). [‘Enjoying’ is different from ‘revering’. I look across the room into the darkness and wonder which it is that I am feeling as the Lansky CD continues to play.]

[Brentano Quartet, Lansky, ‘Ricercare’, 7.6MB MP3]

[50-sec clip, Paul Lansky, ‘Idle Chatter Junior’, 2.0MB MP3]

T he phrasing of the ‘Idle Chatter’ creates a sonic ‘surface’ that’s built-up out of microsound samples—a tonal ‘whole’ that’s astonishingly comprehensible, even if the objects out of which it’s constructed are not… The ‘Ricercar’ montage resembles—to me—the object-code output of a software object-oriented compiler, with object class hierarchies and inheritance and message-passing among instantiated class members…

I have enjoyed Lansky’s compositions for years but now begin to try to analyze why that is—why is it that they seem to me to be so deliciously elegant? [This extraneous thought occurs to me as I examine some Java debugger output in my supposedly day-job now turned to night…]

I should look—look under the ‘veneer’ of Lansky’s music—I must look carefully to see what complexity-hiding and abstraction methods his writing and microsound collocations implement. Maybe you will, too…

A s with complexity-hiding methods in software in general, Paul’s musical complexity-hiding composition techniques assure that ‘users’ (performers; listeners) are not unduly burdened by issues of internal music representation or execution. To a non-programmer, deliberate complexity-hiding may seem a somewhat ‘maternal/paternal’ stance [toward the performers; toward the audience], I guess. But to a software-developer/composer, complexity-hiding is not laden with questions of asymmetrical autonomy of different parts/players. It just represents a set of ‘engineering’/‘compositional’ choices, nothing more. There are philosophically and psychologically more parallels between a software-developer and a composer than most would care to admit...

F or example, to create a software functional specification, we make a process diagram that hides the complexity of certain parts of the operating system or services internals or end-user workflow from the developer of a particular subsystem, or hides the model-checking output from the process-modeller, or hides other things that are irrelevant to the authority and responsibility of other constituencies. To create a system technical design, we make an Application Programming Interface (API) that hides the complexity of software module internals from the system integrator and external application developers. It has been second-nature to do these things in structured software engineering for more than 30 years, but especially in the past 15. Why is it that music composition courses—even ones in computational music—neglect the applicability and relevance of these ideas to music? Probably the answers are that there are just (a) not enough people who are comfortable with and knowledgeable about the computational music software tools that have become available in the past 15 years and (b) not enough hours in the curriculum. If only they knew that inhaling that stuff could help them to make new, more novel, more beautiful, more elegant and surprising compositions! They would make more curriculum hours then, I bet…

M y impression of ‘Ricercare’ and other pieces is that they resemble so-called ‘autonomic’ computing (see links below)—systems that are [by definition] self-configuring, self-healing, self-optimizing and self-protecting, all mostly reflexive, without extensive human deliberation but instead carried out according to pre-defined policies. The road map culminating in autonomic computing architectures encompasses the following five levels of extent or ‘maturity’:

Basic: The product and environment expertise resides entirely in human minds, requiring detailed rehearsal, gestures, messaging, and consultation on even routine procedures.
Managed: Scripting and logging tools automate routine execution and reporting. Individual specialists review information as it comes in, to make plans and decisions.
Predictive: Early warning flags are raised as preset thresholds are tripped. The knowledge base recommends appropriate actions. The proposed resolution of events is leveraged by a centralized storage of common occurrences and experience.
Adaptive: Building on the predictive capabilities, the adaptive system takes action itself based on the situation.
Autonomic: Autonomous algorithm-based policies drive system activities such as allocation of resources within a prioritization framework.

T o allow system components, whether infrastructure or application software, to predict when they are on the verge of violating a threshold, ‘sensors’ must be built-in. Timers or code-instrumentation to monitor software runtime execution are built-in. [Similar monitor structures are present in the compositions of Paul Lansky—I am sure I am not crazy in inferring that they are there in these pieces I am listening to tonight.] In other words, the ‘adaptive’ maturity level of autonomic computing systems calls for the creation of a process that monitors resources, defines a systematic reaction to a set of indicators, and the automation of procedures that alleviate the underlying problem. For example, just as RAID arrays of fault-resilient disk drives can be configured to automatically ‘mirror’ a failed drive to a spare drive, members performing Lansky’s ‘Ricercare’ can be configured automatically to gracefully respond to a situation involving one member of the string quartet…

T he automatic aspect of the response is one guideline for determining whether a given system has matured to the ‘predictive’ level or gone beyond it to become ‘adaptive’. Lansky’s music is at least ‘adaptive’; it remains an open question as to whether some of it goes yet further, into the realm of fully ‘autonomic’…

T he threshold levels defined as part of predictive systems management are not discarded as the [software-engineering/music-composition] ‘environment’ matures. These metrics are captured and analyzed, with the resulting action expressed as ‘recommendations’ and [execution-runtime/performance-practice] ‘options’. As the [system-managers/ensemble-members] gain confidence in the ability of the ‘system’ to monitor and flag events, revised response levels are defined, and first-line ‘defense’ activities are enacted by the system. In this way, the services can be considered to be ‘maturing’ to the adaptive level of an autonomic computing system.

A daptive’ and ‘autonomic’ levels of software systems maturity require systems’ infrastructure to be integrated, to allow ‘smart’ components to recognize when an impending challenge/risk is going to affect them, or where an escalation in technical demand puts their ‘service level agreement’ (SLA) at risk. Dare I say that the same thing inheres in serious music—serious music both composed and improvised? In software, adaptive components can take advantage of alternate infrastructure to ensure uninterrupted operation and uncompromised performance in conformity with the SLA. In music, adaptive components like Lansky’s accomplish the logical equivalent. (What I mean is, after all, are there not implicit SLAs in music? Both in good composition methods and in good performance methods?)

A daptive tools … such as balancing the ‘least recently used’ (LRU) and ‘least frequently used’ (LFU) pages of a memory ‘cache’ or a persistent-message ‘store’ gives better performance than relying on either LRU or LFU strategy alone. It becomes possible to set policy for dynamically responding to situations as they arise. For example, where a key line of motivic development in tonal music requires more resources or more emphasis to achieve the cognitive impact intended, it might prove better to temporarily pull resources from one task to assist a higher priority task than impinge on the service levels defined for that server. Sometimes I get the feeling that Lansky’s minimalist compositions are using a combined LRU-LFU regime to communicate as much as he does, using only relatively modest materials…

Problem management
System/motif availability
Security
Performance and dynamic/adaptive capacity management

T opology hiding: It is possible to hide (e.g. encrypt or mask) information when ‘domain borders’ (administrative or technological or semantic/expressive) are crossed. Hiding of rhythmic or flow-of-control detailed information is different from hiding structural or topological information. Topology hiding is something Lansky does very elegantly, especially in his sampled montage works…

I f a software hacker wants to attack an enemy ‘target’, the target usually must be visible or exposed. To defend a target, you need to obscure it from public view, obscure it from hackers and would-be intruders. By hiding an object, you reduce the attack ‘surface’. In computers, only the part of a progam that is accessible to an attacker can be a target of attack. The pieces of code that’re exposed to public view are usually called APIs (Application Programming Interfaces). The ‘attack surface’ is the union of the code, the interfaces, the services, the protocols, and the processes that are exposed to the system’s users. In software security design, you analyze the attack surface and minimize that surface, sometimes by putting a ‘wrapper’ around it, or placing it in a ‘package’ such that only those entities who have access to the package are exposed to the variables that are within the scope of that package. I think I hear more correspondences—between what Lansky is doing acoustically, and what I am doing writing this Java code here on this laptop…

Is this feature really necessary [in this string quartet movement]? If no, it should be hidden by default.
Is it necessary to offer this feature remotely? If yes, then determine what locations it needs to be accessed from and via what media.
Which users or user-types have ‘need-to-know’ justification for accessing this feature? [The viola? The cello?] Legitimacy of usership needs to be authenticated.
What privileges are required to use this feature? If it needs elevated privilege, how long [How many bars?] does the elevated privilege or access persist before it needs to be renewed?

V irtualization’ is a term that I think deserves to cross-pollinate to music theory and composition, from computer engineering. In computer science, ‘virtualization’ is a design philosophy where the operating environment abstracts the resources involved in its realization. Much virtualization involves abstraction through isolation and ‘hiding’. It’s not only a way to hide the fact that resources/possibilities are limited, but also a way to hide the fact that resources/possibilities may be arbitrarily large, far greater than what is apparent in minimalist evidence from what's happening in the current performance.

N on-disclosure cuts both ways: withholding details may create a credible impression of scarcity, or it may foster an impression that resources may be boundless.

H olger Schmidt’s paper in the Massacci-Redwine-Zannone book (link below) addresses encryption accomplished by a combination of ‘relational renaming’ and ‘hiding’. Disclosures not only happen by way of a process that unfolds publicly; they can also happen via a process’s failing to unfold because of deadlocks, livelocks, or balking, where the failures can imply information about the respondent—especially revealing the limits of virtuosic technical capacity. So Schmidt develops a pattern-based method for encryption that protects against unwanted disclosures, including preventing disclosures through slips or failures. Analogous phenomena (deadlocks, etc.) occur in music—in the score [deliberately], not just [inadvertently] in performance. [Decides to play around with that sometime in the future, in my own composing…]

I n music composition as in computing infrastructure, I look to RT-Cmix and Rubato and PWGL and other algorithmic composition software tools to mature along autonomic computing lines, including the ability to predict and adapt can ensure that the most important systems/motifs are ‘up and running’. Complexity-hiding in Paul Lansky’s writing: great examples to learn from and excellent use-cases to drive software requirements specs for autonomic composition software. Thank you for your interest/indulgence during the foregoing—computer/music fusion reverie with an uncertain [but decidedly autonomic] future.

Paul Lansky page at Princeton
Perry J. The inner voices of simple things: A conversation with Paul Lansky. Perspectives New Music 1995; 34:40-61.
Brentano String Quartet website
Bel B, Vecchione B. Computational musicology. Computer Humanities 1993; 27:1-5.
Cope E. Virtual Music: Computer Synthesis of Musical Style. MIT, 2004.
Cope D. Computer Analysis of Musical Style. Oxford Univ, 1991.
Dibben N. The cognitive reality of hierarchical structure in tonal and atonal music. Music Perception 1994; 12:1-25.
Longuet-Higgins H, Steedman M. On the interpretation of Bach. Machine Intelligence 1971; 6:221-41.
Marsden A, Pople A. Computer Representations and Models in Music. Academic, 1992.
Massacci F, Redwine S, Zannone N, eds. Engineering Secure Software and Systems: First International Symposium, ESSoS 2009 Leuven, Belgium. Springer, 2009.
Miller D. Going with the grain: Granular synthesis. Electronic Musician 01-OCT-2008.
Narmour E. The Analysis and Cognition of Melodic Complexity. Univ Chicago, 1992.
Overill R. On the combinatorial complexity of fuzzy pattern matching in music analysis. Computers Humanities 1993; 27:105-10.
Talukder A, Chaitanya M. Architecting Secure Software Systems. Auerbach, 2008.
Todd P, Loy D. Music and Connectionism. MIT, 1991.
Wallin N. Biomusicology. Pendragon, 1991.
Virtualization page at Wikipedia
Autonomic computing page at Wikipedia
Intl J Autonomic Computing
IBM autonomic page
ACE autonomic SDK
SARDES JADE-based autonomic system

Saturday, April 25, 2009

Australian Chamber Orchestra, Epitome of ‘Embodiment’

O ne of the foremost composers of the French Baroque operatic tradition, Rameau is often cited for his struggle to steer lyric tragedy away from its strict Lullian form, inspired by spoken tragedy, and toward a more expressive musical style… In his compositions, Rameau tried to highlight music’s potential for dramatic meanings. But his listeners, who understood lyric tragedy to be a ‘poetic’ rather than ‘musical’ genre, were generally frustrated by these attempts. In fact, some described Rameau’s music as ‘monstrous’—using an image of deformity to represent the failure of reason and communication.”
— Charles Dill, 1998, Professor of Musicology, Univ Wisconsin-Madison.

L isten to that, will you?” the person next to me could not help but say to me during the intermission. “Have you ever heard an ensemble so in-tune? or so ‘tight’, rhythmically? Each section is playing like they were one person!”

Y are right. I couldn’t disagree,” I said. “And to see them pivoting and rotating left and right, each by as much as 90 degrees, to see each other and physically gesture directly at each other! Their knee-flexing and expressive leaning and swaying... Their stretching of the dynamics, decibels quieter (with mutes on) and louder, beyond the usual interpretation of the as-written dynamics... How intriguing to see instrumentalists do this! It’s like they are opera singers, singing to each other.”

T he Australian Chamber Orchestra performance in Kansas City tonight was exceptionally good. And the exchange above occurred after the Haydn and Mozart before they got to the Suite from the opera ‘Dardanus’.

E nglish pianist Paul Lewis (who trained with Alfred Brendel, among others) joined the Orchestra for the Mozart Piano Concerto No. 12, K. 414... “A lush yet ‘crystalline’ tone... rhythmically ultra-synchronized with the ACO members... Amazing!”

Haydn: Symphony No. 44 ‘Trauer-Sinfonie’
Mozart: Piano Concerto No. 12 in A Major, K. 414
Rameau: Suite from ‘Dardanus’ (1739, 1744, 1760; RCT 35)
Haas: String Quartet No. 2, Op. 7, Monkey Mountains Suite (‘Zopicich hor’)

T he Rameau after the intermission was especially dramatic. Sans recorders and flutes but equipped with antiphonally-distributed horns, we were treated to six short movements with wonderfully rich acoustic colors and textures. Oboes cohabited with violas, the bassoon communed among felicitous cellos, and Lions laid down peaceably with Lambs. The textural and spatial novelty is not gimmicky—it’s innovative, aesthetically defensible in the context of Baroque sensibilities, and stimulates more drama and interest than the music might otherwise do. Very cool. And an excellent way to illuminate Rameau’s complex sense of the interplay between the narrative arc of the libretto and the associated arc of the instrumental music for the opera.

R ichard Tognetti became Artistic Director and Lead Violin in 1990, and under his leadership the ACO has embarked on ever more flexible ensemble configurations: small chamber group; small symphony orchestra; electro-acoustic chamber group; and so on. In ‘live’ performances, only the cellists are seated. The energetic interactions that this deployment encourages are nice—the visual gestures amongst the artists tend to be more demonstrative than when ensemble members are all seated. The standing ensemble members and the antiphonal positioning of some of the instruments add substantially to the excitement of the experience for audience members—these directorial/production decisions were especially appropriate for the dramatic Rameau. Oboes, far far stage-left, their backs to the shell, behind the violins; horn, center-rear; bassoon, far stage-right. These parts speak to each other across the distance, while the strings play with tight precision, as though they were one! Not just stretched dynamics, but dramatically accentuated spatial depth. Uncanny!

T hese decisions and variations also help tremendously realize and vividly accentuate Rameau’s physicality. These directorial/production decisions draw attention to the metaphorical ‘embodiment’ that Golan Gur and other Rameau scholars have written about. ‘Embodiment’ in the field of cognitive science refers to understanding the role of an agent's (character’s; performer’s) own body in its situated interactions with other agents in context—the role that the body has in the agent’s comprehending the situation and those interactions. For example, Dardanus “can’t stand to see” his future relationship with Iphise reduced to questions of pre-arranged nuptials and political alliances—his physical experience of standing, or failing to remain standing, or failing to endure the situation—is an example of cognitive ‘embodiment’, of physicality, which colors Dardanus’s cognition of the situation. His impulsive decision to go kill the monster could hardly arise from anything less than a powerful emotion, mediated by ‘embodied’ cognition.

A CO’s production values are fabulous throughout, but this Rameau really captured my imagination… Richard Tognetti is the Dardanus character! We are so not accustomed to instrumental players physically ‘acting’! These are not a Baroque ensemble performing a Suite ‘from’ Rameau’s opera; they are playing and moving physically on-stage as though they are the opera singers.

I n Greek mythology, Dardanus was a son of Zeus and Electra. The name ‘Dardanus’ in Greek (Δάρδανος) is forboding: it means ‘burnt to a shrivet’, from the verb δαρδάπτω (dardapto)—to grind down utterly, to burn up completely, to vaporize. In the Rameau opera, Dardanus is feuding with King Teucer. Teucer has promised to marry his daughter Iphise to King Anténor, done deal, so much for women’s rights. King Anténor is sucking in his attempts to defeat the monster that’s terrorizing Teucer’s kingdom. Dardanus and Iphise meet, through the probably-superfluous intervention of the magician Isménor, and fall hopelessly in love. An ‘army of one’ and now properly/amorously motivated, Dardanus then single-handedly defeats the terror-monster, saving Anténor’s life. Teucer and Dardanus make peace, and Dardanus finagles Teucer’s release of Iphise from her betrothal to Anténor. Anténor’s happy enough just to be alive, so Dardanus and Iphise marry and live happily ever after, so far as we or Rameau know. So, embodiment, yes; burnt shrivets, no—at least not in this opera or this Suite for chamber ensemble.

Ouverture (Lent)
Premier air (Grave)
Ritournelle
Premier menuet; Deuxième menuet en rondeau
Air de triomphe
Bruit de guerre pour Entr’acte

T he piece by Pavel Haas—Tognetti’s arrangement of String Quartet No. 2 for the larger ACO ensemble—pushed the dramatics up a notch. A dark, dissonant work, filled with punctate staccatos and percussive effects; very ‘cinematic’. Once again, the rhythmic crispness of the ensemble lent extra energy and excitement to the endeavor. The synchronicity of each section was astonishing—every player’s vibrato starting simultaneously, executing with identical width and periodicity, and ending simultaneously; every sautille, identical. To paraphrase Butch Cassidy & The Sundance Kid, who are those guys!??! ACO’s unprecedented dramatics and beautiful physicality brought the audience to its feet.

I f you haven’t previously encountered it, you may like to look at Charles Dill’s book on Rameau from about 10 years ago. Also, as relates to the inter-relationships between text and music, have a look at Peter Kivy’s new book. Several sources regarding ‘embodiment’ and theory of aesthetics and cognition are also included in the links below, for your interest.

The Australian Chamber Orchestra was founded in 1975 in Sydney. Their performances have been very popular at home (more than 10,000 Australian subscribers to their concert series) and heavily attended abroad, too. The ACO’s repertoire ranges from baroque and classical pieces to 20^th-Century and new music. ACO has in recent years commissioned quite a few compositions. The financial sponsorship of the Australia Council for the Arts and the Australian Music Office have contributed significantly to ensuring ACO’s success and sustainability. Would that ACA and AMO examples were emulated by other countries, particularly during the current economic down-turn! (As Tognetti said to the audience, maybe in a more affluent time ACO will not have to conserve on expenses so much and once again tour with their percussionist. That way, Julian Thompson will not need to treat the body of his cello so severely/pugilistically, to create the knocks, bangs, paper-in-the-strings buzzes, and other percussive effects called-for in works like the Haas.)

[ACO, Surfing Chamber Music ‘Musica Surfica’]

Wednesday, April 22, 2009

Orchestrating Virtuosic Saw

Saw ad, Popular Science magazine, March, 1948

T he visit I had to Ripon, Wisconsin, over the weekend was notable for many things—among them an excellent amateur duet, by a violin student accompanied by her father on a musical saw. Recently, composer Ken Ueno has been interested in experimenting with bowed metal bars and sheets. This fortuitous (?gratuitous) coincidence led me to get the Stanley cross-cut saw out of my basement and explore what can be done.

A ny CMT readers who are interested in composing for musical saw should be aware that, as with any instrument, the dimensions of the material determine the pitch range that’s achievable. Almost any saw can produce musically usable sounds, but saws of 26 inches (66 cm) or longer are best. Such a length insures that the ‘sweet spot’—the lengthwise portion of the blade where a bow contacting the flat edge (not the side with the teeth) can set the blade into stable, resonant vibration—is relatively large... several centimeters long. Such a length also insures that flexing the blade in an ‘S’ shape to produce the desired pitches will not place excessive demands on the player’s left-hand strength.

T he majority of commercial woodworking handsaws are tempered carbon steel with a thickness of about 0.040 inches (1 mm). Saws that are specifically made for musical purposes (see links below) may be considerably thinner than this. The thin material allows the S-shape flexion to be achieved with less muscular effort. The preparation of the saw shape and tension before commencing the bow attack can be done more easily, compared to an ordinary commercial woodworking handsaw. The resulting intonation, vibrato, and overall musicality are undoubtedly better—with less performance risk and stress for the performer. (My own views below are informed by only cursory experience with a Stanley woodworking saw, motivated by curiosity about such an exotic instrument. Please accept these comments with a ‘grain of salt’; your own mileage may vary...)

[30-sec clip, Jim Parker, ‘Midsomer Murders Theme’, 0.9MB MP3]

[50-sec clip, Natalia Paruz, Bach/Gounod, ‘Ave Maria’, 1.6MB MP3]

T he majority of musical saws are capable of delivering pitches roughly in the range of a viola. Some 30-inch and larger saws are available and are labeled ‘bass’ or ‘baritone’, but most musical saws are evidently around 26 inches—‘tenor’ instruments. As such, it seems reasonable to score the saw using the alto clef—same as viola. The range in the illustration below represents safe limits for the Stanley hardware-store saw.

I f you decide to purchase a special ‘musical saw’, the backside edge will have been polished and prepared by the manufacturer in such a way as to be suitable to receive bow contact. I examined my Stanley crosscut saw with a magnifying glass and noted that there were some burrs along the backside edge—tiny, sharp projections that could easily damage or cut bowhairs. I removed these with a few strokes of emery cloth, followed by fine steelwool.

O rdinary amounts of rosin on the bow seem to work just fine with the saw. My own brief experiments were with a pernambuco violin bow—too delicate, really, for the 0.040-inch Stanley blade. My suspicion is that a cello bow would offer the stiffness needed to properly articulate the notes on a saw.

H olding the handle of the saw between the legs and jiggling the left leg to produce vibrato in the S-flexed saw is pretty easy. Flexing the blade with the left hand, however, is tricky and quite tiring unless you affix a handle through the hole near the tip of the blade—a technique referred to by saw players as a ‘cheat’. Acknowledging the deprecation of the ‘cheat’ by serious players and the visual aesthetic detractions of the ‘cheat’ for the audience, it seems to me that practical compositions for saw will limit the duration of saw passages to maybe one minute or less at a stretch. After a rest the saw player will have recovered the muscular control to proceed with a subsequent passage.

S urely there may be some virtuosic saw players for whom extended multi-minute passages would present few problems, and for whom the arduous gauntlet-running at prodigious tessitura would afford a proof of their freakish virtuosity. But a composer who wants to score for saw because of its wonderful, peculiar, Theremin-ish timbre ought, I think, to assume that such saw virtuosi are scarce as hens’ teeth. The typical saw artist will be a normal string player who is partial to odd and exotic things, and that sort of person will not care to (a) risk over-stress or injury to their left hand and its musculature, by the weird movements required for saw flexion, or (b) engage in extended practice and rehearsal on the metallic interloper, forsaking their belovèd stringèd axe. In other words, you’ll want to write modest stuff that’s playable without too much trouble.

T he technique of saw is similar to bowing a viola da gamba—underhanded—except for the fact that you are bowing more or less vertically. Like a viola, the saw’s size is somewhat at odds with its pitch range—too large to be comfortably played under the chin, yet smaller than would be ideal for playing between the legs. But, also like the viola, this size-pitch mismatch is probably the reason why we have such a highly distinctive, beautiful tone-color in the saw.

C hromatic passages are pretty easy to perform. Intervals of a major or minor third or perfect fourth can be navigated with little trouble. Fifths and longer jumps seem (to me) to be exceedingly risky. Safe writing for saw would stick with melodic lines—careful attention to voice-leading and no endangerment of the gristle.

T he top end of the saw has a quite distinct ‘nasal’ quality, which can be exploited for short, plaintive effects. The lower range has an apocalyptic, austere, dark sound. If you don’t like that, well, tough. (In fact, I feel that the weird, impending catastrophe demeanor of the saw down there is part of its psychological charm. It’s like a wise, dour octogenarian who perennially and insistently warns everyone in the room that there will inevitably be more serious trouble in the Middle East, no matter what the topic of conversation might be. This ‘wild card’ can have a wonderfully unsettling effect on the other party-goers. Just so with the musical saw...)

T he vibrato technique is basically the same as on the violin, though you may wish to indicate on your score that the performer may vibrate somewhat slower and wider in the lower register. A cautionary notation might be useful, to instruct the performer about the effect that you are aiming for. A musical saw has—What? I am shocked!—a propensity to bring out the ‘ham’; a simple admonishment that ‘less is more’ and that you are looking for ‘subtlety that defies temptation’ should be sufficient.

P ortamento and glissando can be readily executed on the saw—so okay to write those in your score. Obviously, trills or fingered tremolo effects are out of the question. There are no double-stops or triple-stops—at least not without severed digits or ripped trousers. Remember that the performer’s genitalia are facing the sharp teeth of the cross-cut. They (and their significant other) will long remember you if your writing is excessively ambitious.

D o not write sixteenth notes or thirty-second notes for the poor, ‘flickted thing. Doing so will win you no friends.

A rtificial harmonics? Well, my adventures in trying to produce them stopped when my dear spouse told me to “Cut it out!” I can only say that harmonics are in there, and that they will come out to play if you coax them. Touching the blade below the “S”, as the blade heads toward your lap can entice some spooky, scritchy, ethereal stuff out. Hard to control, though, so you may want to choose your practice environment more judiciously than I had done, so as not to tax your loved ones’ ears beyond their endurance...

E levations of the bow—pretty much the same as the viola da gamba, except that the excursions are more vertical. Considerable bow-speed is necessary. Maybe a right-arm “break-away sleeve” for the concert performer would be a practical solution, albeit one that would promote general “hamminess”. For studio work, the player may prefer to wear a comfortable tee-shirt...

B owed tremolo—quite effective on the saw. It has a sharp, biting effect in forte and a silken, decrepit effect in pianissimo, due to the tragic tone color of the instrument.

P izzicati are more resonant on saw than on the violin—a feature that worked wonderfully for the daughter-father duet in the Wisconsin recital last weekend. As with harmonics, the higher pizzicati are usually best left to the violins. Unless, that is, you want the thematically desultory effect of plunked saw rubbing the violin noses in “it”—an entirely valid possibility (You choose!) but more histrionic than would suit most commissions. After all, you do not want your saw opus to end your career.

A djectives in Italian applied to the saw must agree with it in gender (sola or sole, not solo or soli). Although it is the only instrument of the steel family in the feminine gender, it has probably the most masculine tone—Ueno’s bowed steel bar affixed to a timpani head notwithstanding.

O ccasional short solo passages for the saw can be of telling effect for orchestral music, although recent works tend to focus on chamber ensembles. I am not sure whether this is due to the saw’s fall from grace since the 1950s—a time when there were many thousands of saw players throughout North America.

T he solo repertory for saw is scanty—although transcriptions can be effected straightforwardly. Berlioz’s “Harold (‘The Saw’) in Italy”, for example. Or try William Walton or Walter Piston viola concerti.

L ike the viola, the saw is easily obscured or ‘covered’ by other instruments. And saws easily cover and confuse other saws. With that in mind, you will want in general to write open harmonies. Your orchestration should assiduously avoid unisons or minor seconds in the saw section. Do not, I repeat, do not write ‘tutti’ for saw.

I n the olden days there may have been up to eight saws in a symphony orchestra. They used to sit at the conductor’s right, upstage of the cellos, who would in this configuration be out of toothy-edged harm’s way should any slips occur amongst the saws. This was acoustically non-optimal insofar as the blades of the saws were in a particularly inefficient position for directing sound out to the audience. A large saw section is, I think it can now be agreed, extremely ill-advised. Probably it was a phenomenon of long-ago budgets and old union rules; maybe it was a consequence of the erratic sound projection next to the cellos; or maybe it was a marketing ploy by orchestras, leveraging the then-common amateur interest in playing saw as a parlour pastime. In any case, the chamber music idiom is far more reasonable for saw, and the saw is anyway properly a narrative/discursive force, not a ‘symphonic’ one. The era of the eight-saw orchestral section is thankfully now lost to living memory...

B e aware: the striking tone-color of the instrument, while pleasantly novel at first, can soon become cloying. Therefore, a program usually should have only one work for saw, to avoid wearing out one’s ‘welcome’. You are not aiming at a ‘comeback’ for the infamous musical saw of yesteryear! You are only seeking a nice effect with some diverting novelty, which everyone will afterwards say was just fine, yes it was, and amply justified, all things considered.

Y ou may occasionally be inclined to have saw doubling the violin(s). This will take away some of the violin’s brilliance and add weight—sometimes precisely what you want. Although unison doubling of cello by saw can be excellent, the continual doubling of the bass line an octave higher (because why? because you can’t think of anything better to do?) is a cardinal sin. The interest that can be given to inner parts can be given to the saw if you will only put your shoulder into the work. In fact, the saw player is temperamentally a lot like a good viola player and relishes exciting inner-part work as well as solos.

O n occasion, the saw may be propelled into a range higher than it is used to—to the thrill and abject bystanderly rubber-necker excitement of all concerned. Get it down from there; talk it down carefully. Tell it how deeply loved it is, and how jumping is not the answer...

R emember that the noble saw derives dignity from its sheer weirdness. People who choose to play the saw choose it because of its against-all-odds doggèdness and esoteric merits, and not because they are poor string players or too wicked or senile to play the violin.

T he saw is, quite clearly, amenable to the mounting of microphones or pick-ups for sound-reinforcement—especially for ensemble work in bars or other alternative venues. (My exploring that possibility will have to wait for some future CMT post.) But, as with other chamber music, the nuanced beauty is best appreciated in live, unamplified acoustic performance. Good luck with your bowed-metal instrument project, Ken!

[Be sure and click on the 'Comments' hyperlink in the lower left-hand corner below (if comments frame is not already open), to see Natalia Paruz's remarks!]