STUDY OF
OF
THE
THEORIES TARTINI
GIUSEPPE
Figure 1.
This leads him to devote much space in his work to long and abs t r u s e mathematical and geometric proofs in which he seeks to demons t r a t e the validity of this a priori principle and the necessary inclusion of the principles of major, minor, and dissonant harmony in this univ e r s a l principle to thus produce a complete and closed system. To most of his contemporaries these proofs were incomprehen
THEORIES O F GIUSEPPE TARTINI sible, and those who understood them found e r r o r s in T a r t i n i ' s calculations and attacked his system violently. Even T a r t i n i 1 s follower, Benjamin Stillingfleet says: "One cannot without some impression of compassion, s e e him wandering in the perplexing labyrinths of a b s t r a c t ideas, almost without guide, o r a t best with one which i t is most likely would mislead him [, 16]!' And l a t e r on: "What I have already said will be sufficient for my not entering into a detail on this long chapter Ch. 111 a s such a detail would be extremely tedious to some, unintelligible to others and would appear (at least) strange to the only men who a r e qualified to form any judgement on this matter, I mean the mathematicians [6, 17]!' This s e e m s to explain why there has never been a n attempt to account fully f o r this part of T a r t i n i 1 s work, to show where the e r r o r s a r e and in what way they affect his conclusions. Such a n attempt is made in this essay. Tartini's theoretical works a r e : the Trattato di Musica [TI that appeared in 1754, De' Principj delllArmonia [PI that appeared in 1767, and a Risposta di G. T.. alla crittica del di lui Trattato [ a ] that appeared in 1767. His other works a r e two small t r a c t s on ornamentation [A] and violin playing [L,]. He has been credited with the authorto ship of an anonymous ~ i s ~ o s t a l J . J . Rousseau that appeared in 1769. This essay is a comparative study of the Trattato, the Principj o and the Risposta of 1767; although it would have b m i r a b l e t pare these works with the anonymous Risposta, no copy of this work has been available to me. The Trattato is a long and abtruse work which contains the most comprehensive account of T a r t i n i ' s theoretical thought. After a n introduction, where Tartini exposes the m a t h e m a t i c a l p r e m i s e s Tor the demonstrations in the second and third chapters, the work begins with an account of several phenomena which point to the possibility of a physical basis for harmony. Of particular importance here is his a c count ofthe differencetones whichare the most secure physical foundation for his system; in the Principj he claims to have discovered this phenomenon in 17 14 [P, 361. In the second and third chapters we find the bulk of his mathematical calculations and geometric proofs; he t r i e s to prove the harmonic nature of the circle and to derive f r o m the r e l a tion ofthe circumference to the diameter the major and minor systems, the diatonic and chromatic dissonances, and even a n enharmonic system. The l a s t two chapters a r e devoted to the derivation of the scale, the establishment of the basic b a s s progression, and a discussion of several intervals found in the music of his day. The work concludes by answering some possible arguments that could be raised against his system. The Principj is a much c l e a r e r work than the Trattato; the mathematical proofs of the e a r l i e r work a r e for the most part absent here. There is a general reexposition of the principles of the Trattato, but limited to the diatonic system. The work is divided into four parts. The f i r s t deals with the physical foundations of harmony, and contains accounts s i m i l a r to those in Chapter I of the Trattato. The second
[x,
..
a 1 celebre
signor
J.
..J . . .
Rousseau
f o r the m a j o r tone:
THEORIES O F GIUSEPPE TARTINI deals with the systematic foundation of harmony and is concerned with the derivation of intervals. The third t r e a t s what Tartini calls "the musical foundation of harmonyu and consists of the derivation of the scale and the establishment of the basic b a s s progression. The fourth relates the material ofthe three previous chapters into a unified whole. The Risposta is a small polemic work in which Tartini answers the objectibns raised to his Trattato by J e a n Adam ~ e r r e ? It contains v e r y little that has not been dealt with m o r e extensively in the other two books. The most important proposition in the Ris osta is a geometric explanation for the formula of the "third sound (i. e. differencetones). Tartini's system is a closely woven one. Unfortunately i t s author is "quite ignorant of geometry and algebra [B, 51" and his mistaken calculations lead him astray. He firmly believes that he has found the answer to a l l the problems of harmonic theory and has discovered a new s e t of properties for the geometric figures with which his proofs deal. In the Risposta he mentions that he has developed a new science which however he will not yet publish since the few glimpses of i t that appeared in the Trattato were so illreceived 13, 101. The attack on T a r t i n i ' s system by the French theorists led Fdtis to consider T a r t i n i ' s theories a s directly opposed to the theories of viii, 1871. This is denied by Shirlaw [a, 2931; and i t will Rameau be evident to anyone who compares the works of the two theorists that, in spite of T a r t i n i ' s polemic attitude towards Rameau and the attack on Tartini's system by Rameauts followers, the basic theoretical conclusions of both men a r e quite s i m i l a r 3
[z,
In the Trattato, Tartini observes that when a monochord string is plucked, even though i t is one string, i t produces three distinct sounds: the fundamental, the twelfth, and the seventeenth, which a r e in the harmonic proportion of 1:1/3:1/5 [T,101. Likewise, he notices that when a n organ key (with a mixture stop) is pressed many pipes of different pitch sound; but since the sounds produced by the pipes a r e in harmonic proportion, the a u r a l impression is that of a single sound [T, 121. F r o m this he deduces that the essence of harmony is unity, a unity that divides itself into multiplicity only to return again to unity a s i t s basic principle [T,131. He finds further support for this in the fact that instruments such a s the tromba marina produce no sound unl e s s the string is divided into a n aliquot part4 of the total length (unity) of the string [T, 1112]. In this sense the sounds of the monochord string a r e "harmonic monads" and the t e r m aliquot p a r t s does not 121. properly describe them [T, Rameau s t a t e s that musical sound is not one but three 15, 2881.
2 . In the O b s e m t i o n s sur les principes de ltha.rmonie, occasiondes par quelqu&s gcrits modernes sur ce sujet, et particuli&rement par.. le traiix! de th6orie musicale de M W i n i . . . [~eneva:Henri Albert Gosse et Jean Gosse, 17631. This work was not available to me. 3 . There is an excellent account of Rameauys theories bu Joan
Ferrls [8].
4 By an aliquot part of any quantity, line, or surface is meant
. such a part as will measure the whole wlthout a remainder. That is, an
integral divisor.
ALEJANDRO ENRIQUE PLANCHART When Tartini s t a t e s that the three sounds of the monochord string a r e harmonic monads, he does not place himself in direct opposition to Rameau since he considers multiplicity a function of unity and r e g a r d s the division of unity into multiplicity and the resolution of multiplicity into unity a s p a r t s of a complete cycle [T, 131. Rameau's emphasis on multiplicity would not contradictdirectly T a r t i n i ' s assumption of unity a s the basic harmonic principle. Whether o r not multiplicity should be considered beyond the three sounds (1, 113, 115) which a r e audibly produced by the monochord string i s a matter of extensive discussion in both the Trattato and the Principj. In the conclusion of the Trattato, Tartini admits that other theorists claim to h e a r the sounds 1/ 2, 1/ 4, and even 1/6 in addition to the three sounds he has already reported a s being produced by a plucked string. Tartini admits that this may be true but that it would not invalidate his theories since these additional sounds a r e a l s o in harmonic proportion [T_, 1'701. He opens the Principi with a discussion of the specific a s s e r t i o n s of Rameau [& 1941 and dlAlembert [_P, 43ffl with respect to this question. Both claim to have heardthe sounds 112 and 114 produced by a plucked string. Tartini points out that although they make this claim, they ignore their findings in setting up harmonic s y s t e m s based solely on 1, 113, and 115 [P, 31. He admits that he has been unsuccessful in his attempts to determine the presence of the sounds 112 and 114 in the vibration of the plucked string because, a s octave doublings, they a r e swallowed up by the fundamental sound of the s t r i n g [_P, 41. He claims that if Rameau and dlAlembert have indeed heard these sounds, their s y s t e m s invalidate themselves by not making use of their empirical findings [? 31. I, In his s e a r c h f o r a m o r e comprehensive physical foundation for a harmonic system he t u r n s to his discovery of the differencetones, which he calls the "phenomenon of the third sound [E, 511' As reported in the Trattato, he determines their existence through a s e r i e s of experiments in which two instruments play s e v e r a l intervals. If the intervals a r e played with just intonation and loud enough, a third sound, generally lower than either of the two sounds being played, will be heard [ T , 1314]. This account is accompanied in the Trattato by a l i s t of the third sounds for the different intervals (Table I, p. 34): for the octave and the unison no third sound is given since Tartini claims that these intervals produce none. The third sounds of an interval and i t s inversion a r e the same, sometimes with an octave displacement [T,14ffl. Any interval will produce a third sound. If two violins play the following: it will be noticed that the third sound glides
from that of one interval to that of the other, never ceasing to sound even though an incommensurable number of intervening intervals have occurred [X, 171. To him, the third sound is the r e a l physical fundamental b a s s of any given interval and of any given p a i r of melodic lines; the successive third sounds produced by their combination constitute the t r u e fundamental b a s s of the melodies. Any other b a s s would be absurd o r , a t best, artificial [T, 171. Tartini observes that any combination of intervals out of the
THEORIES OF GIUSEPPE TARTINI senario up to the sounding of the whole senario produces a single third sound, which in the Trattato, he mistakenly identifies a s 112, that is, an octave above the fundamental rT, 181. However, in the conclusion of the Trattato he mentions that other theorists have identified the third sound a s 1, o r the fundamental of the s e r i e s . This, he says, does not upset his a s s e r t i o n of the third sound a s the fundamental b a s s and physical root of any interval o r harmonic s e r i e s [T_, 17011. As a consequence of the identification of the third sound of the harmonic s e r i e s a s constant in 112, Tartini believes that he h a s found a novel property of harmonic ratios. Given I / 2: 113, the resultant (third sound) is 112; given 1/3:1/4, the resultant is 112 and s o on [T, 1819]. This goes against the basic p r e m i s e of the system, which cons i d e r s harmony a s a projection of unity. T a r t i n i overlooks this contradiction. In the Principj, Tartini r e v i s e s his view and identifies the third sound, still mistakenly, a s constant in 1 o r in unison with the fundamental of the harmonic s e r i e s [_P, 221; in addition, he gives a mathematical formula for the deduction of the third sound of any interval. The interval of the fifth, 1/ 2: 1/ 3, can be expressed arithmetically a s 3:2. T a r t i n i a s s e r t s that the product of the values that r e p r e s e n t the simplest arithmetic expression of a n interval will give the c o r r e c t third sound. In the case of the fifth, 3 x 2 = 6; thus 6 will r e p r e s e n t the third sound of the interval 3:2. If 6 is expressed a s 1 of the harmonic s e r i e s , 3 will be 112 and 2 will be 113 [_P, 561. When a n interval is expressed in a m o r e complex form, i. e . the octave a s 2:4, the product of the t e r m s of the ratio must be divided by the common factor in o r d e r to a r r i v e a t the c o r r e c t third sound [_P, 61. This formula enables Tartini to determine the third sounds f o r the octave and the unison. He explains that they a r e inaudible because they a r e in unison with one of the generating tones [_P, 51. Tartini fails to notice that he h a s invalidated his a s s e r t i o n that the third sound of the senario is constant in 1. The third sound of the octave 1/2:1/4, which is included in the senario, is, by his formula, 112. In the Risposta, trying to defend the Trattato a t a l l costs, he attempts to justify both 1 and 112 a s third sounds. This is done by means of an involved mathematical proof that attempts to demonstrate that 112 is the geometric expression of the third sound, while 1 is the arithmetic one k, 25ffl. It is important to note that T a r t i n i d o e s not consider the third sound a s a foreign element added to the generating interval; the interval, the third sound, and a l l the harmonic divisions in between, f o r m a harmonic whole [E,671. The formula for the third sound and the operation it r e p r e s e n t s a r e s i m i l a r to a formula proposed by Leonhard Euler. E u l e r ' s formula is based on the multiplication of the t e r m s of the ratio a t which two s t r i n g s vibrate when producing a n interval; the product, together with a l l i t s integral divisors, will f o r m a complex s i m i l a r to the harmonic complex of T a r t i n i l s s y s t e m of the third sound [L 39ffl. T a r t i n i a r g u e s that E u l e r l s formula fails to produce a harmonic proportion. The vibration of two s t r i n g s producing the interval of a fifth a r e in the ratio 3:2, the product of these t e r m s is 6. The integral divisors of 6 a r e 3, 2, and 1, and the complex formed by E u l e r l s law is 6:3:2:1, which is not in harmonic proportion. The complex 6:3:2,
Figure 2 .
Example 2 .
Example 3.
Example 4.
Example 5.
THEORIES O F GIUSEPPE TARTINI produced by T a r t i n i ' s system, i s [E, 891. Moreover, in the case of the minor third, 6:5, the complex a r r i v e d a t by means of E u l e r ' s formula is not even consonant, while the one a r r i v e d a t by T a r t i n i t s formula is. F r o m this, Tartini concludes that the law of the third sound is the law of harmonic proportion and has i t s b a s i s in unity; in consequence, the ratios contained within the senario should be considered a s necessarily basic [_P, 91. To Tartini, the nature of the third sound consists in three properties: first, i t is intrinsically physical harmonic unity a s a principle; second, it produces harmonic proportion with the generating sounds; third, i t s simultaneous union with the two generating sounds and a l l the harmonic divisions in between gives r i s e to the harmonic s y s t e m of which the third sound is the fundamental bass [_P, 111. However, the third sound of an interval which has 1 a s one of the t e r m s of i t s ratio cannot produce a proportion with that interval because it will be identical with one of the sounds of the generating interval. In consequence, any interval which has the unit a s one of i t s t e r m s is only potentially harmonic; the third sound will s t a r t producing harmonic proportion with the generating sounds f r o m the second interval on in any ordered s e r i e s [_P, 151. This makes the second interval of any harmonic s e r i e s the interval that determines the harmonic nature of the s e r i e s [_P, 151. Within the senario t h e r e a r e only two possible s e r i e s , one by difference 1 (superparticular) a s 1:1/ 2: 1/ 3 and another by difference 2 (superpartiens) a s 1:1/3:1/5. Any attempt to construct a s e r i e s within the senario using a l a r g e r difference will never reach the second interval which is necessary to determine the nature of the s e r i e s [P, 121. Returning to the phenomenon of the monochord string, Tartini points out that, while the system of the third sound by difference 1 produces the whole of the senario, the s y s t e m by difference 2 produces the essence of the senario in 1:1/3:1/5, the sounds that he has heard in the monochord string. The phenomenon of the monochord string is then 231. only a partial one while that of the third sound is the general one [E, So f a r , Tartini has worked only with the harmonic division of the string which produces the major system. As soon a s he attempts to explain the minor system, trouble a r i s e s . In the Trattato he gives the third sounds f o r the minor triad, which a r e not one, a s in the major triad, but two (Ex. 2). The resultant is a dissonant combination which causes Tartini to comment that, if the third sounds were more audible, only music with m a j o r triads would be possible [T,671. Minor harmony, derived from the arithmetic division of the string, had been considered by other theorists a s "borrowed" from arithmetic science and essentially foreign to harmony. This amounts to a n admission of m o r e than one principle for any successful harmonic system, and Tartini finds this assumption absurd and fundamentally opposed to the idea of system itself [T,681. Tartini admits the "imperfection" of minor harmony a s compared to major major harmony has but he denies that minor harmony "priority of nature" over minor may s t e m from a different principle and considers it a s a n intrinsically inseparable consequence of major harmony [T, 681. The argument in both t r e a t i s e s is essentially the same. Given a string of determined length, e. g. 60, he divides the string harmoni
ALEJANDRO ENRIQUE PLANCHART cally (Fig. 2, p. 38). Each of the segments taken from A to the point of division produces one of the sounds of the s e r i e s shown in Ex. 3 (p. 38). He calls this s e r i e s the system of fractions and it is nothing other than the senario. These divisions, when taken to B, produce another s e r i e s of string lengths which produce .the s e r i e s of sounds shown in EX. 4. He calls this the system of advances and considers it the physical proof that minor harmony and the subdominant a r e a n inseparable consequence of major harmony [T, 69; 2, 25ffl. He overlooks the factthat the system of advances consists of fractions such a s 516 which a r e not aliquot parts of the string and could not be produced by such instruments a s the tromba marina. Rameau relates minor harmony to the covibration of multiples of the fundamental string [3, 2215 ~ a r t i n iconsiders Rameau's point when he superimposes the systems of fractions and of advances upon a constant fundamental sound (Ex. 5 ) . The resultant third sounds (F and ~b in particular) a r e the s a m e a s the sounds of Rameau's sympathetic strings [_P, 2671. He s e e m s not to notice that greatC is not the fundamental of that s e r i e s and that the successive third sounds a r e the fundamentals of each interval. His s e r i e s , a s Shirlaw points out, is actually made of intervals belonging to different harmonic s e r i e s and a t best it proves only that minor harmony is some s o r t of inverted major h a r mony E, 29561. Tartini is aware that m e r e physical proofs a r e incomplete and s e t s out to s e a r c h for a mathematical foundation for his system. However, mathematical principles, in o r d e r to be valid, must be evident in the observed physical facts. "where the point is to establish a system, it i s necessary to unite the two categories, physical and mathematical, in such a way that they would be inseparable and form a single principle [T_, 201:' The unity which he believes to have found in the physical foundation of harmony is best expressed by the circle which is in itself one. Each of the infinite number of radii that enter into i t s construction can 211. This is not enough. The circle must be proven s e r v e a s unity [T, to be a harmonic unity, and to this end Tartini considers it in connection with the straight line represented in the square. In T a r t i n i t s opinion, this premise has a definite physical basis. The straight line of the monochord string, when it vibrates to produce the harmonic proportion 1:1/3:1/5, produces some s o r t of circular figure, and the a i r m a s s e s s e t in motion by the string take a spheric shape. These
5. Rameau had s t a t e d , mistakenly, t h a t a v i b r a t i n g s t r i n g w i l l cause t o v i b r a t e o t h e r s t r i n g s which a r e i t s r i u l t i p l e s i n l e n g t h a t t h e i r r e s p e c t i v e fundamentals. This would imply a s e r i e s of "undertones" i n i n v e r s e p r o p o r t i o n t o t h e s e r i e s of "overtones!' Accordingly, a fundamental C would generate F , A ~ ,C . I n t h e D & o n s t r a t i ~ a [ ] Rameau J, admits t h a t t h i s observation was erroneous; t h e longer s t r i n g s v i b r a t e i n segments corresponding t o t h e unison of the e x c i t i n g s t r i n g . S t i l l , he continues t o d e r i v e minor harmony from t h e p r o p o r t i o n 6:5:4 and maintains t h a t while major harmony is a d i r e c t product of nature, minor harmony i s " i n d i c a t e d by n a t u r e " [B, 2361.
THEORIES O F GIUSEPPE TARTINI phenomena can be best expressed by the c i r c l e which is the most p e r fect of a l l c u r v e s [_T, 211. The c i r c u l a r line presupposes the existence of a straight line in the radius, which, upon completion of the c i r c u m ference produces the diameter [?_, 211. The diameter, placed tangent to the circumference, becomes the side of a square which c i r c u m s c r i b e s the c i r c l e ( s e e Fig. 1, p. 32). F r o m this premise, Tartini attempts to demonstrate the h a r monic nature of the c i r c l e and to derive f r o m i t the basic principles of major, minor, and dissonant harmony, and chromaticism; and to prove that the limit of consonant harmony is to be found in the senario. Here it is necessary to introduce some of T a r t i n i t s arithmetic p r e m i s e s . In the Trattato p r e m e s s o to the Trattato he s t a t e s that, given the duple r a t i o 60:120, the harmonic mean will be 80, the a r i t h metic mean will be 90, and the counterharmonic mean will be 100. T h e r e is a mean missing, the geometric mean, which Tartini establishes a s 84 [T, 11. The nature of the geometric mean is that the square of the mean is equal to the product of the e x t r e m e s o r that the differences between the e x t r e m e s and the mean a r e i n the s a m e ratio a s the two successive ratios which f o r m the proportion. F o r example, 2 is the geometric n mean between 1 and 4. I consequence, the square of 2 is equal to the product of 1 x 4, and the differences between 1 and 2 and 2 and 4 a r e i n the ratio 1:2 o r 2:4, the two r a t i o s that f o r m the proportion 1:2:4. In T a r t i n i ' s proportion 120:84:60, this does not hold. This proportion, reduced to i t s simplest t e r m s is 10:7:5. It is obvious that the differences between the mean and the e x t r e m e s a r e not in the s a m e ratio since the mean does not produce the s a m e ratio with both ext r e m e s ; besides, the square of 7 is 49 while the product ofthe e x t r e m e s is 50. Tartini admits this when he s a y s that we cannot have a notion of this mean a s geometric because i t cannot be expressed by integers but only by a line [T, 121? Thus he makes a distinction between what he  .t e r m s the complete irrational geometric mean and the incomplete r a i to n a l 21. The f o r m e r would b e m61 the l a t t e r i m (7). Tartini operates with the second one. This is s not a valid premise, but Tartini, through this and other such approximations, attempts to warp geometry to suit his system. Another t e r m which we encounter in T a r t i n i ' s theory is the d 2 c r e t e geometric proportion. By this Tartini means a proportion in which the product of the harmonic and the arithmetic means is equal to the product of the extremes. In this way, the irrational geometric mean that l i e s between the harmonic and arithmetic means can be exp r e s s e d discretely i n t e r m s of rational integers. F o r example, the proportion 12:9:8:6, of which 9 is the arithmetic mean and 8 is the h a r monic mean, is a discrete geometric proportion since both 12 x 6 and 9 x 8 equal 72. T a r t i n i t s f i r s t goal is to prove the harmonic nature of the circle. Taking the c i r c l e AB (Fig. 3, p. 42) with the diameter d i v i d e d r a t i o n ally a t any point, l e t u s say x, T a r t i n i attempts to show that xy will be the harmonic mean and xz the arithmetic mean of the ratio Ax:xB into
[z,
6. The l i n e t o which T a r t i n i r e f e r s is t h e diagonal of a square which i s always i r r a t i o n a l since it involves a length equal t o w .
Figure 3.
THEORIES O F GIUSEPPE TARTINI which the diameter AB is divided. If, Ax:xB = 3:7
=
=
then: Ax2 
X B ~
9
49 21 (see footnote 7) 25 (because it i s r2).
G2
G2
=
=
Expanding the ratio 3:7 (by multiplying by 5) we a r r i v e a t the dis3rete geometric proportion 15:21:25:35, of which 21 is the harmonic mean and 25 the arithmetic mean [T, 22, proposition 111. This s e e m s to indicate that the length of the perpendicular dropped from any point on the diameter to the circumference of a circle is the harmonic mean of the ratio of the line segments into which the diameter is divided by the point on the diameter [T,391. Tartini's e r r o r in this proof l i e s in the expression ofthe discrete geometric proportion of the ratio 3:7. He multiplies a l l the t e r m s by five except the harmonic mean. If we reduce the discrete geometric proportion to i t s original form, then 15:21:25:35 becomes 3:21/5:5:7. However, 2115 does not represent the segment xy which i s actually for which 2 I V 3 i s a close enough approximation to make the e r r o r l e s s evident. Tartini s e e m s to have been vaguely aware that something was out of o r d e r in his expression of the ratio 3:7 a s the proportion 15:2 1:25:35 since he attempts to demonstrate, l a t e r in the Trattato, that the proportion 15 :2 1:25: 35 is a numerical expression for the lengths of the lines in the proportion Ax:xy:xz:xB [T, 241. The second proof of the harmonic nature of the c i r g e is through the same figure with the chord Ay and the hypotenuse Az added to it (Fig. 4, p. 42). Given the circle AB, with the diameter dividedrationally a t x. Tartini says that Ay will be the harmonic mean and Az the counterharmonic mean of the ratio Ax :xz.
If, then:
Ax:xB = 5:9


Ax2
X B ~
=
= =
25 81 45 49 (because it is r2)
2 XY
z2
5
By = Ay2 (70) and + Ax2 =  the theorem of pythagoras? Ax2 + A Z (74). z 2 : z 2 = 70:74, which, reduced to the simplest t e r m s , is ~ 35:37. If AX:^ i s expressed a s a discrete geometric proportion, we
F2
z2
7. i s t h e perpendicular dropped from t h e v e r t e x of t h e r i g h t t r i a n g l e AyB,>ence t h e square of i s equal t o Ax times xB. Substitu t i n g 3 f o r Ax and 7 f o r xB we have 3 x 7 = 2 1 which i s t h e square of
a.
of a r i g h t
triangle i s
ALEJANDRO ENRIQUE PLANCHART have 30:35:37:42, of which 35 is the harmonic mean and 37 the counterharmonic mean. This again s e e m s to indicate the harmonic nature of the circle; and further, that the nature of the square s e e m s to be both arithmetic and counterharmonic [_T, 22ff, proposition 1 1 , 11 Again here, a s in the previous proof, T a r t i n i 1 s e r r o r l i e s in the expression of his t e r m s a s a proportion. If we reduce the proportion 30:35:37:42 to i t s original f o r m we have 5:35/6:37/6:7. H ~ r ethe and Az t e r m s 35 / 6 and 371 6 do not r e p r e s e n t the lines Ax but only approximate them. Tartini never revised his calculations. The Principj contains no calculations of this s o r t , but in the Risposta, published the s a m e y e a r a s the Principj, he s t a t e s "the c i r c l e is a n infinite number of harmonic means [&, 141:' This placed him in opposition to a l l the mathematicians who had demonstrated that the c i r c l e is the locus of the geometric mean of the infinite number of r a t i o s into which the diameter can be divided. He is aware o f t h i s when he a t tempts to justify his view by saying that the sine (xy), a s geometric mean, contains both the arithmetic and harmonic means, the arithmetic a t the point where the sine touches the diameter, the harmonic a t the proposition point where the sine touches the circumference [T, 22ff, IV]. Having a r r i v e d a t the conclusion that the c i r c l e is intrinsically harmonic while the s q u a r e is intrinsically arithmetic and counter) harmonic, Tartini s a y s that the diameter (arithmetic per s ~ should be 491. But f r o m a diameter divided harmonidivided harmonically cally, the only s y s t e m that can be derived is that of consonant harmony and a s such a particular system. The universal system, Tartini claims, must be derived f r o m the chords, complements, sines, and protracted sines a s well, which a r e derived f r o m the diameter divided harmonically 531. Theoretically, the diameter can be divided harmonically g d infih m , but in o r d e r to produce a closed s y s t e m Tartini is forced to find a limit to the divisions of the diameter. As Zarlino does, he s e t s this [T, 567, limit a t the f i r s t s i x divisions of the diameter (the s e n a r i ~ ) proposition VI]. Tartini attempts to prove the necessity of this limit a s follows. Given the c i r c l e and the s q u a r e (Fig. 5, p. 42) with the diameter s e t equal to 120 and divided harmonically through the simplest s i x divisions, the s q u a r e s of the chords and of the hypotenuses in the figure (which according to T a r t i n i a r e the harmonic and counterharmonic means of the proportions considered in the preceding proof) will be:
(m)
(m
[x,
u,
Omitting the s q u a r e s of Aa' and of Aa" since both a r e the same, Tartini constructs the following proportions using the s q u a r e s of the remaining segments:
THEORIES OF GIUSEPPE TARTINI lower extreme harmonic mean counterharmonic mean upper extreme 4000 Ab12 4800 Ab1I2 5200 6000 2700 A C ' ~3600 Ac1I2 4500 5400
The f i r s t proportion is a discrete geometric sesquialtera ( a s 2:3); the second, a discrete geometric duple ( a s 1:2); the third, a disc r e t e geometric proportion of the ratio 2:5; the last, a discrete geometric triple ( a s 1:3) [T, 581. Tartini adds the extremes:
Tartini observes that 100, 90, 84, and 80 a r e the means of the duple ratio 120:60 (which a r e the values assigned to the diameter and the radius of the given circle)? He has managed to base the entire calculation on a circle with the diameter divided in the basic ratios of the genario and has obtained what he considers a unified expression f o r a l l the means of the duple ratio. Tartini r e g a r d s this a s proof that the integral extension and limit of the harmonic s y s t e m a r e in the s e n a r i ~ [ I 601. T His calculations fail because two of his premises a r e incorrect. F i r s t , the t r u e geometric mean of the ratio 120:60 is not 84 6 ( ) b u t m . Second, the squares of the chords, a s we saw in the p r e vious proof, a r e not the harmonic mean of the ratios produced by the sines and the segments of the diameter that form a right triangle with the sines and the chord. The limitation of the harmonic system to the senario c r e a t e s another problem. All accepted consonant intervals a r e included in the senario except the minor sixth which occurs between 115 and 118. This exclusion of the minor sixth brought criticism from S e r r e . In both the Principi and the Risposta, Tartini answers that the minor sixth i s incapable of reduction to a harmonic proportion other than the continuous
84 i s
THEORIES O F GIUSEPPE TARTINI harmonic proportion 1/5:1/6:1/7:1/8, which requires the use of the nondiatonic seventh aliquot part, an absurdity since the harmonic system is essentially diatonic [E,461. However, he adds an attempt to derive the minor sixth from the senario. The harmonic triple, which is composed of a n octave and a fifth can be expressed arithmetically a s 1:2:3. In this form, the proportion is indivisible, but the octave expressed a s 2:4 canbe divided arithmetically 2:3:4 and the fifth, expressed a s 4:6 can be divided arithmetically a s 4:5:6. The s m a l l e r ratios of each proportion (3:4 and 5:6), when put one above the other ( a s musical intervals), form the minor sixth (Ex. 6). To Tartini it is obvious Example 6.
that a n interval produced entirely by arithmetic division cannot be included in the harmonic s y s t e m [E, 481. Once Tartini has established the universal principle of his system in the circle, he undertakes the derivation of the s y s t e m from the r e lationships contained within the figures of the circle and the square. The basic interval of any system is the octave since i t is the f i r s t interval of any superparticular s e r i e s that s t a r t s f r o m unity. The octave itself does not form any system since it is only a ratio, and to have a system i t is necessary to have a proportion. Because of this, the oc24; 2, 221. The tave is considered the a priori ratio of the system second ratio of any system is then the ratio that determines the nature of the system; in the case of the harmonic system, this ratio is the fifth (1/2:1/3). The octave and the fifth a r e then united in the f i r s t harmonic proportion of the system, namely, the triple proportion 1:1/2: 113 1 , 241. 2 All this is exemplified, according to Tartini, in the construction of the circle; the radius being unity which, upon completion of the circle, has been doubled in the diameter and (approximately) tripled in the circumference. This produces the arithmetic s e r i e s 1:2:3, which, inverted, produces the harmonic triple 1:1/2:1/3 [T, 27ffl. In the inversion, the c i r c l e then represents harmonic a s well a s arithmetic unity. Tartini attempts to approximate this proportion through long calculations using Archimedes1 measurements of 7 for the radius, 14 for the diameter, and 22 f o r the circumference. He repeats the calculation two other times, the l a s t time using such values a s 1,000, 000, 000 for the radius, 2,000,000,000 for the diameter, and 6,283,185,507 for the circumference[T, 33ffl. The derivation of the intervals of the senario in the Trattato p r e sents no difficulty since they can be derived from the diameter divided harmonically. In the Principi, where Tartini has avoided proofs concerning the circle, the explanation is m o r e involved. He s t a t e s that the five intervals of the senario must not be derived f r o m the division of an already existent ratio. He finds the system based on the three sounds (1:113: 115) of the monochord s t r i n g unsatisfactory since the
[x,
ALEJANDRO ENRIQUE PLANCHART basic ratio of the harmonic system, the octave (1: 1/21, must be obtained by dividing harmonically the ratio 1:1/3, and the last interval of the senarip, the minor third, cannot be obtained unless the fraction 1/7 i s included [_P, 58ffl. He goes then to the system of the third sound and finds that by the superparticular s e r i e s all the intervals of the senario can be obtained without deriving any interval from another and that the octave finds i t s proper place a t the beginning of the s e r i e s [E 5960]. The derivation of minor harmony follows in bothtreatises a s i m i l a r course which consists in deriving i t from the advances produced by the harmonic division of a line (the diameter) a s was explained e a r l i e r in this essay. However, in the Trattato, Tartini explains the relation of such a system of division, which is essentially arithmetic, to the circle by his r e m a r k that the circle must be considered in conjunction with the a projection of the diameter of the circle which is square [T,361 essentially arithmetic. In the Trattatg, after the explanation of the senario, Tartini gives the following rule: "Of the p a r t s that form the s e n a r i ~ 1, 112, 113, 114, : 115, 116, the terms, 1,1/2,1/4,must not be put to ether [by themselves] although they a r e contained within the senarip!' O The explanation of the rule i s that 1: 1/ 2 : 1/ 4 form a continuous geometric proportion and a s such they a r e the potential principle for dissonant harmony [T, 621. In the Trattato, dissonances a r e derived from the circle in the following manner: Given the circle AB with the diameter divided up to the f i r s t s i x divisions and having the sines and the chords drawn (Fig. 6, p. 46), Tartini deduces the systems given b g o w _ ~ Bequals the pitch C. ~ F r o m the squares of A ,' ; Ab', Acl, Ad', r e t : (See Ex. 7a). F r o m the squares of Ba', El, K 1 ,  E ' , Be': (See Ex. 7b). F r o m the squares of a a l , bbt, c c ' , ddl, eel: (See Ex. 7c). Tartini then places the f i r s t (Ex. 7a)under the third (Ex. 7c), and under F r o m these both the successive formations of the senario (Ex. 8). Tartini observes that, in the second position, the dissonance of the ninth occurs, formed by the two fifths cg and gd' which produce a continuous geometric sesquialtera [T, 73ffl. F r o m the s a m e example, he draws the following law: "Generally, any chord will be dissonant " " that contains two similar intervals of different species except (more by custom than by reason) the octave;'l1 Two fourths, two major thirds, etc. will each produce a continuous geometric proportion and, a s such, a dissonant combination. Tartini's next concern is to determine which interval is the dissonant one when two similar intervals appear in the same chord and he says: "Of two similar intervals of different species, that one will be consonant which intrinsically belongs to the harmonic [major] o r arithmetic [minor] system. The dissonant one will be that which does not [belong] and cannot possibly belong to either of the two mentioned , 10. Che nelle p e r t i i n t e e a l i della sestupla armonica, 1 112, 113, 1/4, 1/5, 1/6, non s i pongano insieme questi t r e termini, 1, 1/2, 1/4, . . . . . benche contenuti nella sestupla [_T, 621. 1 . Che in genere qualunque accord0 m s i c a l e sara dissonante, se 1 s i seranno nell'accordo due i n t e r v a l l i simili d i specie diversa eccetua t a (piu per uso che per ragione) l a ottava [T, 741.
KB,

 
THEORIES O F GIUSEPPE TARTINI systems1112 In the same manner, Tartini deals with the remaining notes of Ex. 7c), arriving a t the following l i s t of dissonances: ninth, eleventh (fourth), augmented twelfth (augmented fifth), and fourteenth (seventh). Tartini observes that the thirteenth is not present in this system, yet he includes it by definition since i t contains two fourths and is t h e r e fore dissonant by the law given previously [T,761. In the case of the second he concludes that the dissonant note is not the upper note of the interval and this he considers a reason against calling the interval a "dissonance of the second" E, 771. He proposes no name for this combination. Any dissonance, he states, should be prepared by a melodic unison; the dissonant note and the dissonant interval should be p r e pared by a s i m i l a r consonant interval, with the bass moving likewise by the same interval 99ffI a s in Ex. 9. The resolution of dissonances should proceed downwards by s t e p o r half step [_T, 82ffl. It
[x,
Example 9.
Example 10.
i s curious that Tartini even resolves the augmented twelfth in this mann e r [_T, 82 and 1301, producing the resolution shown in Ex. 10. In the Principi, Tartini is concerned with the diatonic dissonances. F i r s t he considers the possibility of a system in which the dissonant tones a r e a r r i v e d a t f r o m the ratios outside the senario which have odd numbers a s denominators; i. e . 117, 119, l / 11 etc. These a r e the tones produced naturally by the tromba marina and other s i m i l a r instruments. If the f i r s t three aliquotparts with odd denominators a r e added to the senario, we have a system in which the dissonances produced by the ratios with odd numbers a s denominators outside the senario would have a consonant basis in the three basic sounds of the monochord string [P, 901. This would explain the special treatment of the minor seventh (117) which is not always prepared, since it stands in this system a s some s o r t of mean between the consonant partials 1:1/ 3: 115 and the openly dissonant ones 1/9:1/11:1/13 [E,901. This Tartini finds inconclusive since the oddnumbered aliquot p a r t s produce a l l the outoftune tones which cannot be considered diatonic and which a r e a l 12. Che de due i n t e r v a l l i s i m i l i d i s p e c i e d i v e r s a sar& il consonante q u e l l o che intrinsecamente a p p a r t i e n e a 1 sistema armonico. Sarb il dissonante quello, che i n niun modo appaxt iene, n8 pub appartenere a'due s u d d e t t i s i s t e m i [T, 751.
...
Example 12.
6 :\a
'.3:iL
Example 13.
E l =l
Example 14.
IEa
EsE3
Example 15.
Example 17.
Example
THEORIES O F GIUSEPPE TARTINI tered by skilled players to fit the diatonic system E, 951. He returns again to the s y s t e m of the third sound and constructs upon the fundamental b a s s a diatoic scale. Those intervals between the fundamental b a s s and the notes in the Example 11. scale which a r e not included in the senarlp (Ex. 11) a r e considered dissonant [z, 88; T, 1121. This results in the s a m e dissonances a s the ones derived in the Trattato except for the augmented twelfth which does not appear h e r e since i t cannot be considered a diatonic dissonance.
After a discussion of the u s e of the word "harmony" by Greek theorists, Tartini concludes that since the Greeks did not have simultaneous harmony, their meaning for the word must have been that of successive c o n s o u 143; 2, 49ffl. This successive consonant harmony he deharmony rives from the basic ratio of the system, the duple, with i t s harmonic and arithmetic means. This discrete geometric duple can be expressed mathematically a s 6:8:9: 12 and musically a s in Ex. 12 (p. 50). The tones c and c 1 a r e the extremes of the duple, g is the harmonic mean, and f the arithmetic mean 70ffl. Inthe music of his time, successive consonant harmony is r e p r e sented by the three cadences that can be derived from the duple. The f i r s t and more perfect is the harmonic cadence from the harmonic mean to the extreme (Ex. 13). The second and l e s s perfect is the arithmetic cadence from the arithmetic mean to the extreme (Ex. 14). The third and least perfect combines both means and, in s o doing, both natures and is called a mixed cadence by Tartini (Ex. 15). By o r d e r ing the three cadences according to their degrees of perfection and enclosing them with the extremes of the duple, Tartini a r r i v e s a t the basic b a s s progression for the establishment of a key [T, 107; 2,741 (Ex. 16). F r o m the union of successive and simultaneous harmonies, T a r tini derives the scale, which is the basic principle of melody. The union of both harmonies is accomplished through the use of the "organic" formula, which is in turn derived from the senario. The senario a s we have seen, can be expressed in i t s essence in the three sounds of the monochord string, but i t may be expressed in yet another way by taking only the three upper tones (Ex. 17). It is from this form of the that Tartini derives his "organic" formula: 1, 3, 5 with 28; T, 10121. the 8 added on top a s a duplication of the basstone [E, However, this accounts for major harmony only. Minor harmony is derived from the three lower notes of the arithmetic sextuple [B, 61ffl a s in Ex. 18. The union of both major and minor formulae forms a discrete geometric sesquialtera CG with i t s harmonic mean E and i t s arithmetic mean E~ E, 66ffl. This organic formula can be found in three different positions, and here Tartini recognizes the theory of inversion although the word The f i r s t position is 1, 3, 5, 8; the inversion is never mentioned. 1051. Of the three positions, second, 1, 3, 6; and the third, 1, 4, 6 [T,
[x,
e,
Example 19.
l E . E a
Example 20.
Example 21.
Example 2 2 .
Example 2 3 .
Example 25.
Example 26.
THEORES O F GIUSE P P E TARTINI he considers the f i r s t the most perfect in most cases. However, he adm!ts that the second position is m o r e important in the minor triad thah in the major triad [T, 110111. Tartini examines the natural scale produced by such instruments a s the tromba marina, horns, trumpets etc. and finds that the ratios from one tone to another a r e not those used in musical practice. The 4th and 7th notes a r e "out of tune" [T, 95ffl. This leads to his derivation of the scale, o r basic principle of melody, from the union of the two kinds of consonant harmony: simultaneous and successive. The essence of successive consonant harmony is the discrete geometric duple expressed a s in Ex. 19 (p. 52). Since the note C is duplicated, there a r e only three notes that form the essence of successive consonant harmony: C, F, and G. Upon every one of these three notes a chord is built according to the organic formula, and the three resultant chords a r e the b a s i s for the formation of the scale [T, 98; P, 71ff] a s in Ex. 20. The resultant scale is the Ptolemaic syntonic 771. scale (svntonic ditoniaion) h the T r a t t a b , Tartini observes that such a scale needs to be tempered since three major thirds will not reach the octave and four minor thirds will exceed it. He is quite distressed by the fact that most temperaments s e e m to him quite arbitrary; he mentions that Vallotti's temperament13 is the safest, but offers no solution of his own 1001. The relative minor is derived from the l a s t note of Ex. 7b) (see p. 46) which is E ~ and which Tartini considers to contain inherently , both harmonies (major and minor) because, when placed in i t s proper , place a s arithmetic mean between C and G i t f o r m s a major third with G and a minor third with C. Thus the distance between the .first notes of relative scales is proven to be a minor third [T, 110 111. Tartini gives three fundamental basses for the scale and consequently three ways of adding figures to the scale when used a s bass. However, in both t r e a t i s e s he warns against the confusion ofthe melody with the successive consonant harmony ofthe cadences which results in the misuse of the scale a s a b a s s melody [T, 1061. The f i r s t fundamental b a s s for the scale consists in underlining it with the progression of the ordered cadences (Ex. 21). This fundamental b a s s is completely harmonic in the sense that i t is formed totally of major harmonies. The corresponding figuration for the scale would be a s shown in Ex. 22 1071. The second fundamental b a s s is directly derived from the concept Tartini considers the of the discrete geometric duple C, F, G, c.
E,
u,
[x,
13. T a r t i n i r e f e r s t o t h e Padre Francescantonio V a l l o t t i (16971780) who was o r g a n i s t a t t h e Franciscan seminary i n Padua a t t h e time. V a l l o t t i had s t a r t e d w r i t i n g i n 1735 a t r e a t i s e i n f o u r books, a = However, V a l l o t t i f i n i s h e d only t h e f i r s t book. The o t h e r t h r e e were w r i t t e n by V a l l o t t i ' s d i s c i p l e and successor, Luigi Antonio Sabbatini (17391809). The whole work was published a s V a l l o t t i ' s (under t h e t i t l e T r a t t a t o d e l l a modern8 musica) i n 1950. I n t h e Libro second^, Chapter IV, we f i n d S a b b a t i n i l s e x p o s i t i o n of V a l l o t t i l s t h e o r i e s on temperament.
u .
ALEJANDRO ENRIQUE PLANCHART scale a s two similar tetrachords and the fundamental bass underlines this factor (Ex. 23). The corresponding figuration is a s shown in Ex. 24. This fundamental b a s s is harmonic in the sense that it consists of two harmonic cadences (Ex. 25), but i t underlines, through a strong cadence, the arithmetic mean of the discrete geometric duple C, F, G, c [TI 109ffI. Another characteristic of this b a s s which distinguishes it from the f i r s t is the appearance of minor triads. The minor t r i a d s a r e on D and A, which a r e presently those tones on which the relative minors of the tones underlined by the two cadences would be based IT, 1101. The third fundamental b a s s is produced by the observation that, in the f i r s t one given, there is a tritone relationship between the harmonies of the sixth and seventh steps (Ex. 26). This relation is considered by Tartini a s tolerable when the scale is ascending since it is softened by the passing of the scale from a l e s s perfect mean (arithmetic) to a more perfect one (harmonic). However, when the scale is descending, the relation becomes intolerable since there is nothing to soften it. This is inherent in the harmonic nature which tends to a s cend and not to descend D, 1311. T a r t i n i l s solution is quite s u r p r i s ing since here he r e s o r t s to the seventh aliquot part which he calls the "consonant seventhU(Ex. 27). This produces a fundamental bass that is a retrograde of itself, o r , i n T a r t i n i f s words, a completecircle [T, 1321. The Tartini considers the use of the scale in the b a s s unwise. b a s s motion should be determined by the successive consonant harmony while the realm of the scale is in the upper voices.
The derivation of the chromatic and enharmonic genera is contained entirely within the Tratatto. It appears after Tartini gives a brief account of the Greek tetrachords and adds, "1 have read a l l this in Zarlino, a reasonable man and diligent collector of old things. What may have been the reason for dividing the scale in s o many t e t r a chords. . . I do not know!'14 The possibility of interpolating chromatic tones in the diatonic scale in o r d e r to provide for melodic chromaticism finds i t s basic principle in Ex. 7c) (see p. 46), which includes the notes G# and B~ with the help of which the chromatic genus can be constructed 122ffl (see Ex. 28, p. 55). Through the use of G#, Tartini a r r i v e s a t the following chromatic tetrachord (Ex. 29). This tetrachord takes place in A minor, relative minor of C major,which is the tonality produced by the (given) senario [T,1261. Thus, Tartini relates chromaticism to the minor mode. To derive the enharmonic tetrachord the seventh aliquot part is interpolated in the scale f r o m a to d f which includes the bb (Ex. 30) [T,1271. Tartini admits that some theorists will complain that this new sound is not prepared and not resolved by descending. To this he answers that the seventh partial, having the same third sound a s the f i r s t six and being moreover the true harmonic mean of the fourth a  d f
[x,
[x,
Example 28.
Example 29.
Example 30.
Example 31.
Example 32.
Example 3 3 .
Example 34.
Example 3 5 .
THEORIES O F GIUSEPPE TARTINI is consonant [T, 1281. He claims that the minor seventh before cadences need not be prepared since the difference between the consonant seventh and the t r u e minor seventh is s o s m a l l that the minor seventh is thought of a s a n almost consonant interval E, 1291. His hArmonization of the scale using the seventh partial (Ex. 27, p. 52) b e longs therefore to the enharmonic genus. T a r t i n i considers the augmented twelfth a chromatic dissonance, since the G o c c u r s i n the chromatic tetrachord, and resolves i t a s shown in Ex. 10 (p. 49) [T, 1301. The minor seventh, i n contrast with the minor fourteenth, which is considered a diatonic dissonance, is r e garded a s a n enharmonic dissonance and given the resolution shown i n 1311. Ex. 31 Tartini considers three other intervals and t h e i r inversions: the diminished third, inverted a s the augmented sixth; the augmented s e c ond, inverted a s the diminished seventh; and the diminished fourth, inverted a s the augmented fifth. All these intervals a r e derived f r o m the interpolated diatonic scale of Ex. 28 [T, 1581. In accordance with the derivation, Tartini classifies the augmented second and the diminished fourth a s chromatic intervals and the diminished third a s a n enharmonic interval [T, 1591. He s t a t e s that the diminished seventh is treated a s a dissonance although it is not always prepared. The augmentedfifth (in contrast to the augmented twelfth) is treated a s a nonessential combination since the tone that produces i t is not prepared and it a s cends (Tartini u s e s the word discordanza for it) a s in Ex. 32. The diminished fourth is treated a s a consonance (Ex. 33). The augmented sixth is a l s o treated a s a consonance [T, 157581. The use of such intervals can be demonstrated i n the scale shown in Ex. 34. This scale is based on Ex. 7c) (p. 46). Tartini observes that the s c a l e consists of two equivalent tetrachords, da and adl and gives a n example of a s m a l l piece composed using the scale (Ex. 35) IT, 159ffl. Tartini claims that i n Ex. 35 the only dissonances a r e the 43 suspensions a t the cadences. The only other chord that could be considered dissonant cannot be dissonant in the inversion in which i t is written since the D# cannot be distinguished, even i n the generation of the third sound, f r o m the t r u e harmonic note ~ b (Ex. 36). The funda' mental f o r this chord should then be a s shovl?l i n Ex. 37. Tartini concludes by pointing out the possibilities of such harmonic m a t e r i a l s in dramatic music [_T, 162ffl.
[x,
Rhythm and m e t e r , like harmony, a r e derived by Tartini f r o m the basic ratio of the system, the duple. This ratio provides duple meter. The combination of the duple and the sesquialtera f o r m s the b a s i s for triple m e t e r 1141. In the process of applying m e t e r to the successive consonant harmony and to the melodic voices, rhythm is produced in long and s h o r t accents (long and s h o r t notevalues) which correspond to the long and s h o r t syllables of prosody [T, 1151. However, Tartini observes that musical rhythm, particularly i n triple m e t e r , is not a s t r i c t application of syllabic values to the pitches, but a stylization of the syllabic values which he illustrates a s shown i n Ex. 38. The f i r s t of the two settings is a s t r i c t application of the syllable values that r e s u l t s i n confused m e t e r , while the second, a stylization
[x,
Example 39.
Example 40.
Example 4 1
Example 42.
THEORIES O F GIUSEPPE TARTINI of the syllabic values is clear rhythmically and metrically [T, 1391. Example 38.
Tartini mentions in passing that harmony seems to have a definite influence in determining the long and short accents in music, and he notices the association of downbeats with consonance [T_, 117ffl. In the discrete geometric duple, expressed a s C, F, G, c, T a r tini sees a basis for the three clefs 1201 and the basic tonal realm of tonic, dominant, and subdominant D, 146ff; 2, 98ffl. Tartini treats modulation very briefly. A major key, because of i t s harmonic nature will tend to modulate towards the dominant, a minor key, because of its arithmetic nature will tend to modulate towards the subdominant. Tartini warns against "leaps" in modulation such a s going from a key in flats to a key in sharps and requires that the principal key be made clear a t the opening and closing of any piece [T, 147ffl.
[z,
Tartini's notions about history, and about Greek theory, a r e scattered throughout the two treatises. Greek theory, for him, consists in a statement of the different divisions of the tetrachord [T, 1211 and an explanation of the adaptation of the tetrachords to the strings of the kithara L_P, 531. Tartini believed that the Greeks had an octavespecies based on the arithmetic and the harmonic divisions of the octave thus forming two tetrachords (Ex. 39). Because of this, he considers his diatonic ' system a s essentially the same a s the Greek [P, 49ff; , T 143ffl. Since the Greeks did not have simultaneous harmony, the consonant intervals were the leaps of the fourth, fifth, and octave, while any interval smaller than the fourth was considered dissonant [T, 1431. Tartini considers the ecclesiastical modes a s combinations of fourths and fifths. The authentic modes, in which the fifth is at the bottom, he considers a s harmonic modes, while the plagal ones with the fourth on the bottom a r e considered arithmetic modes. However, he finds the organization of the modes unsatisfactory since the harmonic and arithmetic divisions of the octave do not fall within the same octave (Ex. 40). Tartini then proposes what he considers an improved and "natural" organization of the modes based on the harmonic and arithmetic divisions of the octave cc', ff \ and gg', which a r e derived from the discrete geometric duple (Ex. 41). To these six modes, Tartini adds another s e r i e s of modes based on the arithmetic and harmonic divisions of the octaves Aa, dd', and ee' [T, 1371 (Ex. 42). Tartini had little understanding of the organization of the modes and his new modal system i s nothing but the major and minor scales that form the common tonal realm of the music of his time. In the Tratthto he uses hexachord nomenclature to refer to any given note; this i s dropped in the Princiui. However, in the Rismsta, published the same year a s the Principj, Tartini attacks S e r r e ' s use of
ALEJANDRO ENRIQUE PLANCHART the syllable gi which, according to Tartini, shows S e r r e l s ignorance of "the masterful Italian solfeggio:' that is, hexachord solmisation [R, 70711.
Tartini's opinions on the affections a r e found in the fifth chapter of the T r a t t a t ~ . Regarding the affective power of music, Tartini is quite skeptical. He doubts that music, a s practiced in his time, can express any precise affection. Greek music could do s o since it was monophonic and since particular passions had their own ranges, speed, dynamic qualities, etc. A single voice, with perhaps a single instrument, can most effectively communicate a n affection [T, 139ff, 149ffl. Therefore, some of the old Italian d r a m a s which consisted almost entirely of recitative could most clearly communicate a n affection [T, 1351. However, in most music of his day which employed many voices this would hardly be possible because the different pitches, motion, etc., of the different voices would obscure any attempt to communicate any given passion clearly. The effect will be a very general one leaning toward one affection o r another 1411. The two best means by which to induce an affection in a h e a r e r , Tartini concludes, a r e melodic ornaments and harmony itself 14891. Tartini r e m a r k s on the relation of major harmony to f i r e and joy on one hand, and minor harmony to melancholy, languor, and sweetness on the other. He claims that a harmonic cadence c a r r i e s the characteristics of the major harmony while a n arithmetic cadence has the characteristics of minor harmony. Tartini feels that a study of the affective value of the intervals could begin with this and continue to account for those intervals that a r e better suited to each voice. He observes that while octaves and fifths sound quite satisfactory in the bass, even if used constantly, they a r e quite repugnant in the upper 15341. Such considerations, according to Tartini, would voices require another treatise written by one who, like the Greeks, would be a musician, philosopher, and poet.
[z,
[x,
m,
Bibliography
P. Del principi dell'armonia musicale contenuta nel diatonic0 genere, dissertazione di Giuseppe Tartini. Padua: Stamperia del Seminario, 1767.
L. Lettera del defonto Sig. Giuseppe Tartini alla Signora Maddalena Lombardini inserviente a d una importante Lezione Der i Suonatori d i Violino. Venice: Colombani, 1770. Translated into English by Dr. Burney. London: Bremmer, 1779.
R. Risposta di Giuseppe Tartini alla crittica del di lui Trattato di mu
sics di Monsieur l e S e r r e di Ginevra. Venice: Antonio Decastro, 1767.
THEORIES OF GIUSE PPE TARTINI T. Trattato di musica second0 la Vera scienza delllarmonia. Padua: Stamperia del Seminario, 1754. 1. Euler, Leonhard. Tentamen novae theoriae musicae ex certissimis harmoniae principis dilucide expositae. Petropoli [St. Petersburg? 1: Typ. Academiae, 1739.
..
2. Fktis, F. J . Biographie universelle des musiciens et bibliographie gdndsale de la musique. Volume VIII. Brussels: Leroux, 183544.
3. Rameau, Jean Philippe. Ddmonstration du principe de l'harmonie Paris: servant de base 3 tout l l a r t musical thedrique e t pratique. Durand, 1750.
4.      Nouvelles rdflexions s u r l e principe sonore [bound with the Code de musique pratique]. P a r i s : Imprimerie Royale, 1760.
5. Shirlaw, Matthew. The Theory of Harmony. London: Novello and Co. [1939]. 6. Stillingfleet, Benjamin. The Principles and Power of Harmonx. London: J . and H. Hughs, 1771. 7. Vallotti, Padre Francescantonio [and Sabbatini, Luigi Antonio]. Trattato della moderna rnusica. Padua: I1 messaggiero di S. Antonio, 1950.
8. F e r r i s , Joan. "The Evolution of Rameau's Harmonic Theories:' Journal of Music Theory, III(1959):23156.