|
|
© 2004 Tonalsoft Inc. |
|
|
© 2004 Tonalsoft Inc. |
Paper for the conference of the Center François Viète
Problems of translation in the 18th century
Nantes, January 17, 1997
In RED, translation from Latin by Monzo. My translations are more
literal than Prof. Bailhache's; they have been retained where there
may be a subtle difference in meaning.
In GREEN, Monzo's English version of Prof. Bailhache's French
translation of Euler's Latin (contributed specifically
for this translation).
In MAGENTA (at the end), the translator's notes, and
observations from others.
By the middle of the 1700s, at the time when the music known as classical acquires its letters of nobility with Bach (1685-1750), Handel (1685-1759), Rameau (1683-1764), Haydn (1732-1809), Mozart (1756-1791), etc., two particularly famous mathematicians, Euler and d'Alembert, produce theories of music. This fact is obviously not the result of a "historical chance". It represents on the contrary the prolongation of one tradition.1 By the end of the previous century, several scientists had already become intrigued and put their attention on the same subject: Descartes (Compendium musicae), Galileo (end of the 1st day of Discorsi), Mersenne (the enormous work of the Universal Harmony), Leibniz "as an amateur".2 But in the 1700s, it is the music itself which changes, the traditional harmony dethroning the medieval counterpoint definitively. Thus this change had to be explained, to make the theory of it, and the task could inspire, rightly, any scientist set on music.
Euler (1707-1783) is 24 years old when he writes, in 1731, his
Tentamen novae theoriae musicae ex certissimis harmoniae
principiis dilucide expositae (A attempt at a new
theory of music, exposed in all clearness according to
the most well-founded principles of harmony).
It is a work of 263
pages, written in Latin, which will be published only in 1739. It was
translated into French one century later with the edition of Brussels
of works of the mathematician. I will presently say some words on the
quality of this translation.
In 1739 Euler is already known as a mathematician and is in
St Petersburg, where he will soon occupy the Chair of mathematics.
He is extremely interested by all that touches on music.
He published in Basle, in 1727, a "thesis on sound" where he
compares the sounds produced by the vibrating cords with those
generated by the wind instruments. And about 1726 already, Euler had
projected the plan of a considerable work on music. Apart from
the fact that the sounds were to be noted there by sequence numbers in
the scale, the object of study remained close to musical realities.
The last section, for example, was to analyze the various kinds of
pieces of music (saraband, courrant, etc). But the departure of the
mathematician for St Petersburg (1727) and his other work prevented
him from proceeding in this initial way. It is finally a much
more mature work which was born in 1739.3
Euler remained all the rest of his life interested by
the questions of music. He brought precise details to his new theory
with some articles4
showing, as we will see, that he could take
account of criticisms which one had addressed to his conceptions.5
Plus, I even believe and I will try to show it that he gave up one of
the fundamental assumptions of his theory, without acknowledging it
expressly. Passing on the first chapter (on sound and hearing), I
come immediately to the bases of Euler's musical theory
(he talks in the second chapter "de suavitate et principiis harmoniae").
How to explain that certain sounds appear pleasant and others
not? It seems paradoxical to find a rule, since everyone does not
like the same things, or the taste of the same person can evolve.
Can one lock up an art, like music, in laws? Euler answers that,
as for all the fine arts, it is necessary to trust the opinion of
enlightened people,
therefore in music those whose ear was exerted and who will be
able to see the right laws which dictates nature:
"Sed Musicum similem se genere oportet Architecto, qui plurimorum
perversa de aedificiis iudicia non curans secundum certas et in
natura ipsa fundatas leges aedes exstruit; quae etiamsi harum rerum
ignaris non placeant, tamen, dum intelligentibus probentur, contentus
est." [OO, 3a, I, 224]
["We will answer that the musician must
act like the architect who, worrying little about the bad judgements
which the ignorant multitudes pass on the buildings , builds according
to unquestionable laws based on nature, and is satisfied with
the approval of the people who are enightened in this matter."
This preliminary remark being made, Euler affirms that any
pleasure comes from the perception of the perfection,
any normally-constituted person seeking this:
"Certum est enim perceptionem perfectionis voluptatem
parere hocque omnium spirituum esse proprium, ut perfectionibus
detegendis et intuendis delectentur, ea vero omnia, in quibus vel
perfectionem deficere vel imperfectionem adesse intelligunt,
aversentur." [OO, 3a, I, 225]
["It is certain that any perception of perfection
gives birth to pleasure and that it is a property common
to every spirit, as much to delight itself with the discovery
and contemplation of a perfect object, as to avoid what misses
perfection or exposes its imperfections."
- Bailhache/Monzo]
However, as the example of a clock shows us,
perfection is reduced to order.
"Contemplemur exempli causa horologium, cuius finis
est temporis partes et divisiones ostendere: id maxime nobis
placebit, si ex eius structura intelligimus omnes eius partes ita
esse confectas et inter se conjunctas, ut omnes ad tempus exacte
indicandum concurrant."
[Let us take for example a clock,
whose destination is to mark divisions of time;
we will like it with the highest degree, if the
examination of the structure makes us understand
that the various parts are laid out and combined
by it in such a way that all contribute to indicate
time with exactitude.
"Vicissim igitur etiam intelligitur, ubi sit ordo, ibi etiam esse
perfectionem et legem regulamve ordinis respondere scopo perfectionem
efficienti." [OO, 3a, I, 226]
[Reciprocally, one can say that
where there is order, there is perfection, and
that the rule or law of the order corresponds
to the goal which marks the perfection.
- Bailhache/Monzo]
This notion of order is the key to the theory.
However, it still remains to adapt it to the field of music.
The order of things can be perceived in two ways:
either one knows already the rules and one recognizes the presence
of it immediately; or the rules are unknown, and one must then
seek them, to reveal them (detego); music conforms
to this second process:
"Duobus autem modis ordinem percipere possumus;
altero, quo lex vel regula nobis iam est cognita, et ad eam rem
propositam examinamus; altero, quo legem ante nescimus atque ex ipsa
partium rei dispositione inquirimus, quaenam ea sit lex, quae istam
structuram produxerit. Exemplum horologii supra allatum ad modum
priorem pertinet; iam enim est cognitus scopus seu lex partium
dispositionis, quae est temporis indicatio; ideoque horologium
examinantes dispicere debemus, an structura talis sit, qualem scopus
requirit. Sed si numerorum seriem aliquam ut hanc 1, 2, 3, 5, 8, 13,
21 etc. ascipio nescius, quae eorum progressionis sit lex, tum
paullatim eos numeros inter se conferens deprehendo quemlibet esse
duorum antecedentium summam hancque esse legem
eorum ordinis affirmo."
["There are two ways in which we are able to perceive order;
the one, in which the law or rule is already known to us,
and to that matter being proposed, we simply examine;
the other, in which the law before we were ignorant and
out of the regular arrangement of parts we inquire,
so that that we may construct laws. The example of the clock given
above belongs to the first sort; one already knows the goal
or law of the regular arrangement of the parts, which is the indication of
a time value; thus on examining a clock we only need to discern,
whether its structure is such, as this goal demands.
But if the number series somehow were this 1, 2, 3, 5, 8, 13, 21 etc.
and we had to consider unaware, what is the law of their progression,
then gradually the comparison of those numbers among themselves
will allow anyone to discover that each is the sum of
the two previous and consequently thus is the law of their order confirmed.
We can recognize order in
two manners. When the law which is the reason
is known to us, it is sufficient to examine whether
the object under consideration satisfies it.
But if we lack this data, it is necessary to seek
to discover, in the organization of the parts of the
object, the law which governs their arrangement;
the recognized law, the order, will be the consequence.
The clock quoted previously offers an example of
the first case: the goal or the law of the provision
of the parts is known there, it is the indication of
time. Thus, when we examine a clock, we have
only to check if its construction answers this goal.
But if we have to consider, for example, the series
of numbers 1, 2, 3, 5, 8, 13, 21, etc, without knowing
which is the law of their increase, the comparison of
these numbers makes us soon discover that each
one is the sum of both which precede it; we
consequently know the law of formation of the series,
i.e., we know the order which reigns there.
It remains to determine of what consists the order of sounds.
It is primarily of two types: according to what is called today
the height or pitch
(low or high), or according to the duration.
One could also speak about order according to the intensity,
but this order is not very quantifiable and rather random,
since a composer can indicate only in a vague way the intensity
of the sounds.
And finally, recognizing preeminence of height
over that of the duration, since that one is measured by the
frequencies
of vibration, Euler brings back the evaluation of the musical pleasure
to the arithmetic measurement of the proportions attached to the sounds.
This 'metaphysics' of music is not completely new.
The 'result' of Eulerian metaphysics is found already in Antiquity
with the Greeks, who had based musical science on the theory of the
proportions (cf. the Division of the Canon of the Euclidean school).
In the same way, the idea that the music consists of the indirect
(or more precisely unconscious) perception of the
ratios of the
sound frequencies is stated by Leibniz:
"musica est exercitium arithmeticae occultum
nescientis se numerare animi."6
But, in spite of these identical themes,
the differences should be stressed which mark the originality
of Euler. Initially, as one will soon be able to judge,
the theory of our mathematician exceeds by far the simple
consideration of the ratios of the frequencies of
two sounds. Then, contrary to his predecessors, who had also
founded their theory on the proportions without any
explanation, Euler outlines a philosophical argumentation,
in which by the proportions one arrives at the musical pleasure,
via the order and the perfection. As for the Leibniz source,
one can undoubtedly recognize from it the influence in the
distinction which Euler makes between the two modes of
perception of the order. But it will be noted that, contrary
to the Master of Hannover, he does not venture to speak
about unconscious perception; perhaps he does not admit the
concept of it, or well, quite simply, he prefers to apply its
engineering to develop mathematical calculations rather than
philosophical considerations.
That said, let us examine how Euler evaluates the order of the
sounds according to their height. By the intermediary of the
vibrations of the air, the musical instruments produce regular
blows or shocks [ictus] on our tympanum. With a preoccupation
for clarity, Euler visualizes these blows, in the case
of the simplest chords of two sounds, by the following figures:
- Bailhache/Monzo]
["'Let us contemplate as an example the mechanism of a clock,
whose goal is to show the parts and divisions of time:
it gives us maximum pleasure, if out of that structure we comprehend
how all the parts are intertwined with each other,
and how their actions all converge to indicate the exact time.
"Ex hisce sequitur, in qua re insit perfectio,
in eadem ordinem necessario inesse debere." [OO, 3a, I, 225]
From this it follows, in that which incorporates perfection,
likewise to order must it necessarily be indebted"
- Monzo]
T2
Thus, in any thing where
there is perfection, there is necessarily also order.
- Bailhache]
"Posterior modus percipiendi ordinis ad musicam praecipue
spectat" [OO, 3a, I, 226]
The second manner of recognizing order is present especially in music."
- Monzo]
The second manner of recognizing order is present especially in music."
- Bailhache]
etc...[OO, 3a, I, 231]
This is the place to say some words on the French translation of 1839, since as well the principal topic of our conference [for which Professor Bailhache submitted this paper] is that of the scientific translation. I will be satisfied at a glance on the second chapter, "De suavitate et principiis harmoniae", translated by "On the charms and principles of the harmony". The suavitas will be discussed throughout the work, because one of the major goals that Euler proposes to reach is that to define a scale of "kindness" of the sound aggregates, which he names "suavitatis gradus", "degrees of softness".T4 Useless to notice, under these conditions, that the term of "charm" is not very well selected. Wrote Euler:
"Duobus sonis propositis percipiemus eorum relationem, si intelligamus rationem, quam pulsuum eodem tempore editorum numeri inter se habent; ut si alter eodem tempore 3 pulsus perficiat, dum alter 2, eorum relationem adeoque ordinem cognoscimus observantes hanc ipsam rationem sesquialteram. Similique modo plurium sonorum mutuam relationem comprehendimus, si omnes rationes, quas singulorum sonorum numeri vibrationum eodem tempore editarum inter se tenent, cognoscemus." [OO, 3a, I, 228]
which means [in French]:
"Two sounds being given, we will know the relation which exists between them, if we can grasp the ratio of the number of vibrations carried out for one, with the number of vibrations carried out for the other in the same time. For example, if there were 3 vibrations for the first, while for the second there would be 2 of them, we know their relation and consequently their order, by observing the ratio of the numbers 3 and 2 which is 3:2." [ OCF, 5, 27 ] 7
We see initially that "pulsus" is rendered by "number of vibrations", which corresponds obviously to a desire of modernization of the theory. Literally "pulsus" is an impulse, a shock, and the concept that Euler keeps in mind is certainly that of the shocks (tiny and fast) that the sound bodies produce on the tympanum of the ear by the intermediary of the medium which transmits the sound. In fact, as I explain further on, Euler takes again what one calls now the theory of the coincidence of blows. However, in the following sentence, Euler himself speaks about "numeri vibrationum", which justifies to a certain extent the step of the translator.
We note then that "rationem sesquialteram" is quite simply translated by "3:2". It is again about a modernization, because the sesquialter proportion indicates in Euclid (which Euler takes as a starting point here at least for the terminology) any ratio of the form 1 + 1/n, therefore 3/2 for n = 2.
Let us throw a last glance on the following paragraph:
"Magnum quidem extat in sonorum rationibus percipiendis subsidium, quod singulorum plures ictus percipimus saepiusque eos inter se comparare possumus. Idcirco multo est facilius duorum sonorum rationem discernere audiendo quam duarum linearum eandem rationem habentium intuendo." [OO, 3a, I, 228]
[in French:]
"The appreciation of the relationship between the sounds, is singularly facilitated by this circumstance whereby we perceive several vibrations of each one, and whereby thus we can, for their length of time, compare them several times one with the other. It is thus much easier to distinguish by hearing the ratio of two sounds, than to recognize by the [sense of] sight the same relationship between two lines." [ OCF, 5, 27-28 ]
Again the translation modernizes the term of "ictus" in "vibrations", which however indicates very exactly the "blows" of which I have just spoken.
This rapid examination could lead one to think that the French translation leaves much to be desired. It is certain that it would deserve to be revised to become more precise. However, to judge some by what I know, I believe that one must say all the same that the translator did an intelligent and overall faithful job.
Let us return now to the theory of Euler. The mathematician reasons in the following way.
Then noticing, that 1:5 must be more complex than 1:8 (= 1:23), which has degree 4, Euler allots degree 5 to it and deduces by induction that for a prime number p, 1:p is degree p. Then, this time for p unspecified:
"[...] si ratio 1:p ad gradum, cuius index sit m, referatur, rationem 1:2p ad gradum m + 1 pertinere, 1:4p ad gradum m + 2 et 1:2np ad gradum m + n. Multiplicato enim numero p per 2 ad rationis perceptionem requiritur praeter perceptionem rationis 1:p bisectio aut duplicatio, qua ut simplicissima operatione gradus suavitatis unitate evehitur."
["... if the ratio 1:p to the degree, whose index is m, it is being ascribed, to the ratio 1:2p the degree m + 1 belongs, 1:4p to degree m + 2 and 1:2np to degree m + n. In multiplying a number p by 2 the perception of its ratio requires beyond the perception of the ratio 1:p a bisection or doubling, which is how the simplest operation of one unit of the degree of softness is carried out" - Monzo]
[... if the 1:p ratio belongs to the degree whose index is m, that of 1 with 2p belongs to the degree m + 1, that of 1 with 4p with the degree m + 2, and in general ratio 1 with 2np with the degree m + n; because the multiplication of the number p by 2 gives a ratio only twice as difficult to recognize than the ratio of 1 to p, and increases only one unit the number which expresses the degree of softness of this last. - Bailhache]
And Euler continues by calculating the degree which it is necessary to attach to 1:pq, p and q being again prime:
"Simili modo determinare licet gradum suavitatis rationis 1:pq, si p et q fuerint numeri primi; nam ratio 1:pq eo magis est composita quam 1:p, quo 1:q magis est composita quam 1:1. Ergo rationis 1:pq gradus cum p, q et 1 debet proportionem arithmeticam constituere, unde erit igitur p + q - 1." [OO, 3a, I, 232]
["A similar method is used to determine the degree of softness of ratio 1:pq, if p and q are prime numbers; for ratio 1:pq is more complex than 1:p, while 1:q is more complex than 1:1. Therefore the degree [of softness] of ratio 1:pq with p, q and 1 owes its constitution to the arithmetic proportion, from whence it will be therefore p + q - 1." - Monzo]
[In the same manner one determines the degree of softness of the 1:pq ratio, by supposing that p and q are prime numbers; because between 1:p and 1:pq there is the same difference in composition as between 1:1 and 1:q; consequently the degree of softness of 1:pq must form with degrees 1, p, and q an arithmetical ratio; it will be thus p + q - 1. - Bailhache]
Finally, by observing several times this rule, one can generalize, which produces a formula that Euler does not write, but which amounts to:
the numbers pi being prime and ki being unspecified positive exponents of these numbers.
The cases of more than two sounds are easily reduced to this
formula. Thus, four sounds in the 1:p:q:r
ratios, with p, q, r prime,
will be compared to 1:pqr 8.
When the numbers
of vibrations are not prime, the procedure is a little more
complicated. Let us consider for example the sounds corresponding
to 1:pr:qr:ps (p, q,
r, s being prime). p and r intervene twice,
but being confused by the ear, they should be taken only one time;
so that the unit will be compared to 1:pqrs, which is lowest
common multiple (LCM) of the factors involved.
Remaining to be considered are the
cases where one of the numbers of vibrations is not 1. The
fifth 3:2 is an example. One will start then by if required
reducing the numbers by dividing them by their highest common
factor (HCF) (ex: 4:6:8 give 2:3:4), then, as presently, one
will take their LCM. Thus the fifth amounts to the LCM 6
and is thus fourth degree.
We will stop here in this talk for right now, to note several points.
The idea to measure the degree of softness of
a chord of two sounds by the scheduling of the blows
which one imagines to be the characteristic of a musical sound is not
new. Quite to the contrary, it is well accepted so that it
is agreed to call it today the theory of the coincidence of blows,
i.e., this theory which represented what one could call, within the
meaning of Kuhn, the "normal" theory at the time of Galileo and
Mersenne. According to this theory, a chord is all the more
consonant when the "coinciding" blows resulting from the two
sounds are in higher proportion in the whole of the produced blows.
But, in truth, in the 1600s, calculations are never taken out
to very many places and, what is more, an extremely strange error is
systematically made. Instead of taking the proportion in question,
one considers that of the blows coinciding compared to the number
of blows of the highest sound only. 9
A second remark relates to the step of Euler. It is rather
surprising. All is due in fact in the degree of softness
p + q - 1 attached to the 1:pq ratio,
or, identical in meaning, with two sounds p and q
(these numbers being relatively prime).
We saw that Euler arrives at this result by a kind of
analogical reasoning: the 1:pq ratio "exceeds" 10
1:p as 1:q exceeds 1:1.
It is necessary thus that its degree
is an arithmetical ratio of p, q and 1, which produces
logically p + q - 1. One cannot say that this
deduction enjoys a great rigour and carries the conviction,
the more so as the former result (the degree p of 1:p)
is itself obtained by a synthesis of the same order of probability.
However I say that the result p + q - 1 was directly accessible,
in a way much more rigorous. It was enough for that to correctly
calculate the degree of softness of the p:q
chord in accordance
with the theory of the coincidence of the blows.
If p and q are
two distinct prime numbers (or even relatively prime), it is
quite clear that during one period none of their blows will coincide,
except those of the beginning and the end of the period. There will
be thus one blow coinciding per period (the one period end being
the beginning of the following period). And the total number of
blows is the sum of all the blows, p +q,
decreased by one unit since
the two simultaneous blows are perceived like one only (on the
example of the 5th, fig. 5 above, one sees there that it has a
coincidence and four blows [including this coincidence] per period,
which corresponds with 3+2-1). Coincidence being always single,
p + q - 1 will be able to constitute an opposite
measure of the
softness of the chord of the two sounds p and q.11
One can notice as Yves Hellegouarch did 12
that the
S(kipi - ki) + 1
is a morphism of the
monoid (N, x) in (N, +), which is only one precise manner
to indicate the nature of the function. According to an
oral comment of Jean Dhombres, it should also be noted that
it is the only function which respects the prime numbers
(i.e. for p prime, the degree of softness of 1:p is p).
But, given the empirical way in which Euler builds
this function, it seems extremely improbable that
he was perfectly conscious of the unicity of his solution.
And Hellegouarch, again, is not wrong to say that "many
other functions could be proposed".13
At all events we now have the fundamental elements of the method
used by Euler to build his musical theory.
The question of consonance of
a complex of sounds reduces itself, after division of the components by
their HCF, to calculation of degree of softness of their LCM by
the formula described above. Euler's Essai
is enriched by very many
and often very copious tables; the first that he presents to us is
thus that of the classification of the LCMs according to the the
first sixteen degrees of softness [ OO, 3a, I, 234 ]: I now will focus my attention on the principal part of Euler's
work, namely his mathematical study of harmony, without
going however into all the details of the rules of composition
with which he graces us, nor to tackle the questions of duration
of sounds.
Study of chords of two sounds
The calculation of the LCM attached to each chord,
that Euler calls his exponent, provides, by the preceding
table, a principle of classification of consonances
giving place to a new table ([ OO, 3a, I, 249 ]; I note in
boldface characters the ratios which correspond
to the fundamental consonances and I underline those
which correspond to the traditional dissonances):
14
The classification of the fundamental chords according to Euler
is thus:
For the consonances, Euler finds thus one of the two classifications that Mersenne had already proposed in the Harmony universelle.15 And as Mersenne had also advanced, Euler affirms that there is no clear border between consonances and dissonances; these last are only "bad" consonances, of a high degree. The table clearly shows it, since it places the two minor consonances (the third and the sixth) at the same degree (VIII) as that of the major tone, which is not absurd from the point of view of perceived reality. We will see what becomes of this idea in the last studies of Euler on music.
Study of the chords of more than two sounds
Up to now there is really nothing new.
But the bases of the theory conceal other things well!
As one immediately notices it in the table of the chords, to a
same "exponent" (the LCM) correspond several groups of possible
sounds. Thus, if one considers chords of more than two sounds,
the degree of softness will not change when one adds sounds
corresponding to dividers of the exponent. 1:2:3:6 is not more
"complicated" that 1:6 or 2:3, since all three have 6 for LCM.
Consequently Euler defines
the concept of a complete chord: a chord
will be such if one can add no note to it without its degree becoming
any higher, therefore without its exponent not becoming more complex.
From this it follows that a chord is complete
if it includes all the dividers of its exponent.
According to Euler, the ear will then have the impression of
plenitude in this degree of chord.16
Here will begin the
surprises for somebody who refuses to entirely engage himself
with the fruitfulness of the spirit of the great mathematician.
Indeed let us try "to return" to the music of the time of Euler.
The first chord which one learns (even still today)
is the major triad, the most consonant of all, represented for example
by the notes C E G. Not to complicate the situation, let
us not add the higher octave to it (C),
although that is completely
common. These three notes correspond to the three numbers 4, 5, 6,
whose LCM is 60. That reveals that it is already of the 9th degree and
that the complete chord will be that of the 12 numbers:
1:2:3:4:5:6:10:12:15:20:30:60, i.e. of the twelve notes
C1
C2
G2
C3
E3
G3
E4
G4
B4
E5
B5
B6.
Here is thus a chord which Euler claims that it will be
on the same equal footing of softness as the major triad, a
chord which is spread out over six octaves and which in addition
to the notes C, E, G, will include also the
major 7th B of fundamental C!
It will be more complete, certainly,
but will it be really more pleasant? Will it be also
more practicable? One can doubt it. However, on closer
reflection, that is not as absurd as it may seem.
Because if one refers to the modern theory of the partial sounds
of Helmholtz, it is perfectly exact that in the production of the
sounds C, E, G,
one also hears the higher partials of the same name,
including the B which is the 5th of E and
the major 3rd of G. The whole thing is to know
the intensities; and, actually, the notes B4 B5 B6
will be very weak.
If one introduces "manually" a C2, the chord
C E G C
will correspond to numbers 4, 5, 6, 8, of LCM 120. Complexity
is then raised one degree, the last note of the complete chord
being a B of the seventh octave.
This addition of theoretical complexity
does not correspond to reality accurately. In fact, to add or not
the octave C hardly changes the quality of the chord, with
which Euler himself agrees when he analyzes the genera
T5 (cf. below).
Study of the successions of chords
This chapter, the fifth of Tentamen, testifies, as if it were
necessary, to the originality of Euler in music-theory.
No one who had claimed before him "to explain" the consonances
by the chord of the blows of the musical sounds had gone
further in this way. However music is not only the pure and
simple "pictorial" exposure of several harmonies; what is
perhaps even more significant, it also consists
in a succession, regulated
in time, of a series of harmonies.
Euler's idea is simple: it is the same one as for the
chords taken separately, i.e. the softness of a succession
will depend on the
order which it contains.
This softness will thus be evaluated by the
degree of exponent (the LCM) of the whole complex of
sounds from the two successive chords,
as if they were emitted at the same
time. Two reservations, however, will have to be made.
Initially, as
sounds emitted at the same time sound generally harder than if they
are followed, one will admit worse degrees (i.e. higher) than in the
case of simple consonances.
Then, as it does not act more than one isolated chord, from the
point of view of a succession, the possible common factor of the
numbers of vibrations of a chord is no longer neglected
(i.e., for example, that 2:6:10 are not inevitably any more equivalent
to 1:3:5). That complicates things. I prefer to leave the question
aside in this rapid survey,
17
restricting myself to saying that the
relevance of the interpretation of Euler remains problematic: for
example, the resolution of a dominant 7th chord
on the triad of tonic is not clearly "explained", since the
exponent of the second chord is entirely included in that of the
first.18
The genera of Tentamen
Euler gives to his method a universal value. For a succession of more
than two chords, for a whole piece of music (!) it is the same
principle of evaluation: to calculate the exponent of the whole complex
and from it to deduce the degree of softness.
But any piece is played in respect to certain genera.
The concept of genus is essential, because the instruments with fixed
sounds can produce only one limited number of notes, whose whole
will constitute a genus and will correspond to a certain exponent.
More exactly, according to Euler, the concept of genus rests on
that of octave: a genus is a series of notes which one finds
distributed periodically according to several octaves. Here,
Euler gives no other justification than that of, a posteriori,
design of the existing instruments,
19
which is a defect of his
theory which should obviously be as a priori as possible.
This additional principle gives the result that the exponent of a
genus will be in the form 2mA,
where A will be a product of prime-factors
not containing 2, in other words an odd number. But what,
more precisely, will A be?
Euler answers this question only after having
reviewed various genera of the form
2m.3n.5p.
He quotes Leibniz who,
in a letter to Christian Goldbach, said already:
"Nos in Musica non numeramus ultra quinque, similes
illis populis,qui etiam in Arithmetica non ultra ternarium
progrediebantur, et in quibus phrasis Germanorum de homine simplice
locum haberet: Er kan nicht über drey zählen." 20
["In music we do not go beyond the number 5,
similar to those people, who also in arithmetic have not advanced
beyond the ternary,
and in that German phrase about primitive man of which it is the origin:
They cannot count over 3." - Monzo]
Euler develops a genus all the same where the factor 7 intervenes, but
by way of pure scholarly hypothesis (one will further see his opinion
on this question will change completely in the 1760s):
"Atque sane difficile esset in musicam praeter hos
tres numeros alium, puta 7, introducere, cum consonantiae, in quarum
exponentes septinarius ingrederetur, nimis dure sonarent harmoniamque
turbarent." [OO, 3a, I, 332]
["And it certainly might be difficult in music to introduce
beyond these three numbers [2, 3, 5] another, namely 7,
since [those] consonances, in which exponents of 7 may be engaged,
would sound too harshly and would disturb the harmony."
- Monzo]
And it would certainly be difficult
to introduce in music another number besides those three,
i.e. 7, since the consonances into which the exponent
seven is introduced sound too hard and disturb the harmony.
- Bailhache]
The combination of powers of 3 and 5 produce 18 genera, of which
Euler retained five which have been or still are
of use, eliminating four because they are contained in the others
and rejecting nine either because they are either too simple,
or on the contrary too complicated,
i.e. that they give some sound which is too unpleasant,
either of the two. The
morals of this classification appears clear to me and extremely
significant for the theory: the principle even at its foundation (that
of the order) is not enough to judge admissibility of a harmony;
actually, the historical data play an essential role now, even in
the case of the last genus that Euler studies in detail, the genus
diatonico-chromaticus, which best approaches the music
contemporary with our mathematician.
At this point, it does not seem an exaggeration to say that we
"have just made the tour" of the Eulerian theory contained
in his work of 1739, a brief tour, certainly. Although
extremely abundant, it can be summarized with this: the concept
of order is the basis of the principle of taste.
This order is measured,
as in the old theory, by the coincidence of the blows. However,
Euler goes much further than a simple classification of the
consonances. He also intends to justify the whole of a series
of chords, by measuring its order. But all the definitions,
many calculations, the immense tables which our mathematician
deploys tirelessly should not make illusion. A certain number
of elements foreign to the principle of the order are introduced,
all the more "surreptitiously" that, corresponding most of the
time to principles admitted by the former theorists, they "pass"
without difficulty in front of the eyes of the reader.
Thus from there we go particularly to the essential role that
Euler makes the number 2 play, i.e. with the octave, to the limitation
to only factors 2, 3, 5, and from there to the construction of a finite
number of genera, to final restriction on only one genus. In any case,
these choices do not really explain the harmonic successions actually
used by the musicians. I will quote once again Yves Hellegouarch,
who notices pleasantly: "Is this to say also that a musical work
remains harmonious when it is played backwards?" Ah yes, indeed,
the "principle of the inversion of time", as in traditional
mechanics, applies perfectly to the theory of Tentamen. That is
to say, that it is far from being adequate!
Beyond Tentamen: evolution of Euler's ideas
regarding musical theory
When I first studied Euler's texts on musical theory,
it seemed obvious to me that he had never given up his theory, even
if he had wanted to supplement it or improve it on certain points
until the end of his life. Today I am not as sure of this.
Indeed, the reflections which he added to Tentamen,
in particular those
which relate to the Dominant 7th chord, tend on the
contrary to show that it was with the same bases of his first theory
that he attacked himself.21 In a paper of 1764, Conjecture about the reason of some dissonances
generally received in music, recognizing its practical
importance in the contemporary compositions, Euler examines the
Dominant 7th chord.22 To the notes G B D F
correspond the numbers 36 45 54 64 of exponent (their LCM)
8640 = 26*33*5.
Without the "dissonance" F, the exponent of the
chord (perfect major) would be equal to 60 "and therefore 144 times smaller than before. From where it
seems that the addition of its F spoils too the beautiful harmony
of this consonance so that one can grant a place in the music to him.
However, with the judgement of the ear, this dissonance is nothing
less than unpleasant and one makes use of it in the music with best
success; it even seems that the musical composition acquires of it a
certain force, without which it would be too plain. Here thus a great
paradox, where the theory seems to be in contradiction with the
practice, about which I will try to give an explanation. " Rejecting the explanation of d'Alembert as "too arbitrary and far away
from the true principles of harmony"23, Euler starts by recalling,
extremely skilfully and pertinently, that the ear tolerates slight
variations in the proportions of the consonances (he is not the first
to notice it, many others said it before him, but it does not matter
here):
"Each time it happens,
the perceived proportion is simpler
than the real one and the difference is so small that it escapes
perception." Consequently nothing prohibits to suppose that the ear "substitutes"
the number 63 for the number 64,
"so that all the numbers becoming divisible by 9,
the ratios of our four sounds are now expressed by these numbers
4, 5, 6, 7 whose perception undoubtedly is confounded."
One passes thus to an exponent of 420 instead of 8640 and (what Euler
does not specify) two degrees of softness (degree 15 instead of 17)
are gained. Euler finds even this astute idea: "Perhaps this is the foundation of the rule on the preparation and
the resolution of dissonances, to inform the listener that it
is the same sound, though it can serve as two different ones,
so that they he may not imagine that one has introduced a
completely foreign sound." [OO, 3a, I, 515] Here is an original
interpretation. One knows indeed that
it is usually supposed that preparation and resolution are only used to accustom the ear ,
beforehand
and "subsequently" as one might say, to the dissonant sound by
making it hear in consonances. And Euler concludes: "It is commonly claimed that one makes use in
music only of proportions made up of these three prime numbers
2, 3 and 5 and the great Leibniz 24 has already remarked
that in music we have still not learned to count beyond 5; which
is also incontestably true in the instruments tuned according
to principles of the harmony. But, if my conjecture has validity,
one can say that in the compositions one already counts up to 7
and that the ear thereby is already accustomed to it.
25 It is a new kind
of music, one that has begun to show use and which was unkown
to the ancients. In this kind the chord 4, 5, 6, 7 is the most complete
harmony, since it contains numbers 2, 3, 5 and 7; but it is more
complicated than the common triad which contains only
numbers 2, 3 and 5. If it is a perfection in composition,
perhaps one will try to carry the instruments to the same degree."
[OO, 3a, I, 515] The theoretical changes are much deeper than it first appears.
To start, the dogma of the restriction on only numbers
2, 3, 5 is now abandoned. Then, it would seem that the predilection
for the Dominant chord comes from the fact that it is a consonance and
not a dissonance. Actually, there is much more. Euler wrote in a
paper of 1764, The true character of modern music: "On this occasion it is significant to notice that the
word 'dissonance' is not very specific to express the idea that
one attaches to it; this idea is by no means opposed
to that which one attaches to the word 'consonance', as the
etymology seems to indicate, and therefore, since the consonances
are pleasant to the ear, it should not be thought that the
dissonances are unpleasant for it, or indeed revolting; with that
characterization
the dissonances undoubtedly would have to be entirely banished from
music. The dissonances thus differ from the consonances
in a way that would properly be described as
only because they are less simple or more
complicated, and it is also necessary that this greater complication
is just as pleasant to the ear as the simplicity of the
consonances."
[OO, 3a, I, 517] It is no longer exactly by the impression pleasant or unpleasant
only that consonances and dissonances are distinguished, but only by
their degree of complexity. Thus, at first, Euler had built
the mathematical concept of exponent
to account for the degree of agreement
of the sound combinations. But now, it is the
purely intellectual notion of complexity, measured by the
exponent, which takes precedence.
That same term of degree of softness
should now be banished, since a dissonance (the dominant 7th chord)
has become pleasant [i.e., more in agreement,
in terms of the numbers]. Even the word dissonance is
really no longer of use! 26 I believe that
in this matter Euler went very far, showed himself very "modern",
even if he did not explicitly repudiate the theory of his Tentamen.
T1 Furthermore, on this point as on the whole of his theory, one cannot
form an opinion of its relevance unless one calls
upon more recent explanations, in particular those which are founded
on the perception of the beats between the partial sounds [overtones]
of the
musical notes. These explanations, resulting from the theory of
Helmholtz, entirely confirm the principle of the absence of strict
segregation between consonances and dissonances. The key is in
knowing until which partial sound the ear extends its perception.
However that depends on two factors; on the one hand of the relative
physical intensity of the partial sounds (the experiment shows that,
statistically, 27 it decreases when the order of the partial sound
increases), on the other hand of physiological and mental capacity of
the listener to perceive the beats according to the height of the simple
sounds which produce them. Helmholtz himself will thus take up the idea of the consonance of
the Minor 7th as the "natural one" corresponding to ratio 7/4. 28
But the theory of the beats also immediately indicates the limits
of the theory of Euler. Because in addition to the fact that it
explains why the ear contents itself with approximate ratios,
it teaches us that the partials do not have the same importance and
that thus the "exponent" of a chord is the only number
that characterizes its degree of consonance or of complexity.
In particular, to add all the prime factors of the exponent to
arrive at what Euler calls a "complete chord" will not
improve the chord, as our mathematician imagines it, but will
lead rather to the cacophony of a plethora of partial sounds.
1. During the Middle Ages the
quadrivium indicated four mathematical
"arts": arithmetic, music, geometry and astronomy.
This quadrivium constituted the higher part of knowledge,
in opposition to trivium, the elementary part, which comprised
grammar, rhetoric and dialectic.
(back to text) 2. Cf. our work Leibniz et la théorie de la musique
[Leibniz and the theory of music], Klincksieck, coll. "Domaine musicologique", 1992, 158 p.
(back to text)
3. [OO, 3a, I, x-xiv]. This reference, like the following ones
of Euler, refers to
Opera Omnia,
series tertia, volumen primum, "Opera physica, miscellanea,
epistolae"; Lipsiae et Berolini, 1926.
(back to text) 4. the majority in French.
(back to text) 5. I will leave aside the pages devoted to
music theory
in the Letters to a German princess (hardly sixteen pages
out of nearly six hundred). They contain only one kind of ultra-simplified
extract of Tentamen.
(back to text) 6. "Music is an occult practice of arithmetic in which the spirit
is unaware that it counts." (letter to Chr. Goldbach of
17 April 1712). Cf. our book, already quoted,
Leibniz et la théorie de la musique, p. 151.
(back to text) 7. As above for [OO], we indicate by
[OCF] the references to
OEuvres Complètes en Français
de L. Euler [Complete Works in French of L Euler]
, edited by l'Association des capitaux intellectuels pour favoriser le
développement des Sciences physiques et mathématiques
[Association
of the intellectual capitals to support the development of the
Physical and mathematical sciences],
volume 5, which contains the Essai sur la
musique [Essay on music] and
three other papers on the same subject.
(back to text) 8. The prime numbers resulting from two different sounds must be
distinct. If not,
1:p:p
will be compared to 1:p and not to
1:p2, because the
two sounds p are
taken by the ear as a single sound.
(back to text) 9. Cf. for example our study "Cordes
vibrantes et consonances chez Beeckman, Mersenne et Galilée",
[vibrating Cords and consonances in Beeckman, Mersenne and Galileo]
Sciences et techniques en
perspective, 23, especially $3,
p. 81 - 88. This error is made
by Galileo without any ambiguity:
10. "magis est composita" = exceeds in complexity.
(back to text) 11. Or, if one prefers,
1/(p +
q - 1) will be able to constitute a
direct measure of this softness. It is this value which Galileo,
Mersenne and their contemporaries "should" have found.
(back to text) 12. Yves Hellegouarch, "L'"Essai d'une nouvelle
théorie de la musique" de Leonhard Euler"
[The 'Essay on a new theory of music' of Leonhard Euler],
publication of l'IREM de Caen, 19??, p. 53.
(back to text) 13. Ibid.
(back to text) 14. Euler presents these ratios in the opposite order of
that which is usual today (1:2 instead of 2:1, etc). One could
believe it an archaism, since the Greeks proceeded in this manner.
But Euler simply seeks to place the numbers in an
ascending order.
(back to text) 15. Cf. our article, already quoted,
"Cordes vibrantes et consonances chez Beeckman, Mersenne
et Galilée", p. 82.
(back to text) 16. Consequently: If the exponent is prime,
the complete chord will be composed of two notes: 1:a.
If the exponent is
If the exponent is If the exponent is am bn
cp (a, b, c prime), the complete chord
will be composed of (m+1)(n+1)(p+1) notes (this is the well-known rule
on the number of divisors of a entirety broken up into
products of prime-factors). Euler accompanies his definition by an immense table of the complete
chords of the first 12 degrees.
(back to text) 17. Cf. My former studies:
"Deux mathématiciens musiciens [Two mathematican-musicians]:
Euler et d'Alembert",
Physis, Rivista internazionale di storia
della scienza, vol. XXXII, Nuova Serie,
Fasc. 1, 1995, p. 1-35 (communication in 19th International
Congress of History of sciences, Saragossa, August 22-29, 1993);
"Sciences et musique: quelques grandes étapes en
théorie musicale [Sciences and music:
some great stages in musical theory]", to appear in
Littérature, Médecine,
Société, 13,
Université de Nantes.
(back to text) 18. See my comments in the texts quoted with the preceding note.
(back to text) 19. [OO, 3a, I, 290], $5 et 6.
(back to text) 20. The reference (without quotation specifics) is made in
the Tentamen page 332,
by way of citing authority. Later, in an article of
1764, Euler produces the quotation supplements [OO, 3a, I, 515],
but this time to protest against the principle which it claims to pose.
(back to text) 21. Also, in 1773, the Euler's ideas
as regards physics of the
musical sound became perfectly "modern":
22. Cf. [OO, 3a, I, 509 ff.]
Euler considers also his first inversion, of which it is useless
to speak here.
(back to text) 23. [OO, 3a, I, 510].
(back to text) 24. Euler produces here the quotation
which we reported above.
(back to text) 25. In another article, Euler says still better:
"we will be able to
say with the late Mr. Leibniz that music now learned how to count
up to seven." [OO, 3a, I, 525]
(back to text) 26. "Musicians agree that certain chords [such as the Dominant 7th]
could not be reconciled with the principles of harmony
and they try to support them by the name of dissonance that they
impose on them; but, if they understand by this term a chord where
the ear could not discover any ratio,
one should be able to to use it with as much success
as all other mixture of tone, absurd as it is;
the musicians are quite unwilling to admit that." [OO, 3a, I, 524]
(back to text) 27. I.e. on average according to the various instruments. Only
instruments of very distinctive timbre, like the clarinet,
present partial sounds
of intensity higher than the tenth.
(back to text) 28. Cf. Helmholtz,
Théorie physiologique de la musique, fondée
sur l'étude des sensations auditives,
[French] trans. G. Guéroult, Paris, 1874
[Die Lehre von den Tonempfindungen als physiologische
Grundlage für die Theorie der Musik,
1863], p. 249, 293 and 441 for the
references to the chord of ratio 7/4. However, Helmholtz
finally rejects this chord, like others founded on number 7,
for the reason that its inversion is "worse than itself".
And he concludes that "there is thus a true gap in the series
of the intervals arranged according to their harmony, and it is
this gap which forms the limit between the consonances and the
dissonances." (p. 293).
[English translation of Helmholtz by Alexander Ellis,
On the sensations of tone as a physiological basis
for the theory of music, 1875. Dover Reprint 1954
based on 2nd edition of Ellis. In this edition, Helmholtz
discusses the ratio 7/4 on p 195, and a chord relating to it
on p 344. I do not have the French edition available for
correlation.
feedback welcome]
(back to text)
T2.
I think it is important to translate
'causa horologium'
as 'clockwork' rather than just
'clock',
with an emphasis on the mechanism,
especially since in today's digital age
'clock' does not
give the same picture in the reader's mind as it would have
in Euler's day.
Euler's main reason for choosing a clock as his example
was that in his day it was the most complicated mechanism
encountered by most people in ordinary daily life.
His whole point here is to show how the order in this intricate
mechanism is what results in the perfection of its actions.
I think it is quite an ingenious example, not just because of
this obvious characteristic,
but also because of the other connotations associated with the clock:
the most elusive aspect of music
is its prolongation thru time, and time plays a major
role in the analysis of
frequency ratios, which will turn out
to be the ultimate subject of Euler's paper.
This reference to a clockwork strengthens my belief that
in his mind Euler had a latent model of these rational relationships
as a harmonic lattice diagram,
and thus makes his work a significant precursor to my own.
(back to text) -----------
T1.
This is exactly the position taken by such later theorists as
Schoenberg (Harmonielehre, 1911), Partch (Genesis of a Music,
1947, 1974) and myself (JustMusic, 1995).
See the dictionary entrance for
sonance.
(back to text) -----------
T4.
Prof. Bailhache prefers to translate
'suavitas' into English by the term
'softness',
so I have done so thruout.
However, the term can
be translated by a variety of English words which are not
necessarily synonymous. Some examples are
'sweetness',
'agreeableness', and
'agreement'.
Among the most successful, in my opinion, are
'attractivness',
and one which has a duality of meaning that may (or may not)
have been intended by Euler,
'attraction'.
(back to text) -----------
T5.
Although ordinarily the English translation for the French
genres
would be scales, the term
Euler Genus is already in common use
in English to refer to the particular types of scales under discussion.
(back to text) -----------
T6.
John Chalmers sent me the following:
To the best of my knowledge, Euler never constructed any genera using 7
and Fokker was the first to do so. Correct?
Who first discovered that sounds are not trains of pulses (as the
ancients thought), but sine waves? Who discovered overtones, Mersenne?
Professor Bailhache has contributed:
BIBLIOGRAPHY on Euler
University of St. Andrews (Scotland)
Euler, Leonhard.
English translation by Charles Samuel Smith,
Rameau, Jean-Philippe
[The text translated here]
I would like to thank the author for his generous assistance
"Thus the first and the most pleasant consonance will be that
of the octave, since to each percussion of the tympanum due to
the lowest string correspond two percussions caused by the
highest string: with the time of one vibration on two of the highest string
the effects will thus come to be combined, so that half of the
percussions on the whole will beat the ear together; on their
side two strings in unison, vibrating always together, give the
impression of only one string and for this reason do not produce
any consonance. The 5th also is pleasant, by the fact that
to two pulsations of the lowest string corresponds each time
three pulsations of the highest string: if thus one counts according
to the vibrations of the latter, a third of all the vibrations takes
place together, which means that two solitary vibrations come to be
intercalated between each couple of concordant vibrations;
in the 4th, these are three solitary vibrations which will come
to be intercalated."(
Discours et
démonstrations mathématiques concernant deux sciences
nouvelles [Discourse and mathematical demonstrations
concerning two new sciences], trans. Clavelin, A. Colin, Paris
1970, p. 85. Emphasis Bailhache's.)
(back to text)
etc...
"Incipiamus igitur ab unisono, qui constat perfecta aequalitate
duorum pluriumve sonorum musicorum; cum enim omnis sonitus motu
vibratorio sive tremore in aere excitato producatur, sive iste tremor
fuerit aequabilis sive inaequabilis, in musica alii sorti non
admittuntur, nisi ubi omnes vibrationes inter se sunt isochronae sive
aequalibus tempusculis absolvuntur." [OO, 3a, I, 569,
"De harmoniae veris principiis per speculum musicum
repraesentatis"].
Now, it is no longer a question of ictus or of pulsus and the translator
can correctly say:
"Let us start with the unison, which consists
in the perfect equality of two or several sounds of music. Any
sound owes its birth to a vibratory movement produced in the air,
the vibrations being equal or unequal between them; but in
music one admits as other sounds only those which are due to isochronous
vibrations, i.e. to vibrations carried out in equal times." [OCF, 5, 253]
Not Mersenne alone! It is a complicated story!
Biography of Euler webpage
1739 Tentamen novae theoriae musicae. Sint Petersburg.
In Opera Omnia series III volumen I,
Teubner, Leipzig, 1926.
PhD diss., Indiana University,
Bloomington, June 1960.
1760 Exposition de quelques nouvelles vues mathématiques
dans la théorie de la musique.
Amsterdam.
1764 "Conjecture de la raison de quelques dissonances
généralement reçues dans la musique",
Mémoires de l'Academie Royale de Berlin, vol. 20, 1764.
1764
"Tentamen de sono campanarum".
Novi commentarii Academiae Scientiarum
Imperialis Petropolitanae, tom. 10, 1764.
1774 De harmoniae veris principiis perspeculum musicum
repraesentatis.
Brun, Viggo
1753 Extrait d'une réponse de M. Rameau à M. Euler sur l'identité des octaves. Paris.
Vogel, Martin
1959 "Mehrdimensionale Algorithmen, welche die Eulersche Kettenbruchentwicklung der Zahle verallgemeinern", Festschrift Leonard Euler zum 250. Geburtstag, Akademie-Verlag, Berlin.
Busch, Hermann R
1960 "Die Musikschriften Leonhard Euler", in the preface to Euleri opera omnia, series III, vols. 11-12, Zürich.
Lindley, Mark
1970 Leonhard Eulers Beitrag zur Musiktheorie. Kölner Beiträge zur Musikforschung vol. 58, Gustav Bosse Verlag, Regensburg.
Klösch, Gerhard
1981 "Leonhard Euler als Musiktheoretiker", in Kongreßbericht Bayreuth 1981, Kassel.
Craats, Jan van de
1988 "Von der Eulerschen Konsonanztheorie zum relationalen Tonsystem. Ordnung in der Unendlichkeit der Töne", Mikrotöne II, Edition Helbling, Innsbruck, pp. 51-64.
Muzzulini, Daniel
1989 De Fis van Euler. Aramith Uitgevers, Bloemendaal, 144 pages.
Bailhache, Patrice
1994 "Leonhard Eulers Konsonanztheorie", Musiktheorie vol. 9 no. 2, 1994. Laaber Verlag, Laaber, Germany.
Hellegouarch, Yves
1997 "La Musique traduite en Mathématiques: Leonhard Euler", Communication au colloque du Centre François Viète, "Problèmes de traduction au XVIIIe siècle", Nantes, 17 janvier.
no date "L'Essai d'une nouvelle théorie de la musique de Leonhard Euler", publication de l'IREM de Caen, p. 53.
in helping me with the Latin translations.
