Circulating Air: Inspiration, Voice and

Soul in Poetry and Song

SARAH KAY

Make me thy lyre, even as the forest is:

What if my leaves are falling like its own!
The tumult of thy mighty harmonies
Will take from both a deep, autumnal tone,
Sweet though in sadness.

These lines from Shelley’s ‘Ode to the West Wind’1 call on the wind
to produce music from the poet as it does from the forest, even at the
cost of the poet’s own leaves being torn adrift like those of the trees,
even though what they jointly sound is an autumn knell. The poet’s
words vibrate like the notes of a lyre, as music rather than as text;
the scattering leaves suggest the absence of any fixed support, pure
sound without the permanence of writing. The breath of the west
wind connects Shelley to a poetry that is oral and sung, conjuring him
into a company of former poets who sang in woodlands to the sound
of a lyre. The same wind that infuses him with poetic inspiration has
been blowing and inspiring singers for centuries, although now, in its
autumn, it may be nearing an end. Poetry is born from circulating air
that blows in and through the breath of countless individuals across
time and space.
Among the earlier poets evoked by Shelley’s lines and breathing the
same air as he we can identify the shepherds of Classical eclogues,
medieval troubadours or the Petrarch of the Canzoniere — although
when they were singing the breeze was typically still in its spring
season. Air and leaf are strikingly compounded in Petrarch’s signature
pun laura in which wind, laurel and the beloved all combine. Shelley’s
west wind could be a belated and blustery return of the aura serena

Paragraph 41.1 (2018): 10–25

DOI: 10.3366/para.2018.0247
© Edinburgh University Press
www.euppublishing.com/para
 Circulating Air 11
of the Renaissance poet’s sonnet 196: ‘The calm breeze that comes
murmuring / through green leaves to strike my face, / makes me
recall how Love struck me / the first wound, so sweet and deep, / and
[makes me] see the lovely looks, otherwise hidden from me’.2 Already
in Petrarch this wind is a reiteration of earlier inspirations, not just of
the ‘first wound’ of Love but also of the many invocations of l’aura by
the troubadours, most famously Arnaut Daniel, which in turn go back
at least to Ovid.3
Scholarship on Petrarch, and to a lesser extent on the troubadours,
has been encouraged by texts like sonnet 196 to theorize the
workings of the wind as inspiration. Drawing on an intuitively self-
evident connection between air, breath, poetry and soul, critics
have constructed a genealogy of the poetic image (‘the lovely
looks, otherwise hidden from me’) as produced in the imagination
through the workings of spirit, where ‘spirit’ retains its etymological
associations with wind and breath as well as designating an animating
force or ‘soul’. In-spiration is literally an inhalation which kindles
the warmth of an inner spirit and shapes within it a phantasmatic
object that is an imaginary precipitate of what the poet desires but
has not yet attained in the outer world. The best-known exponent of
this account is Giorgio Agamben for whom ‘Medieval phantasmology
was born from a convergence between the Aristotelian theory of
the imagination and the Neoplatonic doctrine of the pneuma as a
vehicle of the soul, between the magical theory of fascination and the
medical theory of the influences between spirit and body’ (Stanzas,
22).4 The strands which make up this philosophical nexus are carefully
identified and articulated in ‘Eros at the Mirror’ (Stanzas, 73–89), a
chapter-long review of the antique and medieval thinkers — Plato,
Aristotle, Avicenna, Averroes, William of Conches, Thomas Aquinas
— who continue to dominate this tradition of critical thinking about
inspiration, spirit and image.5 This philosophical configuration enables
Agamben to link together chapters on the Romance of the Rose, Dante’s
Vita nova and the poetry of the Stilnovisti, supporting his overall
contention that the medieval doctrine of pneuma and the pneumatic
imagination, framed by the ‘stanza’ or poetic space, provides a unique
relation to the love object. Poetry recasts the object’s constitutive
absence as a phantasmatic presence that is both available for enjoyment
and located within a faculty accessible to cognition (the imagination),
thereby achieving a poetic resolution of the book’s initial problematic,
‘the impossibility, for Western culture, of fully possessing the object of
knowledge’ (Stanzas, xvii).
12 Paragraph
A consequence of this account of inspiration is that it almost
instantly consolidates inspiration as vision, albeit in the guise of the
inner look of phantasy. By the same token, it advances a view of
poetry as more visual than aural, as landscape rather than as soundscape.
This article stems from my sense that to prioritize image over sound
in this way often runs the risk of critical distortion. In the lines of
Shelley with which I began, the wind-tossed trees of the autumnal
landscape are significant mainly for the music they produce, and the
laura of the Canzioniere works both as aural wordplay in Italian and as
an echo of earlier punning in the troubadours’ Occitan. The majority
of troubadour songs are as poor in visual images as they are rich in
sound; a large proportion of their ‘imagery’ is more acoustic than
visual, notably the evocations of birdsong or of other animal noises.
Only one song that I know of, by Folquet de Marselha, refers to an
inner image of the beloved (‘for within my heart, lady, I carry your
likeness’).6 For one canso by Bernart de Ventadorn that features an
image seen in a mirror as an organizing device, two others use echoing
patterns of repetition to produce a ‘mirroring’ that is entirely aural.7
And evidence of dialogue between the troubadours, or of the influence
of one on another, is often traced through shared metrical and musical
schemes.8
Cogent though Agamben’s approach to pneuma may be, therefore,
I do not think it is ideally suited to many of the texts for
which it was devised. In the next two sections I retrace some of
the steps that have served to theorize this pneumatic–phantasmatic
assemblage, rearranging some of the same elements in order to favour a
pneumatic–acoustic relation over a pneumatic–visual one and thereby
articulate a different configuration of spirit and desire. The final
section uses the results of this theoretical discussion to throw light
on the text of the song ‘Amar dei’ (I am compelled to love) by the
troubadour Bernart Marti. Although I conclude with a medieval text,
the circulation of air by definition resists confinement to a single period
and I hope my conclusions may be found relevant to the poetics of
other times.
Returning, then, to what I earlier described as the self-evident
connection between air, breath, poetry and soul, I will insert another
term into the series, that of voice. And rather than filtering their
connection through a Neoplatonist lens, I will stay with Aristotle and
the various texts whose treatment of voice in language or music refers
back to him.
 Circulating Air 13
Voice in Aristotle
At the beginning of De interpretatione, Aristotle states that the voice
symbolizes affections in the soul, whereas writing is made up of
symbols of the sounds made by the voice, referring the reader to his
treatise On the Soul for more detail (16a3ff.).9 Following Aristotle’s
advice is initially rather disconcerting, since the two works appear
to have little in common: De interpretatione concerns language and
logic whereas On the Soul is more a study in comparative biology.
In the major part of the latter treatise — Book 2 and the beginning of
Book 3 — Aristotle inquires how sensation and perception contribute
to an understanding of ‘soul’ defined as ‘the cause or source of the
living body’ (415b9; 661). The centrality of this discussion to On the
Soul bears out Daniel Heller-Roazen’s contention that for Aristotle,
as in ancient thought more generally, sensation takes the place of
what Moderns call ‘consciousness’.10 Aristotle argues that the soul
responsible for the processes of sensation and perception is a soul
common to humans and other animals, often known as the ‘animal
soul’. Anything which is endowed with sensation also has the capacity
for appetition, a term which for Aristotle comprises desire, passion
and wish; all animals possess at least the power of touch, and thus
of desiring (414b1–6). When Aristotle refers to the voice in De
interpretatione as registering ‘affections in the soul’ (16a3; 25), he clearly
has in mind animal reactions such as these.
In his exploration of sensation and perception, Aristotle addresses
the external senses and their objects one by one, starting with hearing
and sound (419b3ff.). Sound is defined as resulting from the impact of
solid bodies against one another and against the air (419b20). Or better,
the air itself is one of these bodies, but it must be hit sharply and hard
so that it ‘does not retreat before the blow, i.e. is not dissipated by it’
(419b21; 668).11 A sound-producing impact — like cymbals clashing
or a whip cracking — unifies the air and hurls it in such a way as to give
rise to a secondary sound or echo when the moving air strikes against a
containing surface. Typically, this echo is inaudible, but it provides the
clue to how hearing works. Air that has been struck moves in a single
mass and enters the containing chamber of the ear, where it impacts
the tympanic membrane (420a4ff.). Hearing is seemingly not simply
the passive reception of this echo-like effect but its active production,
since, as Aristotle goes on to say, sound results both from the striking
body and from the body that is struck. Nevertheless ‘the sound we
hear is always the sounding of something else, not of the organ itself ’
14 Paragraph
(420a16; 668). Sound thus always has a spectral dimension, an unheard
double of what is actually heard.
If sound is a being struck that is also a sounding out, then voice too
is a kind of blow, a particular way of striking air which can be done
only by something animate, that is, something that has soul (420b5).
Voice, that is, is a form of sound, and given that sound requires air, only
animals that breathe have a voice, such animals also having warm blood
(420b9–11). Voice takes the form of a blow struck between the outer
air and this inner soul: it consists in a violent impact on the passage of
breath between the two (420b14–16). Breath serves the dual purposes
of maintaining bodily warmth (through the blood) and enabling
vocal expression, much as the tongue likewise has two functions,
nourishment and pronunciation (420b16–22). Not all animals breathe,
so not all animals have a voice, but conversely voice is not the
unique preserve of human animals. Indeed, in History of Animals 4.9,
where Aristotle elaborates this co-dependence of voice and breath,
the sounds emitted by many kinds of animals are enumerated in
what (compared with the philosopher’s concision elsewhere) feels like
lingering detail, their many different voices carefully distinguished
from the non-vocal noises which insects produce by friction with
the air (535b26–536b24).12 According to History of Animals 536b2,
language is the exclusive domain of humans, but in On the Soul the
lines between humans and other animals are murkier;13 voice and the
whole physiology of speech and hearing are activated by the animal
soul as part of the overall functioning of sensation.
The existence of voice connects human and nonhuman animals, and
also relies on the more basic continuum of circulating air. Aristotle
appears more than once to collapse into a single moment the two
movements whereby air is breathed in and subsequently emitted as
voice. Here, for example, voice is associated with the ‘inbreathed air’:
‘Voice then is the impact of the inbreathed air against the windpipe,
and the agent that produces the impact is the soul resident in these parts
of the body’ (420b28–9; 670). This foreshortening highlights how
voice as a percussive act depends both on connection (breath comes
in and goes out) and on rupture (between external air and internal
soul). Voice marks the boundary between the animated body and the
cosmos outside it, but it also depends on recycled air that undoes this
boundary as fast as it is made.
Although De interpretatione is an introduction to reasoning, through
its reference to On the Soul it presents voice as sound that interrupts
the flow of breath to register the movements and appetites of the
 Circulating Air 15
animal soul. Voice forms part of a soundscape that involves all of the
sounds discussed in History of Animals but is more than just acoustic.
It belongs in the epistemology of sensation and perception, and of the
passions and desires which they provoke. What a voice says may carry
over into different domains (dialectic, for example) but its primary
function is to record these ‘affections’. Voice also belongs in a broad
ontology, one that distinguishes animate bodies that breathe from all
other bodies and especially from plants. Because voice both relies
upon and sharply interrupts the flow of air between what is inside
and outside an individual, voice both bounds body and opens it.
Inserting voice into the series breath–air–poetry–soul articulates it
quite differently from Agamben’s account, with its culmination in a
phantasmatic inner landscape. The Aristotelian configuration is less
mystical, less anthropocentric, more corporeal and material, and more
directly addressed to desires for objects in the external world. Voice is a
physiological manifestation of, and also gives expression to, a breathing
body’s animation as an animal.

Voice in Grammar, Music and the Latin Aristotle

The Latin grammatical tradition that extended from Classical Rome
through the Middle Ages presents voice in a way that often seems to
echo Aristotle; echoes spread to treatises of music; and the terminology
of grammar and music comes back into the Latin rendering of
Aristotle, creating interconnections between all these writings.
Voice is the grammarians’ starting point, just as it marks the
beginning of De interpretatione. Varro, a contemporary of Cicero,
influentially defined voice in terms that identify it primarily as sound:
Utterance (vox), according to the Stoics, is a subtle breath (spiritus) perceptible to
the ear, according to its power. Now it is made either by an impulse of rarified air
(exilis aurae pulsu) or by a blow of air that has been struck (verberati aeris ictu).14
Commenting on Varro’s reference to the Stoics, Marcia Colish
proposes that this vocabulary of breath and spirit reflects the Stoic
doctrine of pneuma.15 Perhaps, but Varro’s articulation of breath
(spiritus), air (aura), and violent pushing (pulsus) or striking (ictus),
which between them release voice as sound, cleaves closely to the
discussion of voice in On the Soul.
The definition of vox as aer ictus, air that has been struck, is repeated
by later Latin grammarians, including Donatus, Victorinus, Probus
and Priscian.16 They also follow Varro in proceeding next to define
16 Paragraph
voice in function of its division into kinds, a move that has been
much studied in recent years.17 Initially, grammarians distinguish two
types of voice, ‘distinct’ and ‘confused’ (sounds that can be written
down versus those that cannot). Later they admit four, each of the
branches of the twofold distinction being further subdivided according
to whether it is articulate (deliberately meaningful) or not. What
seems to be essentially at stake in this revision from two divisions to
four is the clarity with which the human voice is demarcated from
others. It is not privileged in the bipartite model since the quality
of distinctness is common to any human or non-human vocalization
capable of being segmented into letters (a bird saying ‘cuckoo’ just
like a person saying ‘cooking’) and, equally, some human as well as
most non-human sounds are ‘confused’ or not writable (like grunts
and groans). The fourfold division restores the human as such, since
its category of ‘articulate’ or intended meaning is exclusively human.
A distinct voice that is articulate is human, one that is inarticulate is
not – so ‘cooking’ and ‘cuckoo’ are differentiated. Similarly, a human
voice may be both confused and articulate, but not so a non-human
one – so human grunts or groans can have meaning, others not. The
refinement into four categories marks a distancing from voice’s earlier
government by the animal soul, since now the human voice is located
relative to thought, but not so the earlier one. All these divisions of the
voice assume a soundscape — like the enumeration of animal sounds
in History of Animals — that they then proceed to fracture in different
ways.
Another important feature of the grammatical tradition, since
grammarians illustrate their points with quotations from the auctores,
is the implied connection between voice and literary writing. Voice as
sound, as beaten air and as related in some way to animal voice, is the
starting point for the consideration of literature, most often of poetry.
By the time of Priscian the first translations into Latin of Aristotle’s
works were being produced, beginning with the introductory logical
works that make up the Organon. These provided the next part
of the school curriculum after grammar; most of them, including
De interpretatione, the Organon’s second component, were translated
by Boethius. De interpretatione was widely read as an elementary
textbook. Most late antique and medieval schoolboys would therefore
have known that, as Boethius put it, ‘Those things that are in the
voice are recordings of passions (passionum notae) that are in the soul’
(De interpretatione 16a3).18 Although Boethius faithfully includes
Aristotle’s indication to the reader to consult On the Soul for a fuller
 Circulating Air 17
account of what is meant, the latter treatise was not translated into
Latin until the mid-twelfth century (see below). But a definition of
sound that could well have come from On the Soul is included in the
most influential late antique treatise on music, De institutione musica,
which is not coincidentally also by Boethius. Echoing On the Soul
419b20–2, Boethius states that ‘sound is a striking (percussio) of air that
is not dispersed before being heard’ (Book 1.3).19 Only a few lines
earlier, Boethius alludes to On the Soul when he attributes to Aristotle
a view of the rational soul as conjoined with one that is irrational
(inrationabilis; 1.2; 34).
Boethius’ account of voice in his musical treatise takes a different
form since it focuses on distinguishing the singing voice, which jumps
across pitch intervals, from the relative monotone of speech, with
an intermediate category for chant or declamation (1.12; 50–2). But
the grammarians’ definition of voice appears in another discussion of
music, that of Isidore of Seville who, in his Etymologies, reiterates the by
now familiar view that voice is air that has been struck. We might have
expected this formulation in Isidore’s discussion of grammar (Book 1),
but instead it occurs in Book 3, which is about arithmetic and music:
Voice (vox) is air beaten (verberare) by breath, and from this also words (verbum) are
named. Properly, voice is a human characteristic, or a characteristic of unreasoning
animals. But in some cases, with incorrect usage and improperly, a sound is called
a ‘voice,’ as for example ‘the voice of the trumpet bellowed’, and (Vergil, Aen.
3.556):
. . . and voices broken on the shore.
For the word proper to rocks on the shore is ‘sound’ (sonare). Also (Vergil, Aen.
9.503):
But the trumpet far off (made) a terrible sound with its sonorous brass
(aere canoro). (3.19.2)20
Although Isidore begins by repeating the discredited etymology of
verbum (‘word’) < verberare (‘beat’), the connection of voice to language
is irrelevant to his discussion, which is not about verbal substance
but the singing voice. Such a voice is common to humans and
certain non-human (‘unreasoning’, inrationalis) animals, and can even
be attributed — if erroneously — to water and trumpets. This looks
like an invitation to discuss the two- or fourfold divisions of voice
that customarily follow its definition as beaten air. That Isidore does
not do so is the more remarkable since the first example he proceeds
to quote, from Aeneid Book 3, evokes a variety of kinds of voice.
18 Paragraph
In addition to the ‘voices broken on the shore’, waters emit groans
(gemitum) and shoals leap or rejoice (exsultant); as with the human
voice, these sounds are produced by violent impact, rocks being driven
(pulsata) and voices broken (fractas). His second quotation, the passage
from Book 8, similarly recalls the grammarians’ trope of voice as beaten
air: when the trumpet’s ‘sonorous brass’ resounds, the adjective canorus
(‘melodious’) evokes song while aere can be a form of aer ‘air’ as well
as of aes ‘brass’. Even though Isidore contests the ascription of voice
to these examples, by enumerating them he re-invokes the soundscape
fractured by grammarians and re-designates it as the soundscape of
music.
The existence of all these texts means that, when On the Soul was
translated as De anima by James of Venice in the mid-twelfth century,
a Latin terminology for describing the nature of sound was readily
available. By James’s day it was associated with grammar and music,
whereas formerly grammarians and theorists of music appeared to be
echoing Aristotle. Thus, the discussion of sound in the Latin De anima
2.8 deploys terms such as ictus, percussio and verberare to evoke the
striking of air, for example:

419b13–24: As we have said, not all bodies can by impact (ictus) on one another
produce sound . . . What is required for the production of sound is a striking
(percussionem) of two solids against one another and against the air (aera). The latter
condition is satisfied when the air (aer) struck (percussus) does not retreat before
the blow, i.e. is not dissipated by it. That is why it must be struck (percutiatur) with
a sudden sharp blow, if it is to sound.21
420b14–16: But everything that makes a noise does so by beating (verberante)
something against some other thing in something, namely air (aer), so it follows
that only those things produce voice that take in air.22

It emerges from this that a significant number of medieval treatises

on grammar and music attest the same account of voice as that which
Aristotle’s On the Soul elaborates of sound in general. The features
which these texts share are the definition of voice (or sound) as
physically striking or beating air and the relationships posited between
human voices and those of other animals (or other sounds in general),
the decision to posit the human voice as unique being only one choice
among others. Although the relation of voice to the animal soul is
not explicitly asserted outside On the Soul, it is everywhere implied
that voice forms part of a soundscape that exceeds the human. It is
impossible to know how much grammarians knew or shared Aristotle’s
 Circulating Air 19
view of voice as the expression of the passions of the animal soul,
and as consisting both in the circulation of air and in its rupture,
but their definition of voice as aer ictus might be taken as alluding
to these ideas. Demarcating the boundary between air and soul and
reopening it as breath, the voice is positioned to express desire for,
and movement towards, an object of appetite, the theme — however
refined its elaboration — of many troubadour songs.

The Troubadour Bernart Marti

The troubadours cannot avoid singing by means of inbreathed air;
they also sing a lot about it. Their commonest word for it is aura,
‘air’ or ‘breeze’, but the more learned terms for the element, aire or
aer, are also found, as are various words for ‘wind’. In the opening
stanza of a canso by Giraut de Bornelh, for example, the aura comes
with laughter, jesting, singing and a general uproar into which the
troubadour launches his own song, expecting that it will then be sung
by others.23 Here song forms part of a tumult of predominantly social
sound that circulates with the air; but the other sounds accompanying
the troubadour’s voice are not all human, and human sounds are not
necessarily the most meaningful (as here, they can be simply noise).
Many singers revert to the same tropes in their exordial stanzas. Air
circulates through them and also through the voices of birds and other
creatures, some of which would be writable (distinct), others not, and
most of which express ‘passions of the soul’, usually the desire for a
partner.
In ‘Amar dei’ (PC 63.1)24 the early troubadour Bernart Marti sings
of his compulsion to love, introducing his theme with a traditional
natural setting in which movements of air and the surrounding
birdsong set the mezura, the measure (model, or maybe rhythm?) that
he will follow:

Amar dei —25

que ben es mezura
lanquan vei
lo tems en verdura
5 e l’aura es dousana
e refrinh lo chan pels plais
que l’auzels demena,
e·ill nueitz aserena
e floris la mora.
20 Paragraph
(I am compelled to love — / for measure requires it / when I see / the season in
greenery / and the air is mild / and the song resonates through the hedgerows /
that the bird sings / and the night is serene / and the blackberry is in flower.)
I have chosen to write about this song rather than any of the many
others with similar openings because, although the text is sometimes
opaque, in stanzas 2–4 in particular it presents sound and voice as
predominating over sight or image, even describing the singer as
beating the air, while also talking rather frankly about the effects of
sexual captivation by an external object, the lady. Almost nothing is
known about Bernart Marti, but I have suggested elsewhere that he
may have known a form of the liar’s paradox taught in contemporary
schools; and so this school tradition of describing voice and sound may
have been familiar to him too.26
Stanza 2, although flawed by likely copyist’s errors, appears to
continue the trajectory set in the exordial stanza, introducing the
singer’s song as a response to breeze and birdsong.

10 Dunc dompnei
color en peintura,
mas be vei
en plan ma rancura;
cui sa dona enguana
15 tan no·s pagua ni s’irais
que ja l’en sovenha;
d’elir amor terrena
soven chant’ e plora.

(Then I pay court / to painted colour / but I can see my grievance / plainly
indeed. / The man whose lady deceives him / is not so appeased nor so angry /
that it helps him one bit; / he often sings and sobs / at having chosen this worldly
love.)
The opening four lines contrast the delusions of courtship with the
unvarnished reality of grievance; from verse 14 the singer presents
himself, via the third person, as a man deluded who gets no benefit
from either his satisfaction or his distress. With the breeze and the
bird he sings and sobs his enthralment to love, his singing (chantar, 17)
corresponding to (delusional?) contentment (se paguar, 15) and sobbing
(plorar, 17) to distress (s’irasser, 15). Sight of his lady (11–12) nourishes
only illusions — all he can genuinely ‘see’ are his ‘passions of the soul’
that are expressed in the sounds that dominate the end of the stanza;
 Circulating Air 21
some of these, like sobbing, belong in the grammarians’ category of
‘confused’.
The next stanza embraces the delusion, and closes down the field of
vision in favour of darkness. If a voice is heard, it is that of conciliation.

Ges no·l nei

20 que sa forfaitura
no·ill plaidei,
tot per nueit escura
ab leis ses luguana ;
mas, la·m vailla Dieus, la bais,
25 gardan de mal plena
que·l plait destremena
e d’als non labora.

(I don’t at all deny / that I do not accuse her / of her crime / together with her
/ all through the dark night without a light; / rather, so God help me prevail, I
kiss her, / watching as, full of evil as she is, / she ends the suit / and strives for
nothing else.)
Resolution comes in the dark, in the form of sexual satisfaction. But
in stanza 4, this satisfaction has its limits. Again, the text is challenging,
here because of the omission from the manuscript of verse 32 and the
marginal addition of letters proposing a different reading for verse 33:

L’aer correi,
qu’es com folatura;
30 leis non grei,
si·l veils quers pejura
(. . .)
lassa, que·s fara jamais? / las, as que·s fara jamais?
Tan greu cuj revena,
35 tant ha blava vena,
c’uns veillums langora

(I beat the air, / which is like raving; / let her not care / if old skin gets worse /
(. . .) / Ah, wretched woman, what will ever come of her? (or Alas, with whom
will she ever do it?) / I think it will be so difficult for her to get any better, / she
has such evil blood, / so old age is made to suffer.)
The air recycled as singing or sobbing in stanza 2 is now aer ictus, air
that he is beating or lashing. It sounds out as a voice of pure craziness:
anxiety at old age, jealousy, impotent fury. The blow that is struck
22 Paragraph
signals both the continuity and rupture of circulating air: on one side
the aura doussana, on the other the animal soul. What worked for the
birds in the hedgerows which breathe and sing with the same air does
not bring the lover any joy. Perhaps it is in this failure that the limits of
an animal soul make themselves felt. The singer proceeds in stanzas 5
and 6 to parley himself into greater equilibrium; the song’s final stanzas
represent a more conventional, conciliatory tone. But at no point does
the singer think of his inspiration as enabling an inner vision whereby
he can possess his object in imagination.
As in many troubadour songs, the circulation of air in Bernart
Marti’s ‘Amar dei’ follows a different process from that proposed by
Agamben, one that can be described in terms of a widely read tradition
of texts including grammar, logic, music and natural philosophy.

I have used these works to develop a notion of soundscape which

orients song, and poetic composition more broadly, around the voice.
Understood as a kind of sound (including non-human sound), voice is
made by striking air, and marks both the boundary and the connection
between air and soul, whose passions it expresses. The circulation of
air on which voice depends enables it to continue to inspire poets
and singers long after the troubadours were no more than a half-
remembered echo, their songs long ago exhaled in the songs of others.
Sorrowful in its sweetness, their breath is still just faintly audible in
Shelley’s west wind.

NOTES
1 Quoted from G. K. Blank, ‘Shelley’s Wind of Influence’, Philological Quarterly
64 (1985), 475–91.
2 Francesco Petrarca, Canzoniere, edited by Giancarlo Contini (Turin: Einaudi,
1964), 244: ‘L’aura serena che fra verdi fronde / mormorando a ferir nel volto
viemme, / fammi risovenir quand’Amor diemme / le prime piaghe, sí dolci
profonde; // e ‘l bel viso veder, ch’altri m’ascondi’.
3 At least twenty-four named troubadours and five anonymous ones use this
motif, some repeatedly. See Cesare Segre, ‘I sonnetti dell’aura’ and ‘Isotopie
di Laura’, both reprinted in Notizie dalla crisi: Dove va la critica letteraria?
(Turin: Einaudi, 1993), 42–65 and 66–80; Barbara Spaggiari, ‘Il tema “west-
österlicher” dell’aura’, Studi medievali 26 (1985), 185–291; Luciano Rossi,
‘Per la storia dell’Aura’, Lettere italiane 42 (1990), 553–74; Andy Lemons,
‘The Metamorphoses of Aura in Ovid and Arnaut Daniel’, Digital Philology:
A Journal of Medieval Cultures 2:1 (2013), 35–59.
 Circulating Air 23
4 Giorgio Agamben, Stanzas: Word and Phantasm in Western Culture, translated
by Ronald L. Martinez (Minneapolis: University of Minnesota Press, 1993).
First published as Stanze: La parola et il fantasma nella cultura occidentale (Turin:
Einaudi, 1977).
5 As regards the troubadours in particular, see Gaia Gubbini, ‘L’haleine de
l’esprit’ in Les Cinq Sens au Moyen Age, edited by Eric Palazzo (Paris: Editions
du Cerf, 2016), 745–58; and ‘Soupir, esprit: Bernard de Ventadour, Can
lo boschatges es floritz’, Romanistisches Jahrbuch 65/66 (2015), 86–102. On
Petrarch, see, for example, Gaia Gubbini, ‘Pneuma: la scuola siciliana, la
scuola salernitana e l’eredità in Petrarca’, Romance Philology 68 (2014), 231–47
and essays in Corps et esprit au Moyen Age: Littérature, philosophie, médecine,
edited by Gaia Gubbini (Berlin: De Gruyter, 2017).
6 Folquet de Marselha, ‘Qu’inz el cor, port, domna, vostra faisson’ in Le poesie
di Folchetto di Marsiglia, edited by Paolo Squillacioti (Pisa: Pacini, 1999), 285,
verse 9 (PC 155.8); and cf. the delightful illustration of this song in New York,
Pierpont Morgan Library, MS M. 819 ( = troubadour chansonnier N),
fol. 59r, http://corsair.morganlibrary.org/icaimages/8/m819.059ra.jpg/,
consulted 30 July 2017, 12:45 p.m. EST.
7 Bernard de Ventadour, Chansons d’amour, edited by Mosché Lazar (Paris:
Klincksieck, 1966), 180–3 (PC 70.43) compared with 206–11 (70.25) and
212–14 (70.26).
8 For example by Jörn Gruber, Dialektik des Trobar. Untersuchungen zur
Struktur und Entwicklung des occitanischen und französischen Minnesanges des 12.
Jahrhunderts. Beihefte zur Zeitschrift für romanische Philologie 194 (Tübingen:
Niemeyer, 1983) and Simon Gaunt, Troubadours and Irony (Cambridge:
Cambridge University Press, 1989).
9 In this section, all translations of Aristotle are from The Complete Works of
Aristotle. The Revised Oxford Translation, edited by Jonathan Barnes, 2 vols
(Princeton: Princeton University Press, 1984). De interpretatione is 1:25–38
and On the Soul, 1:641–92. Paraphrased passages are identified by Bekker
number only; when the translation is quoted verbatim, page numbers in this
edition are additionally provided.
10 Daniel Heller-Roazen, The Inner Touch: Archaeology of a Sensation (New York:
Zone Books, 2007), 30.
11 I am grateful to Alison Cornish for showing the importance of violence
in Aristotle’s account of sound: see her ‘Words and Blood. Suicide and the
Sound of the Soul in Inferno 13’ in Susan Boynton, Sarah Kay, Alison Cornish
and Andrew Albin, ‘Sound Matters’, Speculum 91:4 (2016), 998–1039
(1015–26).
12 History of Animals, in The Complete Works of Aristotle, 1:774–993.
13 Aristotle implies that thinking and reasoning are found only in humans, but
are dependent on the senses, differing from them mainly in that thought,
unlike sensation, is capable of error (427b7ff.).
24 Paragraph
14 ‘Vox est, ut Stoicis videtur, spiritus tenuis auditu sensibilis, quantum in ipso
est. Fit autem vel exilis aura pulsu vel verberati aeris ictu.’ Quoted by Marcia
L. Colish, The Stoic Tradition from Antiquity to the Early Middle Ages, 2 vols
(Leiden: Brill, 1985), 1:321. As Colish explains, Varro’s words are known
because they were reprised verbatim by the fourth-century grammarian
Diomedes.
15 Colish, The Stoic Tradition, 1:322.
16 Colish, The Stoic Tradition, 1:326; Christopher Cannon, The Grounds of English
(Oxford: Oxford: University Press, 2004), 116–17; Daniel Heller-Roazen,
‘De voce’ in Du bruit à l’œuvre, edited by Christopher Lucken and Juan Rugoli
(Paris: MētisPresses, 2013), 37–48 (39).
17 Jean-Marie Fritz, Paysages sonores du Moyen Age. Le versant épistémologique
(Paris: H. Champion, 2000), 190–202; Cannon, Grounds of English, 117–20;
Elizabeth Eva Leach, Sung Birds: Music, Nature and Poetry in the Later Middle
Ages (Ithaca: Cornell University Press, 2007), 28–53; Heller-Roazen, ‘De
voce’, 39–43; Sarah Kay, ‘The Soundscape of Troubadour Lyric, or, How
Human is Song?’ in Susan Boynton, Sarah Kay, Alison Cornish and Andrew
Albin, ‘Sound Matters’, Speculum 91:4 (2016), 998–1039 (1002–15).
18 My translation is of the Latin text quoted from the Aristoteles Latinus
Database (Brepols), henceforth ALD: ‘Sunt ergo ea quae sunt in voce earum
quae sunt in anima passionum notae’.
19 Boèce, Traité de musique, edited and translated by Christian Meyer (Turnhout:
Brepols, 2004), 34. Henceforth page numbers to this edition are included in
each reference, following the book and section number.
20 The Etymologies of Isidore of Seville, translated by Stephen A. Barney, W. J.
Lewis, J. A. Beach and Oliver Berghof (Cambridge: Cambridge University
Press, 2006). The Latin original reads: ‘Vox est aer spiritu verberatus,
unde et verba sunt nuncupata. Proprie autem vox hominum est, seu
inrationabilium animantium. Nam in aliis abusive non proprie sonitum
vocem vocari, ut: ‘vox tubae infremuit’ (Virg. Aen. 3,556): ‘Fractasque ad
litore voces’. Nam proprium est ut litorei sonent scopuli, et (Virg. Aen.
9,503): ‘At tuba terribilem sonitum procul aere canoro’. The Latin Library,
http://www.thelatinlibrary.com/isidore.html/, consulted 21 January 2017,
3.40 p.m. EST.
21 Here is the complete passage from the ALD: ‘Sicut autem diximus, not
quorumlibet contingentium ictus sonus est [. . . ] sed oportet firmorum
percussionem fieri adinvicem et ad aera. Hoc autem fit cum permaneat
percussus aer et non solvatur. Unde si velociter et fortiter percutiatur, sonat’.
22 ALD: ‘Sed quoniam omne sonat verberante aliquo et aliquid et in aliquo (hoc
autem est aer), rationabiliter utique vocabunt hec sola quecumque accipiunt
aerem).’
23 ‘Can creis la fresca fueil’ els rams / E l’ombra s’espeiss’ el defes, / M’agrada
l’aur’ e. l temps e. l mes / e. l gaps e. l ris e. l iois e. l chans / e. l douz mazanz / que
 Circulating Air 25
.
creis quant s’aizina l matis. / Si no.m ganda / Mos Seigner covinenz et manz,
/ Fora m’enanz / A far un vers que fos per cels chantatz / Cui iois e pretz e
cortesia platz’ (When the branches break into new leaf and the shade deepens
in the coppice the air and season and month are a delight to me, and a delight
the laughter and jesting and joy and song, and the sweet uproar that grows
with the morning’s approach. If my Lord does not withdraw what she has
told and promised me, it would be good for me to compose a song for those
to sing who take pleasure in joy and renown and courtliness). PC 242.58, vv.
1–11, quoted from The Cansos and Sirventes of the Troubadour Giraut de Borneil:
A Critical Edition, edited by Ruth Verity Sharman (Cambridge: Cambridge
University Press, 1989), 162–7.
24 I have consulted editions and translations by Fabrizio Beggiato, Il Trovatore
Bernart Marti (Modena: Mucchi, 1984), 55–64, and by Simon Gaunt,
Troubadours and Irony, 80–5, but decisions about the presentation and inter-
pretation of the text are my own. The sole witness is Paris, BnF fr. 1749
(chansonnier E), fol. 100, http://gallica.bnf.fr/ark:/12148/btv1b6000801v/
f114.image/, consulted 21 January 2017, 3.45 p.m. EST.
25 Italics show where the text printed here diverges from what is in the
manuscript (for which see Beggiato’s transcription or directly on Gallica).
26 Sarah Kay, Courtly Contradictions: The Emergence of the Literary Object in the
Twelfth Century (Stanford: Stanford University Press, 2003), 134–7.