You are on page 1of 11

Theinred of Dover on Consonance:

A Chapter in the History of Harmony

A Summary of Harmonic Theory to the Twelfth Century


Harmony characterizes Western music; in no other music
has that element developed to such an extent, at the expense of
others. The road to harmonic thinking, however, was long and
tortuous, wending its way from Greek and Hebrew beginnings
through the still little-understood development of that greatest
body of Western monophony, Gregorian Chant, to organum,
polyphony, figured bass, and beyond. The actual origins of harmony are now lost in the mists of time; mention of experiments
with organum dates from the tenth century. Beginning with a
small number of permissible intervals, the harmonic vocabulary gradually expanded until, by the thirteenth century, all of
the diatonic intervals had been fitted into a theoretical system
that assigned them relative consonance-values. The development of polyphony posed many problems for both practical and
speculative musicians: intonation, the definition of consonance, the classification of intervals. It is the purpose of this article to examine developments in harmonic thinking which took
place roughly contemporaneously with the Notre Dame
School, with particular emphasis on the contributions of
Theinred of Dover.
The Middle Ages inherited only a part of the legacy of antiquity. Medieval music theorists had to depend largely on the
work of Boethius, in which the contributions of some ancient

authors were partially preserved. In brief, classical music theory recognized six consonances: the diatessaron (perfect
fourth), diapente (perfect fifth), diapason (octave), diapason
cum diatessaron (octave plus perfect fourth, i.e., perfect eleventh), diapason cum diapente (perfect twelfth), and bis diapason (double octave); Pythagorean intonation was assumed.
Greek theory had originally conceived of consonance in melodic terms, but by late antiquity both melodic and harmonic
consonances were distinguished. 1 This is the approach taken by
Boethius himself and is the point from which medieval theorists
took their departure . 2

1 "The Classical Greek concept of harmony considered consonances in melodic terms; if the mese (a) was sounded following the hypate (e), the result
would have been considered a consonance, that is, a melodic consonance"
(Calvin M. Bower, Boethius' "The Principles of Music," An Introduction,
Translation, and Commentary [Ann Arbor: University Microfilms, 1967], p.
440). In post-classical times this notion was replaced by that of simultaneous
consonances. The following passage from Boethius' De institutione musica
IV .1 is instructive: "Consonant sounds are those that sound sweet and intermingled when struck at the same time, while dissonant sounds are those which
do not sound sweet and intermingled when struck at the same time" (Bower, pp.
212-213). Bower states in notes to this passage that the phrases italicized above
were added to an earlier definition by Boethius or an intermediate source. For
a fuller discussion of this question see Bower, pp. 440-43.
2 lbid., p. 443.

Downloaded from http://mts.oxfordjournals.org/ by guest on January 31, 2014

John L. Snyder

Theinred of Dover on Consonance

Then [the composer] should place the first note of the organal voice,
i.e., the inception, with the cantus, either at the octave below or
above, or at the unison, or at the fifth or fourth, or sometimes at the
third or sixth. The organum may never be at the second or seventh
from the cantus, because it sounds badly. The notes in the middle,
between the beginning and the predetermined last two, he should
place either at the fifth or fourth or third or sixth, but more frequently
at the fourth or fifth, because it sounds better.
Postea primam vocem organi id est inceptionem ponat cum cantu vel
inferius in diapason vel superius, vel in eadem vel in quinta vel in
quarta, aliquanto et in tertia vel sexta. In secunda autem vel septima
voce a cantu numquam erit organum, quia male sonat. Medias autem
voces, inter primam et ultimas duas preelectas, ponat in quinta vel
3 Hucbald and the author of the Enchiriadis treatises accepted Boethius'
views on the subject; Guido d'Arezzo and John of Afflighem were not specifically concerned with consonance but seem also to have accepted traditional
thought on the matter.
4 MaiMnder Traktat, ed. Hans Heinrich Eggebrecht in Ad Organum Faciendum (Mainz: B. Schotts S6hne, 1970), pp. 111-15. See also the edition with
English translation by Jay A. Huff in Ad organum faciendum & Item de organo
(Institute of Medieval Music, 1969).
5 Traktat von Montpellier, ed. Eggebrecht in Ad Organum Faciendum, pp.
187-88.
6 "Quasi dulcis fistula." Both Eggebrecht and Huff translate fistula as flute,
which seems unnecessarily narrow.

quarta vel tertia vel sixta, sed frequentius in quarta vel in quinta, quia
pulcrius sonant. 7

These excerpts show a definite, if not very systematic, preference for thirds and sixths over seconds and sevenths; however,
the authors of these early organum treatises did not attempt to
fit the thirds and sixths into a comprehensive consonancedissonance scheme. 8
Theinred of Dover
Theinred's treatise, De legitimis ordinibus pentachordorum
et tetrachordorum, survives in only one manuscript, MS Oxford, Bodley 842, folios 1-44v 9 Theinred is not otherwise defini
tely known. 1 He was long thought to have written in the late
fourteenth century, but modern investigations indicate that he
lived in the twelfth century. 11 Although the main thrust of
Theinred's treatise deals with the problem of non-diatonic tones
in plainsong, the work also presents some interesting views on
consonance. Indeed, the only passages from the treatise to have
;

Traktat von Montpellier, 4-6 (p. 187).


Gut ("La Notion du Consonance chez la Theoriciens du Moyen
Age," Acta Musicologica 48 [1976]: 20-44) interprets these remarks as indicating that the authors regarded thirds (and sixths) as "intermediate consonances"; as we shall see, however, actual recognition of such a category is a
much later development.
9 No definitive edition of the text has yet appeared; the only available source
is the preliminary reading included in John L. Snyder, "The De legitimis ordinibus pentachordum et tetrachordum of Theinred of Dover" (Ph.D. diss., Indiana University, 1982).
wOn his possible identification with a Tenredus mentioned by John of Salisbury in his Metalogicon, see Snyder, pp. 16-18. This identification was first proposed by Clement C. J. Webb in an article, "Tenred of Dover," in the English
Historical Review 30 (October, 1915): 658-60.
"Snyder, chaps. 1 and 2. In a recent article ("Transposition and 'Key' Signatures in Late Medieval Music," Musica Disciplina 33 [1979]: 27-41) Gilbert
Reaney suggests a later thirteenth-century date for Theinred. The problem
cannot be dealt with in its entirety here, but some aspects of the issue will be
discussed later in this article.
7

8 Serge

Downloaded from http://mts.oxfordjournals.org/ by guest on January 31, 2014

Early medieval theorists did not venture beyond this position. 3 With the rise of polyphony, however, a more comprehensive theory was neededone that would define the relative
consonance of all the diatonic intervals. The beginnings of this
process may be seen in two of the earliest organum treatises,
the Treatise of Milan (late eleventh century) 4 and the Montpellier Treatise (early twelfth century). 5 In the former, a major
third (C and E) before a cadence by contrary motion (on D) is
said to sound "like a sweet pipe. " 6 In the latter, thirds and
sixths are allowed in organum, except that they may be used
only sparingly at beginnings and never at endings:

111

112

Music Theory Spectrum

Consonances are those sounds which, if made simultaneously, are


concordant by musical reasoning. Dissonances are those sounds
which, if made simultaneously, are discordant by musical reasoning.
Whence it happens that some diversisons are consonances and equisons, but others consonances and inequisons; moreover, that some
are dissonances and equisons, but others dissonances and inequisons.
Consoni sunt qui si simul fiunt musicae ratione concordant. Dissoni sunt qui si simul fiunt musicae ratione discordant. Unde contingit
diversisionorum, alios quidem esse consonos et aequisonos, alios
vero consonos et iniquisonos, alios autem dissonos et aequisonos,
alios vero dissonos et iniquisonos. 13
12 See, for example, the New Oxford History of Music (London: Oxford
University Press, 1954), vol. 2, p. 339; Gut, p. 29; Fritz Reckow, Der Musiktraktat des Anonymus 4 (Wiesbaden: Musikwissenschaftliche Verlagsanstalt, 1970), p. 78.
"Liber I, capitulum i sententiae 8-10. Hereinafter, citations of Theinred's
treatise will be made in the form <I,i,8-10>. (When a capitulum contains
more than one paragraph, the paragraphs are indicated with lowercase letters.)

The scribe has provided, in the margin, a diagram of this


scheme:
Soni
unitas

diversisoni

aequisoni

inaequisoni

consoni

dissoni

This diagram is not entirely accurate, however; a better representation of Theinred's classification of sounds follows:
Soni
unisoni

diversisoni
aequisoni

inaequisoni

consoni dissoni

consoni dissoni

The category "equison and dissonance" is possible because


Theinred considers multiple octaves beyond the double octave
to be dissonancesor at least non-consonances (see below).
This, with the fact that Theinred separates the unison from the
octave(s) in his scheme, indicates that the notion of octave
equivalence was not well developed in his mind.14

"See the discussion of the eleventh, below.

Downloaded from http://mts.oxfordjournals.org/ by guest on January 31, 2014

heretofore appeared in print are the incipit and a few sentences


dealing with the consonance of thirds. 12 Theinred's treatment of
the issue is, as will be made clear presently, much more
complexand much less settledthan has been generally
known.
The treatise is divided into three books, with a Prooemium.
The discussions concerning consonance occur primarily, but
not exclusively, in Book II. In the first chapter of Book I ("On
the proportions of musical sounds"), Theinred sets forth a dichotomous classificatory scheme for musical sounds. All sounds
are either unisons or diversisons (non-unisons); the latter are in
turn divided into equisons (octaves and multiple octaves) and
inequisons; each of these latter categories is divided into consonances and dissonances. These final categories are defined subjectively:

Theinred of Dover on Consonance

' 5 Ptolemy, Harmonics 1.4.9-10 and 1.7.15-16, quoted in Boethius, De institutione musica, ed. Godofred Friedein (Leipzig: Teubner, 1867), V,v and V,
xixii.
16 [f.31". . .non a numero sonorum, sed a sonitu musicae melodiae tonum
dixerunt." For a discussion of various etymologies for the term tonus, see the
remarks of John of Afflighem in the translation by Warren Babb (Hucbald,
Guido and John on Music, ed. Claude Palisca [New Haven: Yale University
Press, 1978]), p. 116 and note. It seems unlikely that Theinred knew John of
Afflighem's treatise.

Theinred then presents two sets of definitions for the series of


terms monochordum through octochordum. In the first set
these terms are treated as generic interval designations: major/
minor second through perfect/augmented/diminished octave.
Theinred's inclusion of altered octaves is interesting in itself;
these intervals reappear in his theory of species (Book 111). 17
The second set of definitions refers to the number of "melodically proximate sounds" contained in the designated span: e.g.,
a pentachordum is comprised of five different adjacent sounds
(the diatonic set is assumed). The latter definition is retained in
the treatise; the former is somewhat superfluous, as specific interval designations (diapente cum tono or cum semitonio) are
used everywhere else and are more precise.
Except for two brief chapters on the genera, Book II is devoted to the question of consonance (which had been raised,
but not settled, in Book I). Theinred takes the position, inherited from antiquity, that intervals of simpler ratios are more
consonant than those with relatively complicated proportions.
However, he also makes allowances for the imperfection of the
ear <I, i, 12-13>:
These are the ratios under consideration: very many dissonances are
admitted into the number of consonances by the coarseness 18 of the
ear, as are the ditone, diapente cum semitonio, triple and quadruple
diapason, or those intervals that, by a small amount, exceed or fall
short of the proportions of a consonance; of these, a man cannot at all
hear the dissonance.
[f.2r] Sunt rationes namque consideratione: dissoni quam plurimi
auditus ruditio in numerum admittuntur consonantiarum, ut sunt
"See Snyder, chap. 6.
18 Probably from ruditium (coarseness), derived from rudire (to be rude), a
known medieval verb. Ruditio is also a known medievalism for eruditio; the
obvious necessity for an ablative here would require double emendation: fel ruditio[nd , so the other reading has been preferred.

Downloaded from http://mts.oxfordjournals.org/ by guest on January 31, 2014

Theinred clearly derived his scheme from one proposed by


Ptolemy and quoted by Boethius. 15 The earlier approach, however, was not strictly dichotomous, having quadripartite division at the final stage: consonances, dissonances, "melodic intervals," and "unmelodic intervals." This is because Ptolemy
(and Boethius) recognized harmonic and melodic consonances; Theinred is here concerned only with harmonic consonance and needs no extra categories at the lowest level.
Chapters two and three essentially repeat, very briefly, the
classificatory apparatus given above, in slightly different terms.
Their function is primarily rhetorical.
In the fourth chapter ("Whence the names of individual harmonies"), Theinred assigns specific names to the members of
his various classes of sound. The first named are the diapason,
diapente, diatessaron, and tonus. He says that the diapente and
diatessaron are so called because they contain five and four
sounds, respectively. The interval containing the diapente and
diatessaron and having eight sounds is called diapason. The
"principal interval" (the one distinguishing the diapente and
diatessaron) is not named for the number of sounds it contains,
but for the sound of musical melody: tonus. 16 The remaining
intervals are to be named by "adding and subtracting" the four
previously defined. In reality, the names are nearly all derived
by addition (e.g., diapente cum tono); Theinred has also failed
to mention the semitone, which is derived by the subtraction of
a ditone from a diatessaron, but which is named by division.

113


114

Music Theory Spectrum

[f.2v] termini ditoni, diapente cum semitonio, ter et quater diapason,


vel qui a consonantia minima particula gravioris et acutioris termini
superabundant vel deficiunt. Quorum dissonantias auditus percipere
vel vir etiam nullatenus potest.

For if nothing were heard, there would not have existed any argument
concerning sound whatsoever. . . . the Pythagoreans take a certain
middle path. For they measure consonances themselves with the ear,
but the distances by which consonances differ among themselves they
do not intrust to the ears, the judgments of which are unclear, but to
rules and reasonas though the sense were an obedient servant, and
the reason were truly a commanding judge. 19

Theinred also attempts "a certain middle path"; however, he


succeeds mainly in confusing the issue.
Theinred begins by setting up a series (progressio) of proportions for each of the two types of proportionality capable of producing consonances, from simple to complex, as follows
(<II,ii>, f igv) 20
Multiples: 2:1 3:1 4:1 5:1 6:1
Superparticulars: 3:2 4:3 5:4
.

19 Boethius

I,ix; quoted from Bower's translation, p. 58.


and medieval mathematics dealt with proportions in terms of
five classes: (1) multiples are fractions in which the numerator is a multiple of
the denominator (e.g., 2/1, 3/1); (2) superparticulars are proportions in which
the numerator exceeds the denominator by one (e.g., 3/2, 4/3); (3) superpartients are fractions in which the numerator exceeds the denominator by two or
more, but is less than twice the denominator (e.g., 5/3, 7/4, 7/5); (4) multiple
superparticulars are proportions in which the numerator contains the denominator twice or more times, with a remainder of one (e.g., 5/2, 7/3); (5) multiple
superpartients are fractions in which the numerator contains the denominator
twice or more times, with a remainder of two or more (e.g., 8/3, 12/5). It is
20 Classical

The reasons for these orderings are not given and are not at all
clear.
Theinred states that when the point of interruption (the
number five) is reached, the progression of consonances is terminated. Thus, when the multiple series reaches 5:1, this interval (double octave plus major third) and all succeeding
multiple-proportion intervals will not be consonances. This rule
excludes from consonance the sextuple proportion (6:1, two octaves plus fifth) and the triple octave (8:1). Likewise, the superparticulars are consonant through the sesquitertia (4:3, perfect
fourth); superparticulars from the sesquiquarta (5:4, just major
third) onward are not consonant. Since the simplest superpartient (5:3, just major sixth) is not a consonance, no superpartient
assumed that proportions are stated in simplest terms: 14/4, for example, is not
a multiple superpartient but another manifestation of the multiple superparticular 7/2. Musical consonance, except for the problematic case of the diapason
cum diatessaron (to be discussed below), was associated with the multiple and
superparticular proportions.
21 Respectively: octave, twelfth, double octave, and double octave plus
fifth.
22 Respectively: perfect fifth, perfect fourth, tone.
23 Respectively: minor seventh, major sixth, minor third.
24 Respectively: octave plus major ninth, major ninth, double octave plus
perfect fourth.
25 Respectively: perfect eleventh, double octave plus major sixth, major
thirteenth.

Downloaded from http://mts.oxfordjournals.org/ by guest on January 31, 2014

Theinred expresses some doubt about the subjective approach,


however, wondering if "to admit the judgment of the ear is to
deny reason" (<I,i,14>). Boethius had also addressed this
question, arguing that not all judgment should be given to the
senses, which are fallible (reason is held to be infallible); yet he
admits that the art of music depends upon hearing:

Theinred does not provide series for the other types of proportionality, as only these two classes (theoretically) produce consonances; other types of proportions are represented, however,
in the series of intervals, which are grouped according to
proportion-type (<II,v> , f.19r):

Multiples: 21 2:1
3:1
4:1
6:1

Superparticulars: 22 3:2
4:3
9:8

Superpartients: 23
16:9
27:16 32:27
Multiple superparticulars: 24 9:2
9:4
16:3

Multiple superpartients: 25
8:3
27:4
27:8

Theinred of Dover on Consonance

And so: the first (simplest) proportion in any type of proportionality is in sound the most consonant. And so far as a proportion is removed from the simplest [of its type], so much less is it consonant
unless it seems to approximate the proportion of some consonance, as
the ditone, which is near in sound to the sesquiquartian proportion
(5:4), which is the first [superparticular] after the sesquitertia (4:3):
the former exceeds the latter by one eighty-first, 27 which is difficult to
hear. Or the semiditone, which is near to the sesquiquinta (6:5), which is
the second [superparticular] after the sesquitertia: the latter is exceeded by the former by only two twenty-fifths of the smaller term;
but this is easily heard. 28

26 <II,vi,be>

(f.19v).
has the relationship right, but we would prefer to say that the
ditone (81:64) exceeds the sesquiquarta (80:64) by one eightieth, not by one
eighty-first.
28 Theinred's treatment of the two minor thirds is a sleight of hand, so far as
the proportions are concerned: he says that the semiditone (32:27) is exceeded
by the sesquiquinta (30:25) by two twenty-fifths of the smaller term. The semiditone is indeed larger than the sesquiquinta, but to compare the denominators
in this way, without a common numerator, is senseless. The actual difference
27 Theinred

[f.19v] Omnis namque generis, prima proportio in sonis consonissima est. Et quanto magis [f.20r] a prima elongatur, tanto minus consona invenitur, nisi propinquitate alicujus consonae proportionis consona videatur, ut ditonus, qui sesquiquartae sonorum proportioni
quae prima sequitur sesquitertiam adeo propinquus est, ut octogesima prima tantum parte majoris termini hic superet hanc: quod
auditu percipere difficile est. Vel semiditonus, qui sesquiquintae
sonorum proportioni quae secunda sequitur sesquitertiam adeo propinquus est, ut duabus tantummodo vicesimis quintis partibus minoris termini, haec superetur ab hoc: quod auditus aut facile discernit.

Theinred thus hedges on his own scheme: he admits that the imprecision of the ear (mentioned earlier) may cause intervals of
complex ratios to be accepted as consonances, if they closely approximate the proportions of some recognized consonance. But
here he allows thirds on the grounds that they approximate the
5:4 and 6:5 intervalssounds which were previously excluded
from the canon of consonances! Theinred's vacillation on this
issue is certainly an honest reflection of his musically revolutionary times, but it is not otherwise very helpful. Theinred concedes that 5:4 and 6:5 follow 4:3; he thus seems to imply that the
boundary between consonance and dissonance is not a matter
of black and white, but rather a process of continuous shading
from one to the other. Theinred does not go so far as to create a
category of "intermediate consonances" 29that idea appeared

between the two major thirds is, of course, exactly the same as that between the
two minor thirds. Fritz Reckow quotes this passage in a lengthy footnote in his
edition of the treatise of Anonymous IV, p. 78, and suggests emending the last
sentence: ". . .quod auditus aut [ = haud] facile discernit" ("which the ear does
not easily discern"). This suggestion is quite tempting, as it at least renders the
last clause as a true statement, but I have not adopted it because it seems to
make a correct conclusion follow faulty premises. Theinred has certainly erred,
but I think that the mistake is in his comparison of the semiditone and sesquiquinta, not in his subsequent reasoning.
29 Gut interprets Theinred's remarks as equivalent to the use of a category
of "intermediate consonances." This is reasonable as a summary but does not
reflect the actual complexity (or confusion) of Theinred's discussion.

Downloaded from http://mts.oxfordjournals.org/ by guest on January 31, 2014

can be a consonance; likewise with the multiple


superparticularsthe non-consonant status of the simplest
(5:2, just major tenth) insures that no ratio of this type of proportionality will produce a consonant interval. 26
Chapter viii is a reiteration of the above, in different terms.
Chapter ix summarizes the discussion to that point by naming
the accepted consonances: diapason, diapente, diatessaron,
diapason cum diapente, double diapason, and diapason cum
diatessaron. They are given in order of increasing complexity of
their proportions.
Theinred's theory of consonance, if somewhat confused,
does not depart radically from the ideas of his ancient models.
However, a twelfth-century theorist must account in some way
for the growing acceptance of thirds and sixths as harmonic consonances, a problem which Theinred addresses in a telling passage <II,vi,f>:

115

116

Music Theory Spectrum

slightly later 30but it is clear that he was wrestling with the


same problem, and, in a way, came to a similar conclusion.
The passage quoted above may be contrasted with Walter
Odington's oft-cited remarks:

Superpartiens vero nec integrum servat nec singulas admittit partes


sed plures partes, et ideo ipsa relicta est. Et per consequens ditonus et
semiditonus non sunt symphoniae.
Verumtamen quia vicinae sunt sesquiquartae et sesquiquintae habitudinis quarum unitas facit differentiam, iccirco plurimos estimant
consonas esse. Et si in numeris non reperiantur consoni, voces tamen
hominum sua subtilitate ipsos ducunt in mixturam suavem et penitus
in consonum quandoque dulcedine nota fit consonum, in vocis ruditate offendit auditum. 31

While Odington attributes the apparent consonance of thirds to


performance, Theinred regards the phenomenon as a matter of
the imperfection of human perception.
30 Particularly

in the work of Johannes de Garlandia; see below.


of Odington, Summa de Speculatione Musicae, ed. Frederick
Hammond, Corpus Scriptorum de Musica 14 (Rome: American Institute of
Musicology, 1973), II. 10. 13-16 (pp. 70-71). Gut, p. 22, quotes this passage
(beginning with "Verumtamen. . .") and cites the Hammond edition; what is
printed, however, is the reading of Edmond de Coussemaker (Scriptorum de
musica medii aevi nova series [Paris: 1876], vol. 1, p. 199), which differs from
Hammond's edition in several places.
31 Walter

Why the ditone and semiditone may nevertheless


be admitted into organa
And so of these, the ditone and semiditone, with their octavecomplements: on account of the octaves they are admitted into organum with the other consonances, the former (with the diapente cum
tono) more often, because they more closely approximate the proportions of a better consonance; but the latter (with the diapente cum
semitonio) less often, as they less closely approach the proportions of
a lesser consonance.
[f.20r] Quare tamen tonus semiditonus admittantur in organa
Unde horum ditonus et semiditonus cum sibi convidentibus
aequisonas consonantias, propter aequisonantiam cum consonantiis
admittuntur in organa; hic quidem tamen cum diapente cum semitonio saepius propter consonioris proportionis majorem propinquitatem, hoc [MS: hic] vero cum diapente cum tono rarius, propter minus
consone proportionis minorem propinquitatem.

Although he has ascribed consonance to these intervals on the


basis of their octave-complement relationships, Theinred cites
usage patterns for pairs of intervals that are not octave complements. Furthermore, Theinred's argument itself is fallacious:
seconds and sevenths (and augmented fourths and diminished
fifths) also combine to form octaves, and, according to
Theinred's line of reasoning, should likewise be allowed in organum. It must be admitted that the thirds and sixths are closest
to the consonances, but Theinred's reasoning is basically faulty.
Beyond the theory, his mention of the relative frequency of occurrence for these intervals is of some interest: is Theinred reporting observed patterns of use or prescribing on the basis of
his theory? Theinred gives us no clue as to what music he had in
mind, if any, but we may observe that the thirteenth-century
hymn to St. Magnus, "Nobilis, humilis," of English origin, 32
32 Historical Anthology of Music, comp. and ed. Willi Apel and Archibald
T. Davison (Cambridge: Harvard University Press, 1949), sel. 25c. The date

Downloaded from http://mts.oxfordjournals.org/ by guest on January 31, 2014

But a superpartient neither preserves integrity nor gives out single


parts [as remainders], but gives out [as remainders] many parts, and
therefore it has been passed over. And as a result, the ditonus and
semiditonus are not consonances.
But again, since one unit makes the difference, [the ditone and
semiditone] are very close in property to the sesquiquarta and sesquiquinta ; on this account many regard them as consonances. And if
the consonance is not found in numbers, nevertheless, the voices of
men by their subtlety lead them [i.e., these intervals] into a sweet mixture and entirely into a consonance; and when a note is made sweetly
consonant, in the coarseness of the voice it offends the ear.

Theinred must then deal with the appearance of thirds and


sixths in organum <II,vii>:

Theinred of Dover on Consonance

of this piece has been a matter of some debate: Apel regards it as a twelfthcentury composition, but see, for example, John Beveridge, "Two Scottish
Thirteenth-Century Songs," Music and Letters 20 (1939): 352-64, and Anselm
Hughes' discussion in the New Oxford History of Music, vol. 2 (rev. ed., London: Oxford University Press, 1967), pp. 341f.
33 Reaney, pp. 29-30.

colorful (and sometimes obscure) metaphors, the last and clearest and most musical of which runs thus <II,x,a,6-7>:
Lyres, differing from each other by a tone, when heard singly, delight
the ear by the sweetness of melody, but moved simultaneously, they
roughen the sounds. So the superpartient, which by itself could not be
consonant, is consonant when multiplied, just as the multiple superparticulars can in no way be consonant, though the multiples and superparticulars are in themselves consonant.
[f.201 Lirae autem tono in omnibus sonis suis distantes, singulae
autem audientis auditu[m] modulationis suavitate oblectant; simul
vero motae, voci[b]us exasperant. Ita superpartientia cum multiplicate consonat, quae per se consonare non possit sicut multiplicitas
[et superparticularitas in margine] per se consonant [quae conjunctae
consonare nequeunt in margine]

Theinred apparently recognizes the weakness of his position, as he offers two further explanations for the consonance
of the eleventh (<x,b>). The first is dependent upon the fact
that consonances, which are not directly associated with superpartient proportions, are nevertheless divisible into (unequal)
superpartient ratios. Theinred presents interesting and apparently unprecedented divisions of the octave and fifth: 8:11:16
and 10:13:15. 34 The point would seem to be that if the proportion of the eleventh could be similarly divided, it should be considered a consonance. Theinred attempts no such division,
even though one is readily available (3:5:8); perhaps he considered this sufficiently obvious to his readers that only a hint was
necessary.
The foregoing is still weak, so Theinred invents yet another
means of justifying the diapason cum diatessaron as a consonance <II,x,b,6-8>:
Furthermore, if any consonant interval is subtracted from an
equisonal consonance [i.e., an octave or double octave], in fact that
34 Respectively:

551 + 649 cents; 454 + 248 cents.

Downloaded from http://mts.oxfordjournals.org/ by guest on January 31, 2014

contains 26 major thirds and 19 minor thirds. Unfortunately,


there is not sufficient material from the twelfth century to warrant firm conclusions.
Gilbert Reaney has expressed the opinion that this chapter
on thirds "fits better into the later 13th century, when thirds
were becoming important in practical music (for instance, in
gymels like Edi beo), and when they were being discussed by
Walter Odington (early 14th century)." 33 However, discussions
of thirds in organum occur from the end of the eleventh century
on (see above); furthermore, Theinred's discussion of the issue
would have been badly outdated if it had come after Garlandia's
(see below). Additionally, it should be noted that Theinred
does not treat rhythmic considerations at all, so that his remarks
are relevant mainly to relatively early organum. Therefore, it
seems better to regard this treatise as belonging to the twelfth
century.
Finally, Theinred must deal with the diapason cum diatessaron (the perfect eleventh), the consonance of which had been
a matter of much controversy from antiquity. The ratio of the
eleventh is 8:3, a multiple superpartient (in fact, the simplest
ratio of that class). Superpartients had been excluded from the
consonances altogether, so Theinred resorts to an argument
that is roughly "the whole is of a different character than its
parts." That is, that the superpartient type of proportion, which
cannot by itself yield consonances, can do so if combined with
the multiple type of proportion. Conversely, the multiple superparticulars are not considered consonances, although the multiples and superparticulars, individually, do produce consonances. To illustrate his point, Theinred employs several

117

118

Music Theory Spectrum

which remains will be a consonance. But if the interval subtracted is a


dissonance, that remaining will be a dissonance. And so, if from a
double diapason a diapente is subtracted, a diapason cum diatessaron
[remains], and [thus] ought to be counted among the consonances.

We may contrast this with Boethius' solution:


[The diapason] is like the number 10; for if some number within ten be
added to it, the number is preserved just as it was before. . . . if you
add [2] to 10, you will produce 12, and the 2 has been preserved in
conjunction with the 10. . . .
In the same way the diapason consonance can be united to any
other consonance, and it preserves the consonance rather than
changes it, and it does not yield a dissonance from a consonance. For
just as a diapente consonance joined to a diapason preserves a diapason and diapente consonance, of course in the triple proportion, thus
if a diatessaron also be joined to the diapason, it produces another
consonance. Thus according to Ptolemy, the diapason and diatessaron should be added to the list of consonances. 35

Theinred cannot adopt this approach, however, as it would allow the triple diapason to become a consonance (double diapason plus a diapason), and so on. His subtraction method apparently is contrived to allow (1) the diapason cum diatessaron as a
consonance, but (2) still limit the total number of consonances
to six.
To summarize his position: Theinred defined consonance in
terms of harmonic compatibility of sounds. In this he followed
then-current trends in music theory. Unlike his immediate
predecessors, however, he proposed a rather elaborate, if
mathematically confused, rationale that was intended to ac35 Boethius

V,x; quoted from Bower, pp. 309-10.

After Theinred: Resolution of the Problems


Although the fundamental problems of the categorization of
consonance/dissonance and of the number of consonances were
not solved for some time after Theinred, some new ideas on
these subjects began to appear within a half century or so of his
work. About the year 1200, the anonymous author of the De
musica libellus proposed a threefold division of consonances:
It is to be noted that the unison and diapason are perfect consonances,
the ditone and semiditone are imperfect consonances, and the diatessaron and diapente are called intermediate consonances.
Notandum est quod unisonus et diapason sunt consonantie perfecte;
ditonus et semiditonus sunt imperfecte; diatessaron et diapente dicuntur medie. 36

In his treatise, De mensurabili musica, 37 which appeared about


a century after Theinred's (i.e., about 1240), John of Garland
continues the process already begun:
36 Anonymus VII, De musica libellus, ed. Edmond de Coussemaker, op.
cit., vol. 1, p. 382.
37 Johannes de Garlandia, De mensurabili musica, ed. E. Reimer (Wiesbaden: Steiner, 1972).

Downloaded from http://mts.oxfordjournals.org/ by guest on January 31, 2014

[f.2 1 1 Amplius si quid absciditur ab aequisona consonantia, si abscisum consonantia est, et quod relinquitur consonantia erit. Si vero abscisum dissonantia est, et quod relinquitur dissonantia erit. Si itaque a
bis diapason, diapente absciditur, et diapason cum diatessaron inter
consonantias computari debet.

count for the accepted consonances, explain the phenomenon


of the thirds, and limit the total number of consonances to six.
His ambivalent treatment of thirds and sixths foreshadowed
their later classification as intermediate consonances, but
Theinred himself did not subdivide the consonances except into
equisonal and inequisonal groupsand it is not clear just where
the thirds and sixths fit into that scheme. Furthermore, his failure to recognize the principle of octave equivalence led him to
regard such intervals as the triple octave as technical dissonances (though he recognized all octave multiples as equisons).
This in turn led him to rather contorted efforts to prove the consonance of the eleventh.

Theinred of Dover on Consonance

Sequitur de consonantiis in eodem tempore. Consonantiarum


quaedam dicuntur concordantiae, quaedam discordantiae. Concordantia dicitur esse, quando duae voces iunguntur in eodem tempore,
ita quod una vox potest compati cum alia secundum auditum. Discordantia dicitur contrario modo. Concordantiarum triplex est modus,
quia quaedam sunt perfectae, quaedam imperfectae, quaedam mediae. . . .
Et sciendum, quod supradictae concordantiae possunt sumi in infinitum. . . .
Discordantia dicitur esse, quando duae voces iunguntur in eodem
tempore, ita quod secundum auditum una vox non potest compati
cum alia. Discordantiarum quaedam dicuntur perfectae, quaedam
imperfectae, quaedam mediae. . . .Iste species dissonantiae. . .possunt sumi usque in infinitum sicut et concordantiae. 38

John has thus proposed two different and radical changes in


harmonic theory. First, he has introduced subcategories not
only of consonance, but also of dissonance, at which earlier
writers had only hinted. Secondly, he has recognized the principle of octave equivalence. He thus replaces the older notion of
a fixed number of consonances with a more flexible concept: a
certain number of species of consonances. This new approach
saves both the inconsistency that was inherent in Boethius'
38 Ibid.,

pp. 67-68, 70-72.

work and the contortions Theinred needed to circumvent that


inconsistency. Of course, these new ideas did not gain immediate acceptance. Anonymous IV, writing about 1275, indicated
that thirds were widely used as consonances, though there was
some disagreement on their classification:
The odd-numbered notes of the first mode in the duplum are
aligned in consonance with the odd-numbered notes of the first mode
in the first voice, or tenor, and this is done according to the consonance of the unison or diapason for perfect concords, of the diatessaron or diapente for intermediate consonances, or of the semiditone
or ditone for imperfect consonances, although among some the ditone and semiditone are not thus reckoned. Nevertheless, among the
best composers of polyphonic music, and particularly in certain
places, such as in the part of England called Westcountry, they are
called the best consonances, because among such people they are so
much in use.
Puncta imparia primi modi in duplo se habent in concordantia ad
puncta imparia primi modi in primo sive tenore, et hoc secundum
concordantiam unisoni vel diapason pro concordantia vel concordantiis perfectis vel diatessaron diapente pro concordantiis mediis vel
semiditono ditono <pro concordantiis imperfectiis>, quamvis ditonus et semiditonus apud aliquos non sic reputantur. Tamen apud
organistas optimos et prout in quibusdam terris sicut Anglia in patria,
quae dicitur Westcuntre, optimae concordantiae dicuntur, quoniam
apud tales magis sunt in usu. 39

A quarter-century later, however, Walter Odington still questioned the consonance of the diapason cum diatessaron, eventually classifying it as a dissonance; nor did he accept sixths as
consonances:
[The diapente cum tono] consists in 16 plus 11 of its parts (that is, sixteenths), and this genre [of proportionality] was previously rejected
[as a source of consonances]. Nevertheless, in the music of this time,
the diapente cum tono is many times used as if it were a consonance.
39 Reckow,

pp. 77-78.

Downloaded from http://mts.oxfordjournals.org/ by guest on January 31, 2014

The following concerns simultaneous harmonies. Some simultaneities are called "concords," others "discords." It is called a concord
when two voices are joined at the same time, such that one voice can
be accepted with the other by the ear. A discord is defined as the reverse. Concords are of three classes, for some are perfect, some imperfect, and some intermediate. . . .
And it is to be known that the above-named concords can be extended into infinity. . . .
It is called a "discord" when two voices are joined simultaneously,
such that, according to the ear, one voice cannot mix with the other.
Some discords are said to be perfect, others imperfect, and others intermediate. . . .These species of dissonance . . . can be extended
into infinity, just as the concords.

119

120

Music Theory Spectrum

When that shall happen, since it is uncommon, let it be made more


smoothly. . . .
There are six discordant harmonies: the semiditone, ditone,
diapente cum tono, diapason cum semiditonus, diapason cum ditonus, and diapason cum diatessaron. About these enough has been
said.

The consonant/dissonant status of particular intervals continued to be debated for some time (in a few cases until our own
century), but a basic framework for the discussion had been established. The remaining difficulty with which Theinred wrestled, the difference between "pure" and "Pythagorean" thirds,
4 11.12.5-7 (p. 74), 11.15.2-3 (p. 75).

Downloaded from http://mts.oxfordjournals.org/ by guest on January 31, 2014

Continent enim 16 et eius 11 partes, scilicet sextas decimas, et haec


species prius repudiata est. Verumtamen per diapente cum tono proceditur multotiens in cantilenis istius temporis quod fit ut concordia.
Cum acciderit, quia rara est, fiat suavior. . . .
Concordes discordiae sunt sex: semiditonus, ditonus, diapente
cum tono, diapason et semiditonus, diapason et ditonus, diapason et
diatessaron. De his satis dictum est. 4

was of course not resolved until the invention of tempered tunings during the Renaissance.
In conclusion, we may say that Theinred of Dover both followed and kept pace with musical-theoretical history, and that,
if he had little actual influence, he did anticipate some important later developments. He generally accepted and worked
from the notions of consonance and dissonance he had inherited from antiquity but struggled to reconcile those venerable
ideas with new trends emerging in his own day. Thus, he defended the classical canon of six consonances but sought to justify the use of thirds and sixths in organum. He regarded multiple octaves beyond the bis diapason as technical dissonances but
admitted that the ear cannot perceive their dissonance. He
clung to Pythagorean intonation but explained the growing acceptance of thirds partly on the grounds that they approximate
another (just) intonation. Theinred was surely only dimly aware
of the ramifications of his work, yet, the fact is that he was an
early exponent of ideas which, when their time had come, produced far-reaching results. Despite his inconsistencies, he has
earned a place in the ranks of original musical thinkers.

You might also like