OFFICERS OF THE SociETY
78.9)
Present
£4. KENNEY
Count:
CL AUSTIN (he
A.B WALLACEHADRILL (Meigs Secretary),
D.N.SEDLEY (Programme Secretary),
RG.G. COLEMAN, PE EASTERLING, AM SNODGRASS
Euitors of the Proceedings: E.3, KENNEY, 1. DIGGLE
The objects of the Society are he furtherance of clases studies in genera, and
ture anu evlzation of Greeee
mmofamember resent
on of aresident research stents 2.50, The subse
| member resident ehewhere 1 £2.50 annually. Members receive notices ofall meetings ofthe Soaety
aod he Procredings
‘Any library may subscribe tothe
snes £5.00 pee anim,
ety and receive comtes ofthe Proctedings. The subscription for
The Sox
Proceedings ofthe Cambridge Pilotagical Socety con octet and other
ls by members, appears annually, Contributions intended fr the Proceedings should be addressed
to Professor E. 1. Kenney, Peterhouse, Cambridge CB2 IRD.
‘Supplements to the Proceedings, consisting of monograph, are a occas
1 sponsible fortwo series of oublcations,
ning papers read to
sees appearing less
ioworxsofintermediaesize,
i proposals for monogrannsto be
Campridge CR?
eavons for membership, and all other coresondince rtaing to te Socety sould be
addressed to Dr, Austin. :
(OI KAAOYMENOI * APMONIKOE:
THE PREDECESSORS OF ARISTOXENUS
A cursory glance at the reports ofthe later students of harmonic theory isenough
to give a clearif perhaps artificially systematic picture of the character and relations,
of the major conflicting schoo's of thought in the first century or so A.D. In the
centre of the field are the supposed followers of Aristoxenus. lined up against the
forces of the so-called Pythagoreans. Each side is linked with a more or less tunatie
fringe; to the right of the Pythagoreans those mathematical extremists who find no
place in harmonic studies for alo®ors at all, and to the left of the more empirical
Aristoxeneans a collection of persons known as opyaviKoi, whose work, Whatever
1 was, 18 based wholly in perception and in familianty with the properties of
musical instruments, and who find no place for theory or for the pursuit of the
itiat of harmonic truths.
‘Though Aristoxenus would have been horrified at some of the views espoused by
és reputed followers, and though what came together under the title of
“Pythagorean’ doctrine had many diverse sources, there 18 no doubt that at least a
solid amount of what is attributed to these schools by such writers as Ptolemy and
Porphyry quite genuinely goes back to the fourth century, to Anistoxenus on the
one hand, and perhaps 0 Archytas and his followers on the other. It may be this
fact that has tempted commentators to read back the line-up of forces which
Ptolemy found in his day into the late fourth century itself The picture presented is
of Anistoxenus, himself a self-conscious mnovator, im arms primarily against
entrenched battalions of Pythagoreans, and occasionally hitting out m an
sncidental way at a relatively significant super-empirieist fringe to. whose
members he gives the name GpyoviKoi.
1 shall argue that this preture is quite wrong. Let us begin with the Pythagoreans,
supposedly Anstoxenus’ prineipal opponents. It might perhaps be thought strange,
bbut is 1s not too surprising, that he never once mentions them by that name. There
‘might be all sorts of reasons for that. Its possibly a shade more surpstsing that he
never mentions by name any individual whom we could now recognise as a
Pythagorean ~ Archytas, for example. Thirdly, we are entitled to wonder why, in
stating those of his doctrines which plainly do run counter to Pythagorean
positions, Aristoxenus fails to mention the fact that this conflict exists, and makes
no attempt to explain it. As to explicit criticisms, out of the 35 or 40 passages in the
Elements sn wich Aristoxenus dizecty attacks the vicars of contemporaries and
predecessors, only one (32.20, is gertainly dirécied against {he Pythagorean, and
on my understanding of it only One other every might be. Pinall), i Senis'édd. to
894382 A. BARKER
say the least, that m a number of passages Aristoxenus makes quite general claims
about his predecessors’ failure even to consider certain topics, when in several
significant cases we have good evidence that Pythagorean theonsts had already
looked at them in some detail. These points must at tcast suggest that there is
something amiss with the commonly imagined scenario of a fight to the death
between Aristoxenus and the Pythagoreans. On the contrary, he seems largely to
hhave ignored them.
‘On the other side, his complaints against the apyovikoi, the more obviously
are very frequent. But I shall argue that this is not to say tha
regards them as representing @ school of thought opposed to his own, or as some
specially characterised group of eccentrics. He recognises, |
group of dpriovixoi out on the far left, from which he wishes to
rather, he gives the label apovixds to anyone whom he recognises as working in
the same field as nimself, and would, I think, have cheerfully accepted thi
appellation for himself. He eticases those who fall under this neading not as
‘methodological heretics or eccentrics, but as having failed to do their work properly
~ work which he will now do much better. Of course, he may be quite wrong about
this, rather as Aristotle has been criticised for seemg the Presocraties as trying and
failing to be Anstotie: but for the present we are trying to see the relation as he
imself perceived
| shall suggest that the reason why he has solittle to say about the Pythagoreans is
imply that he does not recognise tiem as attempting the samme task, If their work
with which he was certainty familiar 1s significant at all, it1s so ina different field,
and is not relevant to him. Thus his view of his own intellectual environment is
entirely different from that of the historians of the Aristoxenean and Pythagorean
schools of later antiquity, despite the fact that the basis of these latter accounts 1s
Ay to be found in the fourth century. There are even those in the fourth
whose understanding of the situation is not dissimilar: Theophrastus
4s conceivably a case in point.* But Theophrastus is not Aristoxenus.
‘The significance of these pretiminary remarks will, hope, come outin the sequel,
since my object in this paper isto say something about the background of previous
research against which Aristoxenus is working. In order to do this T must pursue
this same theme a litte further, with the aim of providing some evidence for the
‘latms I have made so far: having done that I shall devote the remainder of the
paper to the attempt to see what, if anything, can be said about the work of those
whom Anstoxenus does regard as his rivals and predecessors. It will not be much,
since they are scarcely known except from Aristoxenus himself, and his comments
are almost invariably negative; but the attempt seems worth making, since sofar as
can discover remarkably litle has previously been written about those early
musica! theonsis who cannot comfortably be sheltered under the “Pythagorean
umbrella
I shall begin by summarising the evidence on which I base my contention that
‘THE PREDECESSORS OF ARISTOXENUS 3
Aristoxenus found tne work of the Pythagoreans im this field no more than
irrelevant.
‘The central part of the Pythagorean programme seems to have been the analysts
of the primary intervals of the scale m terms of mathematical ratios.” Given this
preoccupation it was natural for them to ask what these ratios are ratios of, and this
question leads directly to the analysis of sound in terms of movement, and of
variation in piten in terms of variation m the ‘number’ of that movement, 1.6, it
speed.‘ There were various theories as to the character of the variable velocity or
whieh piten depended. Two were of particular importance, the view adopted by
Plato, probably by Archytas, and certainly by a number of writersin the Peripatetic
tradition, that it was the velocity at which a sound is propagated; and the thesis
ibed to that rather shadowy figure known as Heraclides Ponticus, that it wa
the frequency of the vibration of the moving ai.* But Aristoxenus dismisses all such
speculations out of hand. pr} taparrét@cay 8° Ais ai tv etc KwHOEIs aysvtE\
‘ois @BSyyoUug S6Eat Kai KaBSKOU THY pov RivtIOW Elva pacKsvTOV, Hesays a
12.46. His point, emphasised several times on Meiboms pages 9-11. 1s that suct
considerations have no bearing on the proper subject of harmonies, which is th
character of musical sound as heard, The physical conditions ofits production an
fof no harmonic imterest.§
32.20 ff. is the one passage which deals plainly and directly with Pythagoreas
theory, Here Aristoxenus describes one class of his predecessors a
aiorpiodoyotvees Kat THY uév Glonow eKkhivovtes ac odeay OOK AxorPT
vonrac Bt xaraoKeiZovtes alttag Kat paoKOVTES ROyoUC Te twvas aplBUBY elva
‘eat tdign pds GAAMAA EV Ole TS Te OED Kai tO lupo yiyvera, All the marks o
Pythagorean theory are here ~ the preference given to reason over perception, th
analysis of piteh relations by means of numerical ratios, and the account of
ai tO apd in terms of physical velocities. The authors of such speculations ar
ily stigmatised as aXAotpwortroug ASyous héyOvtES Ka EvavTL@TATONS Tol
‘patvouévoic; and apart from the remark in the following paragraph that harmon:
is not geometry, and must make use of precise and accurately trasned alan, th
school of thought is never referred to again It is evident that Aristoxenus sees n
need to purse them on points of detail ~ the1r entire methodology rules them ou
from scrious consideration in this context.
Thirdly, as an example of te way A\ 1e neglects Pythagorea
researches, consider his repeated assertion (2. -13, and elsewhere) thi
none of oi Eunpoo€ey studied any of the musical genera other than th
enharmonic.” The claim is puzzling in a number of ways, and Ishall return to st, bu
for tne moment the poist 1s sumplc and ciear. Unless our information in Ptolem
and Porphyry is altogether wrong - and though there are serious difficulues ¥ don
think that i 1s ~ all three of the genera had been studied and given complete,
controversial, mathematical analyses by Archytas.* Even if we reject this eviden«
the difficuity remains, since there 18 no doubt at all nat the scheme of ratios used4 A BARKER
Plato in the Timaeus yields a scale which is not enharmonic, but diatonic.’ Thu
fact that at teast this other form of the scale had been given a quantitative analys
can scarcely have been unknown to Aristoxenus: Dut the analysis is Pythagorean,
set in terms of K6yo1 dpiOudv, and is therefore irrelevant, Hence, though the
‘question of the origin of a systematic distinction between the genera rematns vexed
and problematic ~ I shall say more about it below ~ these passages provide further
evidence of a high order that Anstoxenus did not consider Pythagorean analyses as
falling within
he
harmonics many items of substantive
le with the Pythagorean approach. Two
‘obvious ones arc his assertion and proof that tne interval ofa fourth is
and his assumption that no problem is presented by thi
‘Yet he nowhere mentions that such claims are
Further, there was in the Pythagorean system one
glaring and obvious flaw, that while the interval of an octave plus a fourths clearly
heard as a concord, 1n the system of Pythagorean ratios it has to be classified as &
discord.!? The interval in question appears in Aristoxenus’ list of concords at 45.8-
22, without any relevant comment; and one could hardly suppose that anyone wit
so waspish a tongue as he could have resi ning, even
in passing, so unambiguous an embarassment to his opponents, had he considered.
rem to be opponents at all
but that the label 1s used by
those contemporaries and predecessors whom he takes
in his own field. The case 1s not easy to make out
nm to have presented proof: but the
considerations which I stall mention seem to me to carry some Weight
First let me recapitulate and elaborate briefly the position whieh T want to
contest. Pozphyry, m his commentary on Ptolemy's Harmonics,
‘mus, and Ptolemais of Cyrene, to
en Anstoxeneans and Pythagoreans
understood." Broadly speaking there are two major di
uyse pitch-relations, concords, and intervals generally in terms of
}¢ Anistoxeneans treat pitches as points on an audi
rated by certain distances. This a fascinat
{shall say
tas not my main focus here. Secondly, and for present purposes more crucially,
they differ m their understanding of the roles of 2.0705 and dicot. Both bee
from dioOnorc: both recognise subsequently a role for Aoyos, but for the
Pytnagoreans, once the starting point has been adopted, Aoyo% sth eri
its results confi io®norc, this merely proves the latier’s ingecuracy and
Anistoxenus to designate.
to be working
‘THE PREDECESSORS OF ARISTOXENUS
unreliability. For the Aristoxeneans, apjovuna is concerned wholly with that whic
can be heard, and it studies the laws of music as heard, All judgments mu
therefore eventually be brought back to the test of dlo®norc, andif reason purpor
gone astray,
Aside from these two main-stream approaches, Ptolemass also recognises (
Prec ana who mal :
Shame, put in'that they Ehe no demonstrations or arguments fo support
Seromons, Skin praieforms he soe bass for he celosons hey Teach 3
he pte them the hame opyaveo! or @ovOKNO
Now ts erect ae tnara amber of thee
eto the effec that tey Ae overempin
the stuctore of etl instruments, at hey present
évapndvioc.) It would,
term to be recognised here as the
for ol Euxpoodey, and the pun is made possible by a combination of the comm
ise of épovi6s to refer to a student of harmonic theory with the special sense
pnovia Which designates the enharmonic genus.
the paragraph which follows Anistoxenus attributes several shortcomings
these predecessors, One 1s therr failure to consider any but the enharmonic gen
e attributes ‘whose use t
fal nog 11-17). They ct6 A. DARKER,
he arrangement of the octave, and not of the other yeyéBy and orpiwara
@.17-25), No one, he says, has dealt with these other matters, and what they have
discussed they have treated imperfectly ~ as ne himself has shown previously dee
emzaxonotyiev ras vv apuovixdy 8Eas (2,25-30)
Iisa great pity that tne work he here refers tos lost. But at leas
that he is making no di
generally,
Passages from elsewhere nthe work confirm this first impression. The failure of
his predecessors to consider any but the enharmonic genus 1 referred to again at
35.1-13: here the word &pyovtrds does not occur: we are simply told that no one
hhas worked on the other genera. We seem to meet the ‘diagrams’
joned in section 2 mn at least three other passages, all associated with the term
Karanixvwots, The reference s apparently tothe representation of scales ona map
or diagram consisting of a line divided at sntervals of an entiarmonie diesis, ot
quarter-tone, and I shall go into its significance in some detail later. One passage
(53.3-11) mentions in connection with the concept of continuity, 8 2£¥¢, and uses
the phrase of cls tv xaturtxvecty Bhénoveas: this makes
Atistoxenus’ predecessors employed xt. In an eat
concerning continu
tats rv Stay py
seems clear
tion as yet between ot Appovixoi and ol EuRpotey
(27.34-28.3) also
We are told Enrntéov 6 x6 cuvezts ody dc ot appovrxor Ey
katanvevoceow axodtBévat neipiv plain
the term karabxvoots and the use of the
as of tipyovexol. In yet another passage
speaking of modulation and the proper relation of ovarywava to toxot,7
he says that weshould dicuss them on npg thy katandxvoo BASroveas KaQanap
of appovixot: and again (7.30-2) that some of ot apyovixo! have touched on the
matter xavé-tiyny, not with this issue in mind but xararvrvGout Bovowevors 19
‘Stéypaywa. Again we seem to have a clear association of of dppovixoi with
Karantikvoots, and this might easily be taken toindicate a reference to some special
school: yet the very same sentence continues KaQ6Aov 8 odSevi oxeBOv ty Tote
Eunpoatey pavepov yeyévnra: r008 Ayty. which
the reference of the term apuovexot isin fact quite general, certainly sug,
are al any rate not justified in following Macran’s preyudi
rendering oi8evi as ‘no single writer belonging to this sch
If the associations of
own, my case would be at best flimsy. The evidence, ifanything, would go the other
way; ana these passages more than balance out the argument which I based on his
references to his predecessors’ obsession
story, however, when We come to hi them for failing to deal properly
with oyripera ang yeyé6n other than the octave. We have seen that these failures
were described am his work concerning tac tiv opovixidy 86Eac. On Meibom's
pages 5-6 this criticism is elaborated, with reference ex
Eratocles, who is not known from any other source. Is
return to him later, but
“THE PREDECESSORS OF ARISTOXENUS,
conclusions.
generally with a
there are the
‘whagoreans, who are irrelevant, and there arc
(ag Ka dnoBeiEews, and who do not even make
adequate enumeration of abta ra gatvoueva (32.20-31). It ean hardly be do
that Eratocles is here being taken as represent
at least the expressions oi Euxpoo¥ey and ot GpHoviKot
the same reference.
fe to mention briefly
‘0 other passages. At 37,8-35 he scems to
Tsnould vs
proof of his assertion that no one (00
the fact that the treatment of this subject
mentions three conflicting groups of claims made about tovor by diff
rely random; an
cho
ol approver
Gppovixoi, This at least shows that the apHovecot are not a mono!
seems also to indicate that 2 bundle of information abou
tantamo 1n about all his predecessors.
wath the thesis that the aim of his science
T shall discuss further below, is si
attributed after the long passage 1s com}
aapuovixot are named, ina rather ambiguous expression embedded in the cia
ev obv Sv ayvoLay THY OASAMYLY CATHY EoZAIKAOLY Ot KaotLEVO! GPHOVE
(40.25-6). Now this expression might easily be read as “the so-called Gpjo\
and taken to name this supposed special school, But if that 1s right.
‘Aristoxenus not make the attribution at the beginning of the argument? Hi
done nothing to show that the people he 1s talking about betong to some sp
schoo! ~ they are so far merely tives. The point, I think, 1s contemptuous
‘Anstoxenus is here in a particularly scathing mood. “These so-called 6p}.ov
‘means ‘these people who dignify inemselves with the honourable title of studer
deserve it’.
point fe this survey of the evidence in the te
he notion that ol apyove
id expect to find spectal criticisms also attached to other sp
ing of the kind 1s to be found, apart from one off-hand
meomprehensible reference to the much older musteian Lasos of Hermione
4), At least one Would expect to meet eniticism of kinds of approach, met!
doctrine which are not those associated explicitly with ot apovixot; but
so either. As far as \age, though there 1s one focus of ensticism, the v
katarikveots, which Is always linked with the term appovixos, th
no significant focus of8 A. BARKER
Pythagorcans apart ~ there 1s no schoo! of musical thought known to Aristoxenus
for whose members the label apuovixoi would have seemed inappropriate.
1 mught add two small pieces of indirect evidence. The word apyovixoi, used in
reference to students of apuovext,, occurs in Theophrastus Fr. 89, where it
Gistinguishes those who make important use of aigdnaxg from the naOnuacuxot,
presumably Pythagorean, who do not; and though Theophrastus concedes that
some people
yurposes the important poi
group. There are the apyovixoi and there are the
it might be extracted from the famous passage at Republic $31a-c,
‘Though Plato does not use the word 4pyovixdc, there seems a pretty direct relation
between his ypnotoi, desperately listening for minute differences in pitch, and
Aristoxcnus’ apuovixol: there is even the verbal link between Aristoxenus’
katanvevoorg and Plato's descriptive phrase xuxvawat’ eva 6vopdtovtec. I stall
return to this passage later: for the present my point is that once again - Plato’s own
doctrine apart ~ only the two ‘schools’ are mentioned, these ‘empiricists’ and the
Pythagoreans.
I shall now move on to the remaining part of my programme, to see what can be
Said about the methods and aims of these predecessors of Aristoxenus. T shi
mention a number of issues, but [shall {
4s What it is ~ known as KataxOxvoat<,
because Aristoxenus obviously considers it important, and partly because it gives
us at least a toe-hold in something post ‘shard to build up much of a picture
con the basis of a list of things which the apuovixoi did nor do,
At its most straightforward, what xataxdxvweic seems to mvolve 1
construction of a diagram in the form of a line, with a series of marks at equal
the boundaries of the smallest m
\gram, in fact, represents a series of quarter-tones. On to this line
n be mapped in a way which revealed the sizes of the component
‘my picture partly
intervals: the
scales could th
sntervals.*®
If Aristoxenus is t 'd, the purpose of this construction is to be
connected with the fol ters,
(®) The explanation of the natures of musical ofpata and weyéOn: that is, the
analysis of the mtcrval-structures of the various scales, tne smaller components of
which they are constructed, and the complex arrangements which can be formed by
their conju
jons between rOvor (‘modes’); HiEIe oF ‘mogulation’; and the
appropriateness of particular kinds of scale to particular trot, or rang
‘Anistoxenus scems to say that the Katanbxwaatg diagrams could have been
deal with these matters, but (hat the apovtKol ignored them or treated them only
kota tyqV, in passing: this suggests that for the ApHovexot their real value Lay
elsewhere (7.4-34),
‘THE PREDECESSORS OF ARISTOXENUS 9
Gi) The nature of +8 88}, musical continuity. Of this Aristoxenus complains
that the Gpovixoi thought of their quarter-tone series as representing 0 tEiis;
‘whereas, he argues, no more than two quarter-tones can ever occur successively ima
Proper musical sequence, and nence the phinciples determining which mn
roperiy follow another must lie elsewhere (27.34-28.17,and'53.3-15)>"I shall have
alittle more to say about this
‘We may note further, in a pr
to have had rather general assoc
structure of the enharmonic octave, and incompleteness and inaccuracy generally
(2.7.28: ef. 6.1319, 32.29-3 35-38.5), (b) an excessive empiricism, which
excluded the formulation of aitiat, explanations for the conclusions reached (6.13-
14, 32.29.30), and (¢) an emphasis on correct quantitative notation as an important
aim of the science (39-41)
‘One might well wonder why Aristoxenus should have regarded the procedure 1
fhave described as more than harmless tive. There are, T think, quit
1m the present context ist
interest or importance to its own
feems strange to suppose that the construction
ave been a central part of anyone’s
programme for harmonic science. But I think that it can be shown that such
appearances are deceptive, and that the development of the xutaxixvoots
diagrams must have been both significant and difficu!
Consider the circumstances, Before the fourth-century theorists began to work
‘on them, the scale-forms employed by Greek musicians 1n actual muste seem to
have becn multifarious and chaotic. As far as scholars have been able to judge,
different kinds of music may well have employed differently organised series of
notes. but no one Nad sought or discovered any determinate principles by which
they could be related and compared to one another. The characters of the different
kinds may have been more or less recognisable, but there had been no systematic
{tack on the question what it was that generated these different characters. Itis not,
even clear that these differences were primari as differences mn mnterval-
structure rather than e.g, relative height and depth of pit as during the fourth
century that these distinctions were first. pinned down to differences in
mnterval-structure, and these differences compared, organised, and eventuall
fossilised, and itis in the context of the project of developing these comparisons that
Karambevioats is, I believe, to be understood,””
‘The investigators who began this work were hampered by the lack of any
appropriate tools. There was of course no key-board, on which the relative sizes of
there any no! system
vals as the modern stave does. Such
nas is known identifies gSyy0t not by pitch, but by Sévayic, thats, by the10 A. DARKER,
functional role performed by that note in the scale.” Aristoxenus himself argues for
the view that the notes in the scale are to be considered! as Suvéuetc, restricted more
oF less narrowly to a particular tdoxs or pitch by the nature of the relevant péhos,
land against the tiesis which he ascribes to his predecessors that @66y/ot are to be
simply identified as téec (see two extended passages, 39-41 and 46-50)
to infer from what he say the view
‘given note in the scal
natural note ofthat scale. The idea t
them, between two notes of an accepted scal
depends on the possibility of thinking of the a
independently of given s inking m this Way 15,
plainly a precondition of the quantitative comparison of the interval structures of
the different accepted scales. | suggest, then, that this idea 1s not one which arises
directly out of common musical practice, but one developed by Aristoxenus'
jecessors as a tool of analysis. It 18 previous theorists and not practical
ans Who are under attack in the passages cited above.
itatively compared,
next note up a given scale isthe same distance
he quest
tackled is how the relative sizes of the steps of the scale can prop
Its here that the relevance of the xatamtxveocts di
abstract schema on to which any scale, indeed any coll
be plotted, and by means of which the sizes of their intervais can be compared: and
it provides for the first ume the possibility of a quantitative notation, allowing us to
write down unambiguously any hoard sequence of pitches. Aristoxenus, as we have
seen, treats the view Uhat an accurate quantitative notation is the crucial aim of
har contempt; but if my argument 1s near the truth his disdam 1s
scarcely justifie
‘We can thus
least guess at one major objective of those who employed the
‘must have been extremely