Page  104 ï~~Automated Music, interdisciplinary aspects of a developmental history1 Carola Bohm, music@computing Nerotalstr. 6, 55124 Mainz, Germany email: Abstract: This paper discusses a historical aspect of computational music, the developmental history of musical automatons.and algorithmic methods of composition. In my research studies I have found that there has been a continuous tradition of using automation as a method to compose and play music since the classical antiquity, a tradition closely linked to the innovations in the natural and mechanical sciences and immanently related to a societies' philosophies of life 1. Introduction The chronology of computers is just a new chapter of the history of mathematics, the sciences and the automatons. Its roots lie even further back, when mankind began to count, to sort, to draw, to weave and to make music.2 Looking at the sagas of the ancient world, it can be seen that the urge to automate a process, including musical processes, has existed from the beginning of written history and stemming from certain philosophies of life. The term "computing" means to determine something by certain (numerical) methods (com - mentally, together in mind; putare - to think, ponder). The term "automaton" derives from the greek word "automatos": acting by itself. These two words clearly show that from early on there existed a bipolarity in the reception of automated processes. "Computare" was the theoretical use of strategies in a solution-making process. Today we would call it the algorithms, i.e. the software. Automatons, even in the ancient sense, were the machines and machinerie, that had the ability of acting by themselves, i.e. the hardware. It is another presentation of the duality of theory and practice, thinking and doing, ratio vs. sensus. In the specific field of computing, and even more specific in music computing, the act of thinking in musical algorithms stands besides the conversion of these algorithms into mechanisms to automate a process. The ability to think of music in terms of sets of rules stems from the Middle Ages, whereas the skill to build mechanical mechanisms based on sets of rules roots back to the Rennaissance. It can be said that these two changes in the way one thought about music, created today's basic ability in computer music to compose through a creation of an algorithmic system and its conversion into computing machinerie. 2. Algorithms in the Middle A2es In the Early Middle Ages the term "algorithm" emerges for the first time. It described the methods of certain rules in computing with arabic numerals. The persian-arabic mathematician Ibn Musa AlChawArismi is regarded to be the author of a book about rules for computing with arabic numerals, that is the then relatively new decimal system. The term "algorithm" itself is derived from AIChawArismi's name. It seems that along with it, a whole new way of thinking was developed: an algorithmic way of thinking using sets of rules to apply to problem-solving processes. Musicology has always accepted the major achievement of the Middle Ages to be the creation of music notation. An aspect too often neglected is the above described appearance of a new way of thinking. This change can be seen in contemporary methods used for composing, in the contemporary music theory and in the then current mathematical computational methods. In the preceding decades music theory and mathematics had been a theory of knowledge. Jamie James described the antiquity with its philosophy of life as folows: "Everything you can see or hear and know is an aspect of the ultimate truth: (...) all are reflections of the essential perfection of the universe. (...) It is not a simple matter of faith: the best philosophical and scientific minds have proven that it is so." 3 In the Middle Ages mathematics and music theory now evolve to become a fixed set of rules and were finally used as abstract tools. Composing became a process of using certain rules to create music. A treatise of Anonymus IV (1290) concerning notation techniques describes just this major change: "And thus they [the older music theorists] tried with great difficulty to understand just that, which is nowadays understood effortlessly by those, who take on the help of rules to achieve the same result in one hour, for which the elders of the past took seven hours. "4 Bohm 104 ICMC Proceedings 1996

Page  105 ï~~Forerunner for this way of thinking is Guido d'Arezzo with his musical algorithms of the 11th century, which are often seen as the first algorithms to automate the process of composing. He described in his "Micrologus"5 a method, with which melodies could easily be created out of any given text by designating the five vowels to the notes of an ascending scale. Quite certain is that d'Arezzos objective did not lie in the automation of a musical process. The used texts of his day were always the "word of god", thus being the most important part in a composition. He argues: "These five vowels we will use, may they lend their melodious sound to the words"6 His algorithm supported the asthetic intention of his period, which lay not yet in the methodical structure of composition as it does in our days. His algorithmic way of composing was for him a method of laying stress to the words of god, even deriving the melodie out of just these words. But even accepting the "unintentional use" of his musical algorithms it can be stated that his way of composing introduced a new way of thinking about music. Formal rules were brought into the compositional process, rules which metamorphosed music theory from a theory of knowledge into a set of rules. Another early example for this new way of thinking about music is again found in d'Arezzos "Micrologus" with his attempt of comparing grammatical to musical structures. With categories of grammar d'Arezzo was able to define the structure of gregorian chant, not only to describe it exactly but also to design new gregorian chants using these categories of grammar as instructions for new compositions. 7 In later periods these new music theory paradigms climaxed in the almost automated composition processes of a Philip de Vitry (13th century) with his isorhythmic motets or the strict canon art in the Netherlands of the 15th century. If we see our modern computer music as an end product in a history of composition, than we find important elements of our current music deeply rooted in these Middle Ages. Algorithmic thinking made it possible to describe a system with a set of rules and afterwards use these in a further creation process. Without the mechanisms for automation, these rules represented the "software" of a computing process; a set of rules, an algorithm without which modemn day computing in music would not be possible. 3. The Rennaissance, from algorithms to mechanics The Rennaissance has been oftenly called the "the age which raised mechanics to art". It introduced another aspect into mathematics and music. One could say that it brought the "hardware" into arithmetic computing and composing music, for rules were now put into mechanisms, creating calculating nrmachines, musical automatons and composing machines. A major prerequisite for the creation of these machines was the bidirectional approach of theory and practice. Whereas in the precedent periods theorists kept apart from the practicioners of their own art, a new ideal of a scholar emerged. In the musical arts of the 16th century the term "musico perfetto" now meant the ideal musician, who would combine practical experience with theoretical knowledge. Gioseffo Zarlino not only influenced but also personified this ideal, having written the "Istitutio harmoniche"(1558) in which he united the "Musica theorica" and the "Musica practica".8 Such works as Vicentinos "L'Antica Musica ridotta alla moderna Prattica"(1555) (transl. "Music of the antiquity reduced to modern practice") couldn't have been written without this closing gap in scholarly knowledge. In the mathematical science the theoretical invention of logarithms by John Napier around 1600 was followed almost immediately by the creation of mechanical tools to aid this calculating method, as for instance the arithmetic slide rulers. Having already worked out excact theoretical methods of describing and composing music in the preceeding ages, it was logical to also invent tools to aid in this compositional process and to build the mechanical counterparts to already existant theory. Athanasius Kircher's "Musica Arithmetica"9(1650) looked quite similar to Napier's arithmetic slide rulers. It consisted out of a box with wooden slide rulers. These rulers were inscribed with figures representing the notes of a scale, the measure of time and the rythms. The user had to solely combine the rulers to simply read the notes, measure and rythm for up to four voices. This machine should not be thought of as a avantgarde jest of the time. It was rather an attempt to shed light upon the mysteries of music with existant progressive methods. Even the name of this machine "musica arithmetica" implies the use of arithmetics to describe a system excactly and completely, define the system of practiced music of the time with a set of rules and realize it as a mechanical automaton. Other "machines" or mechanical tools that followed were for instance A uctor Lampadius ICMC Proceedings 1996 105 Bohm

Page  106 ï~~(1500-1559) "Tabula compositoria" besides the numerous aoutomatons with simpler replaying mechanisms, as the automatic organs of Athanasius Kircher, Robertus Fluctibus and others. Again, if we think in terms of our contemporary electronic and digital music, another major element was founded in the Rennaissance: the knowledge of the possibility to describe a closed system, such as a compositional method, define it complete and excact with rules and realize it as a mechanical counterpart. This knowledge has its roots in the the age of Rennaissance. From this era on, abstract tools such as slide rulers or arithmetic machines have been used to compute certain problems and these abstract tools, these mechanisms were theory put into practical mechanics, or as we would say in our computing age, they were the hardware to the software. 4. Conclusion The way music is defined and practiced has changed throughout the history. Especially in the "automated music" these changes were closely related to other major developments in the sciences and mathematics. Music theory changed from a science (antiquity) to a set of rules (rennaissance) to representations of certain views of music (modern times), just as mathematics changed from a theory of knowledge to an abstract tool. With the disappearance of the gap between the practical and the theoretical fields of science the predominant technology changes from mathematics to mechanics, and later in our times to electronics. The connection to the developmental history of "automated music" exists and can directly be seen in the composition methods and the instruments used. Antiquity Renaissance Modern Age Music theory it mpositional rules music systems/ music LJepresentatis Theory of Algorithms Complete and excact knowledge! or sets of rules descriptions ofcolsed I II Imusical systems Major occurences of "automated music' Musical Music automatons Music computer und automatons and composition computer music, depicting machines that con- creationof own closed Icertain I verted sets of rules[ systems as works of art nature I Imachinerie I Ialgorithrnicmusic) I water organl I~cher, ' I These epoches founded the ability to think in terms of algorithms and mechanical systems, making computation possible. Thus our current computer music, which seemes to have closed the gap between mathematics and music again, stems in its most inner prerequesite from these ages and their "automated music". This research work can be seen as a historical view on computing in music, thus belonging to the early history of computer music, or rather its own prehistoric times. 1 This work is expected to be published in bookform in german language under the titel "Automatisierte Musik, interdisziplinare Aspekte einer Entwicklungsgeschichte" by the end of this year. For more information contact the author. 2 Hans Jochen Schneider (Hg.), Lexikon der Informatik und Datenverarbeitung, Wien 1991, p.151. 3 Jamie James, Music of the Spheres. Music, Science and the Natural Order of the Universe, London 1993, p.3. 4 Anonymus IV concerning notations, in: Willi Apel: Die Notation der polyphonen Musik, 900-1600, Leipzig 1962, p.268. Original: "et nimio tempore laborabant antequam scirent bene aliquid quod nunc ex levi ab omnibus laborantibus circa talia percipitur mediantibus predictorum, ita quod quilibet proficerit in una hora quam in septem ante quod longum ire." 5 Guido von Arezzo, Guidonis Aretini Micrologus, in: Corpus Scriptorum di musica IV, ed. Jos Smits van Waesberghe, American Institut of musicology, Rome 1955. 6 Authors translation based on Mich. Hermesdorff, Kurze Abhandlung Guidos iber die Regeln musikalischer Kunst, Trier 1876. Original: "Itas itaque quinque vocales sumamus, forsitan cum tantum concordiae tribuunt verbis..." 7 See Karlheinz Schlager, Aspekte der mittelalterlichen Musik, in: Propylaen Geschichte der Literatur, Frankfurt 1988, p.411. 8 See Renate Groth, Italienische Musiktheorie im 16. und 17. Jahrhundert (= Frieder Zaminer (Ed.), Geschichte der Musiktheorie, Bd.7), p.2 -3. 9 Athanasius Kircher, M.usurgia universalis, 2 Bde., Rom 1650, Fascimile Hildesheim, New York 1970. Bohm 106 ICMC Proceedings 1996