You are on page 1of 16

.

-m : MNEMOSYNE
\ ral ^r A Journal
'••> 0/
Ciassicai Sluditi
BRILL Mnemosyne 66 (2013) 634-648 lirilUnm/mncm

Did Varro Think that Slaves were Talking Tools?"

Juan P. Lewis
The University of Edinburgh, School of History, Classics and Archaeology,
Doorway 4, Teviot Place, Edinburgh EH89AG, Scotland, UK
Juanplewis@googlemaiLcom

Received: May 2011; accepted: July 2011

Abstract
This article challenges the widespread notion that Varro's Res Rustica 1.17 defined
slaves as 'talking tools'. Instead, it argues that the genus vocale instrument com-
prised both slave and free workers within an economic unit. In this context, instru-
mentum does not mean 'tool', but what is needed to run a farmstead, including the
human work-force, in accordance with the views that were prevalent among the
Romans. Varro drew upon the agricultural literature of his time to build his three-
fold division of the instrumentum. His definition was unoriginal and void of any
moral implications. As such, it has to be seen as a mere rhetorical device. Wbat
Varro attempted to do in this passage was to cast the pedestrian claim that human
labour is needed to run a farmstead into a language that resembled an academic
and philosophical discourse.

Keywords
Varro, Res Rustica, instrumentum, slaves, talking tools

The idea that in Res Rustica 1.17, Varro defined slaves as dehumanized talk-
ing tools has become a trope among most students and scholars of ancient

*' I wish to thank David Greenwood, Alex lmrie, Ulrike Koth and the anonymous reader of
Mnemosyne for their helpful comments and suggestions, and for helping me to make the
English of this article readable. It goes without saying that all remaining errors are entirely
my own.
i<) Kiininklijku lirill NV, l.i-iilcn, 2.113 DOI: l().llB:S/l.'5(j(«2.=il2XH17(i2:i
J.P. Lewis /Mnemosyne 66 (2013) 634-648 635

slavery.'' We can trace its origins at least as far back as Wallon's Histoire de
l'esclavage dans l'antiquité, published in 1847, where he stated that:

[A slave] was an instrument, active and alive, an instrument equipped with


voice and intelligence, by all means a man, even though a degraded one.^'

In a footnote, he added that, according to Varro, the instruments of agricul-


ture were ranked in three kinds, the talking kind, the semi-talking one and
the mute kind.^' A century and a half of modern scholarship has established
the rendering oí instrumentum vocale as 'talking tools' as a sort of scientific
fact. Many renowned scholars have given credit and support to this idea.*'
So seem to have done most modern translators of the text.^' This under-
standing of Varro's passage is found even in Karl Marx's Das Kapital.^^ Only

" All references in the body of the text are from Varro's fies Rustica, unless otherwise
stated.
2' Wallon 1847,189: "[L'esclave] c'était encore un instrument animé et actif, un instrument
doué de parole et d'intelligence, un homme enfin, quoique dégradé" (my emphasis).
3> Wallon 1847,189 n. 2.
*' Barrow 1928,152, Westermann 1955, 81, Lauffer 1961, 385, Staerman 1969, 82, White 1970,
358, Martin 1971, 250, Milani 1972, 203, Hopkins 1978, 123, Wiedemann 1981, 131, Patterson
1982,337, Carandini 1988,12,35,302 and 322, Schumacher 2001,268, Heath 2005,203, McCar-
thy 2004,22.
^' In his Loeb translation. Hooper (1979,225) renders instrumentigenus vocale as "the class
of instruments which is articulate". Heurgon's French edition for Les Belles Letres (1978,42)
translates it as "matériel vocal". In his first critical edition. Flach (1996,187) translates it as
"der sprachfähige Teil der Gutsausstattung" ('the talking part of the equipment'). In the
commentary, however, he renders it as "sprachfähiges Werkzeug" ('talking tool'; 1996,270).
In his new 2006 one-volume edition. Flach slightly changed the German translation to "die
sprachfahige, die halbsprachfähige und die stumme Gattung der Gutsausstattung" ('the
talking, the semi-talking and the mute kind of the equipment'), which is closer to the Latin
original (2006, 215). The German "Gattung" can be translated as 'genus', 'class' or 'kind'.
Tirado Bendi's Spanish translation (1992, 29) renders it as "los instrumentos que hablan"
('the instruments that talk'). Traglia (1979, 635) understands genus vocale instrumenti as the
"vocal means" of cultivation of the fields. The whole passage is rendered as "Altri li [i.e. i
mezzi con cui si coltivano i campi] distinguono in tre tipi: vocale, semivocale e muto" ('Oth-
ers distinguish [the means with which land is cultivated] in three types: vocal, semi-vocal
and mute'). Earlier Italian editions, however, had "istromento vocale" ('talking tool'); see for
example Pagani 1854,50.
••' Marx 1968 [1867], 210 n. 17. Weber (1924, 244) compares the instrumentum vocale, which
he defines as "Sklavenkapital" ('slave capital'), with idle machines that exhaust profits. He
does not quote Varro's passage, but Cato's Agricultura.
636 J.P. Lewis /Mnemosyne 66 (2013) 634-648

a few scholars such as Gerhard Perl and Jens E. Skydsgaard have reacted
against this trend. Both have noticed that with the locution instrumentum
vocale, Varro referred to the entire work-force of a farmstead, indepen-
dently of their legal status.^'
In what follows, I will argue that there are no grounds to think that Varro
thought that slaves were tools, let alone 'talking tools'. The Latin of the pas-
sage in question leaves no room for doubt: in this context, the word instru-
mentum did not mean 'tool'. Moreover, as Perl and Skydsgaard have
suggested, the genus vocale of the Instrumentum did not refer to slaves only,
but to all agricultural labourers, both free and slaves. I will also argue that
there is nothing particularly derogatory or morally degrading about slaves
in this passage, which is a purely factual description of what is needed to
make a farmstead work. Finally, I will suggest that Varro was not innovat-
ing, but drawing upon the agricultural literature of his time. At most, he
merely introduced a new manner of putting forth old ideas.
First, we have to understand what type of book we are dealing with. Res
Rustica is not so much a handbook of husbandry, as a treatise on morals
and an exercise in rhetoric and logical argumentation—and, possibly, a
pedantic display of Varro's encyclopaedic knowledge.**' Varro was very
keen to present himself as a methodical and careful thinker.'" Unlike Cato,
Columella, or Palladius, who organised their books as prescriptive manu-
als, fies Rustica discussed agriculture and husbandry in the form of an aca-
demic dialogue, as if the subject was the matter of a philosophical tract
rather than of a companion to farming."" The subtext of his treatise is a
philosophy of history in which agriculture is not only older than city life,
but also morally superior (2.1.3-4; 3.1.4).'" He reproduced all the prejudices
of absentee landowners, their romantic ideas about rural life and their

7) Perl 1977,424, Skydsgaard 1980,68. Echoed by Carlsen (1995,18).


8' On Varro's reputation for encyclopaedic knowledge, see Lindberg 2007, 137, Cardauns
2001,82-4.
9' Skydsgaard 1968,36.
"" The two main contestants in Book 1 are Cn. Tremelius Scrofa and C. Licinus Stolo, two
contemporaries of Varro who wrote on agricultural matters. There has been some polemic
about the real identity of Scrofa; see Brunt 1972,304-8, Martin 1995,86, Münzer 1937.
'" Conte 1994,219: "Thus, not intended (except superficially) for the practical instruction of
the steward, but written rather to foster and gratily the ideology of the rich land-owner, the
De Re Rustica in a way estheticizes the farmer's life", quoted by Kronenberg (2009,12 n. 46).
J.P. Lewis /Mnemosyne 66 (2013) 634-64S 637

contempt for commerce and city luxuries (3.2.15). The way in which Varro
organized the text and presented his arguments is even more important
than the accuracy of the information provided. The treatise is divided into
three books, each covering a particular aspect of this subject-matter.'^' The
first book is about agriculture proper, the second one about livestock breed-
ing, and the third deals with poultry and hunting. Each book is subdivided
into two or more parts, each one in turn subdivided in dyads, triads and
tetrads and their respective sub-genera. Nonetheless, despite his rhetorical
efforts and his claims to academic knowledge and personal experience on
the subject (1.1.11), Varro seems to have added little to the common under-
standing about farming of his time.'^' Very often, he simply described how
farming was done in different regions without making any judgement
about the effectiveness of those methods. He often made mistakes about
agricultural techniques and misquoted other authors.'"*' Some of his reports
are so extraordinary, that one sometimes wonders how much he relied
on hearsay and popular myths.'^' When he did add something, it was usu-
ally no more than platitudes such as 'the size of the familia should be in
accordance with the local conditions of each^nc/us' (loose paraphrase of
1.18.6-7). As Skydsgaard has noticed, he was not always consistent and the
plan of the discussion set at the beginning of the book is very often "only
partially followed"."''
Leah Kronenberg has recently argued that some of these pitfalls may be
due to the fact that fies Rustica is rather "a parody of academic debates in
the Republic", and that it should not be taken too seriously.'^' Although
Kronenberg argues her case quite persuasively, an author's real intentions

'^' On tbe structure of tbe first book of Res Rustica, see Skydsgaard 1968,10-25.
™ Varro formed part of tbe commission establisbed by Caesar to distribute Campanian
land and enforce bis agrarian laws of 59 BC, Pliny Nat. 7.176. Land distribution, bowever, is
not necessarily related to busbandry and farming. Being in cbarge of a land reform does not
necessarily mean tbat one is well versed in agricultural matters.
"*' Skydsgaard 1968, 64-88 and 118.
'^' For example, be explained tbe bigb reproductive rate of bares quoting Arcbelaus' tbeory
tbat bares grow more birtb canals as tbey get older (3.12.4).
""' Kronenberg 2009,84, Skydsgaard 1968,12. Tbis bas led some scbolars to wonder wbetber
tbe cbaracters of tbe dialogue are only parodies of otber people's opinions. Most of tbem,
bowever, bave rejected tbis conclusion; see Kronenberg 2009, 76 n. 2, wbere sbe quotes
among otbers Flacb 1996, 28, Skydsgaard 1968, 37 and Heurgon 1978, xl.
'" Kronenberg 2009, 76.
638 f.p. Lewis /Mnemosyne 66 (2013) 634-648

are always difficult to pin down. The argument developed here is not
affected by how this question is answered, however. As will be seen towards
the end of this article, the passage in question could have been written
without changes no matter whether Varro wrote fíes Rustica as a serious
piece of scholarship or as a satire of the works of other intellectuals of
his time.
Res Rustica 1.17 belongs to the second part of Scrofa's discourse on the
four chief components of agriculture (1.5.3), which comprises more than
half of the first hook (1.3 to 1.37.3). The theme of this section is the part of
agriculture 'which must be in the farmstead for its cultivation' {pars, quae
[...] in fundo dehent esse culturae causa; 1.5.4, Hooper's translation), in
other words, the instrumentum. Literalism should not confuse us here.
Although instrumentum can mean 'tool' or 'utensil' in Latin, that is not the
only meaning of the word.'^' The Latin verb instruere means 'to furnish' or
'to equip'.'^' In this context, therefore, instrumentum is the nominalization
ofthat verb. Accordingly, it does not mean tool, but all the means a farm or
any other productive unity is furnished with in order to make it work,
including "the human work-force",20) a definition found in Ulpian's twenti-
eth book adSabinum {Dig. 33.7.8.pr.):

In instrumento fundi ea esse, quae fructus quaerendi cogendi conservandi


gratia parata sunt. Sabinus libris ad Vitellium evidenter enumerat: quaerendi,
veluti homines, qui agrum colunt, et qui eos exercent, praepositive sunt his,
quorum in numero sunt villici, et monitores, praeterea boves domiti et pécora
stercorandi causa parata, vasaque utilia culturae
The instrumentum of the farmstead is comprised of everything which is pre-
pared for the sake of procuring, gathering and hoarding the fruits (of agricul-
ture). Sabinus enumerates them in his books to Vitellius: procuring, such as
men who work on the land, and those who oversee or preside upon them.

'^* For instrumentum as 'tool' or 'utensil', see e.g. Cicero Dom. 62; Livy 42.53.4; Quintiiian
¡nst. 2.21.24 and DecL 281.3; Pliny Ep. 10.33.2. Used metaphorically to refer to astrologers and
their trickery, see Tacitus Hist. 1.22.2 [multas secreta Poppaeae mathematicos, pessimum
principalis matrimonii instrumentum, habuerant).
'^' E.g. Cato Agr. 10,12 and 14, quoted by Varro R. 1.18.1; (blumella 9.6.3; Livy 29.4.2 (instruere
naves).
^'" Bradley 1994,58. This was true for a farm and for a taberna as well: Dig. 50.16.185. On the
instrumentumfijndi, see Steinwenter 1942, 24-40.
J.P. Lewis /Mnemosyne 66 (2013) 634-648 639

among whom there are bailiffs and overseers. And also tamed oxen and sheep
used for fertilizing the soil, and vessels and other utensils used for the cultiva-
tion ofthe land

There, men who till the land are mentioned as forming part ofthe instrumen-
tum, not as tools themselves, together with the bailiff, cattle, and agricul-
tural implements such as vessels, ploughs, scythes, wagons, etc.-^" As Scrofa
himself states, the part of agriculture he is about to discuss encompasses
both men, 'through whom the land is cultivated, and the rest ofthe equip-
ment' (de hominihus, per quos colendum, et de rellquo instrumento, 1.5.4).
Loyal to his compartmentalising habits, Varro made Scrofa claim that
some divided the instrumentum into two parts or genera, others into three
parts: the genus vocale instrumenti, the genus semivocale and the genus
mutum (1.17.1), wherefrom the locution instrumentum vocale or 'talking
tool' has been derived by modern scholarship (1.17.1):

Quas res alii dividunt in duas partes, in homines et adminicula hominum, sine
quibus rebus colère non possunt; alii in tres partes, instrumenti genus vocale
et semivocale et mutum
These things some divide in two parts, in men and the accessories of men,
without which they cannot work the land; others in three parts, the talking
component ofthe instrument, the semi-talking one and the mute one^^'

Here is where the core ofthe confusion lies. If we pay close attention to the
text, we can see that whereas instrumentum is in the genitive, the three
genera are in the nominative. The adjective vocale, which is in the nomina-
tive, is therefore a modifier oí genus, not of instrumentum. What Varro was
doing here is what our brains do intuitively when we are encountered

^" Quoting Sabinus, Ulpian used homines qui agrum colunt ('men who work on the land'),
but he did not give any status indication. We do not know whether he meant only slaves or
labourers of both free and servile status.
^^' The fact that Varro used res (things) to denote both agricultural labourers and the rest of
the equipment ofthe farm may have some bearing on the traditional interpretation of this
passage. A modern reader could always understand this use of res as an attempt to dehu-
manize both free and slave labourers. In Latin, however, both animate and inanimate nouns
can be coreferential with either a neuter form of a demonstrative pronoun or the noun res
in its stead, as in Dig. 43.i.i.pr; see Kühner and Stegmann 1912 [1877], 61-2.
640 J.P. Lewis /Mnemosyne 66 (2013) 634-648

with an amorphous cluster of entities, namely we group them together in


collections or subsets of the larger set.^^' All subsets are distinguished by a
certain characteristic or condition which they do not share with other sub-
sets.^'*' Varro's three/yesera, therefore, are subsets of the larger set known
as instrumentum fundi. The characteristic or condition that distinguishes
Varro's ñrst genus from the other two is the ability to use language, hence
the qualification as vocale.
Even \{ vocale modifies .^ewws and not instrumentum, it could of course
be argued that Varro's classification does not necessarily exclude the exis-
tence of an instrumentum vocale, i.e. a 'talking tool'. In Lingua Latina 8.21,
for example, Varro stated that there were two classes of word derivation,
voluntary and natural {dectinationum genera sunt duo, voluntarium et
naturate).^^^ In that passage, the adjectives modify genera. Conversely, in
L- 9.35, we find the same adjectives modifying declinatio: itaque in volun-
tariis declinationibus inconsequentia est, in naturalibus constantia. If that
was the logic behind 1.17.1, it would follow that the three genera of the
instrumentum could be instrumenta themselves. Although this possibility
cannot be entirely ruled out, no passage has been preserved in which
instrumentum is modified by any of the adjectives that modify genus in
1.17.1.^^' Moreover, as far as what we know from what has been preserved of
his writings, Varro never conflated the two meanings oi instrumentum. In
L. 5.134-40, he used the word in the restricted sense of'tool' to denote only
the third genus of the generic instrumentum.^''^ Nothing of the sort is ever

^' Holmes 1998,15.1 am well aware of the risks of anachronism in using modern concepts
to talk about ancient realities. Modem set theory was initially developed by Cantor and
Dedekind in the 1870s and Varro was not explicitly using anything of the sort. Nonetheless,
as the mental operation of clustering entities into sets is almost certainly hardwired into our
brains, this understanding of Varro's text stands.
^*' This is called the axiom ofspecification, which states that "to every set A and to every condi-
tion S(x) there corresponds a set B whose elements are exactly those elements of A for which
S(x) holds", and which can be formalised as follows: B = {x e A : S(x)}; see Halmos 1974,4.
^^' By declinatio (derivation) he meant the formation of new words from other existing
words, such as the male name Ephesius from the name of the city of Ephesus.
2"^' The reason may be that whereas genus is undoubtedly used to denote two different
types of word formation in L 8.21, the three genera of the instrumentum are in fact its parts
or components.
^''* He listed all sorts of fanning tools such as hoes, ploughs, harrows, sickles, etc. under the
rubric of instrumenta rustica.
J.P. Lewis /Mnemosyne 66 (2013) 634-648 641

found with regard to the other two subsets or genera. Whereas there is no
doubt that he saw the elements of the mute subset as 'tools' in the restricted
sense of the word, there is no evidence that Varro ever referred to the ele-
ments of the first two subsets explicitly as such.
Contrary, then, to what many scholars think, Varro's subset of talking
elements is not comprised of slaves only. The confusion springs from the
following clause of 1.17.1, where Varro expands on the components of each
subset of the instrumentum:

genus vocale, in quo sunt servi, semivocale, in quo sunt boves, mutum, in quo
sunt plaustra
the talking part [of the instrumentum], in which slaves are, the semi-talking
one, in which oxen are, and the mute one, in which wefindwagons

As Skydsgaard has correctly noted, although he mentioned slaves only, but


no other type of workers, Varro did not equate the genus vocale of the
instrumentum to slaves here. Instead, Skydsgaard points out, "the words in
quo introduce a typical Varronian exemplification".^^' Here, Varro was nei-
ther defining the subsets nor stating what each of them was comprised of,
but merely mentioning some of the elements of each subset as an example.
Accordingly, of the genus semlvocale,^^^ he mentioned oxen only. Later in
the text, when he elaborated on his definition, he mentioned not only cat-
tle, but donkeys, and sheep and even dogs (1.19-21).^°' Of the^enus mutum,
he listed wagons only. In his elaboration, he also included utensils like bas-
kets {corbulae), iars {dolia), and the like (1.22). If we turn our view back to
the subset of talking elements, we will notice that slaves were not the only
elements of it. Straight after his definition of the three genera, he added
that (1.17.2)

omnes agri coluntur hominibus servis aut liberis aut utrisque


all land is worked by men, slaves or free, or both

28' Skydsgaard 1980,68.


^^' That is to say the subset C whose elements are those elements of the set A for which
the condition T(x), namely to make sounds that are not language, holds.
^°' This was noted by Perl (1977,425), who also noticed that Varro did not mention horses.
642 /. p. Lewis / Mnemosyne 66 (2013) 634-648

In Other words, Varro thought that slaves, as capable of speech, were human
beings like any other free labourer. He never called slaves 'tools'. There was
nothing particularly contentious about this in ancient times. It was widely
believed that the ability to use language was what most distinguished
human beings from animals and non-living things.^" As beings who could
use language, therefore, slaves could easily be seen as forming part of the
human kind.^^* Further confirmation that Varro saw slaves and free as

a" Heatb (2005) bas discussed tbis for Greek autbors. Tbere is no equivalent for Latin ones,
tbougb.
^^' Contrary to wbat is commonly assumed, Aristotle's views on natural slavery were fiercely
contested among tbe Greeks, a fact Aristotle bimself acknowledged in PoL i253b2o-3. He did
not mention wbo tbe opponents to tbe tbeory of natural slavery were. It bas been suggested
tbat among tbem was a certain Alcidamas, a disciple of Gorgias. Alcidamas was known for
bis Messenian Oration, in wbicb be proclaimed tbat 'God created us all free, nature bas
made no-one a slave' (ÈXeuSépouç àffjKS 7ravxc(Ç SEOÇ, oúSéva SoûXov Y¡ çuaiç 7r£7roíy)K£v), quoted
by Aristotle in Rh. i373bi8; see Cambiano 1987. Tbe tbeory of natural slavery bas never been
attested among tbe Romans. From a very early date, tbe communis opinio among tbe Romans
was tbat according to tbe laws of nature all men were free and equal and tbat slavery was tbe
product of force and convention. According to Dionysius of Halicarnassus 4.23.1, wben king
Tullius granted citizensbip to freedmen, be summoned tbe assembly of tbe people and
expressed surprise tbat some patricians tbougbt tbat slaves and free differed in nature
ratber tban in fortune. It is difficult to say bow mucb of tbis anecdote is just Dionysius'
invention, but be would certainly bave found a sympatbetic audience to bis claim. Legal
sources are even more unambiguous. Tbe jurist Florentinus defined slavery as tbe legal con-
dition of being somebody else's property against tbe laws of nature; servitus est constitutio
iuris gentium, qua quis dominio alieno contra naturam subicitur {Dig. 1.5.4.1). In bis book on
Sabinus, tbe jurist Ulpian stated tbat witb regard to natural law, all men were equal; quod ad
ius naturale attinet, omnes homines aequaies sunt {Dig. 50.17.32). We owe tbe standard
Roman definition of social status to (îaius. He stated tbat according to tbe law of persons, all
men were eitber free or slaves (Gaius Irist. 1.8 = Dig. 1.5.3). T^be words be used to talk about
slaves were homo anàpersona (Gaius Inst. 1.48-52 = Dig. i.6.i.pr.). He was not tbe only one to
do so, and it is almost beyond doubt tbat bis was tbe normal way of referring to slaves from
a very early age. In tbe second century BC, tbe jurist Brutus stated tbat tbe offspring of a
female slave did not belong to tbe usufructuary, but to ber master. Tbe reason be summoned
was tbat a buman being, i.e. tbe offspring, could not be regarded as tbe^ucfus of anotber
buman being, i.e. tbe female slave; see Cicero Fin. 1.12; Dig. 7.i.68.pr. Wben a slave sale was
performed, tbe vendor bad to sbow up before tbe magistrate and, grabbing tbe slave, be bad
to declare tbat tbat 'man' belonged to bim by Quiritarian law (Gaius Inst. 1.119, 2.24). Tbe
way in wbicb tbe wbole process was formulated indicates tbat it may bave been ancient, but
still in use in tbe second century AD. We can be confident, tberefore, tbat slaves were recog-
nised as part of tbe buman species tbrougbout Roman bistory.
j.P. Lewis / Mnemosyne 66 (2013) 634-648 643

equal in kind—even if unequal in legal status—is found in the second book


of Res Rustica, where the reproduction of slaves is referred to as fetura
humana, i.e. as human breeding (2.10.6).^^'
So, where do all the tools come from? A common error, already found in
Wallon, is to think that when Varro defined his instrumentum and the sub-
sets thereof, Varro was following Aristotle, who saw slaves as opyava £|.t4'uxa,
that is to say, as animate tools {EN ii6ib4) and xxy||^axa è'fxvjjuxa, or animate
possessions {PoL i253b32).^*> There is no evidence to support that view,
however. There is no doubt that Varro knew Aristotle's work reasonably
well. In fíes Rustica, he mentioned Aristotle by name at least twice (2.1.3;
2.5.13), and quoted him indirectly possibly around thirty times.^^' In this
particular case, however, Varro and Aristotle's views were clearly different.
On the one hand, as pointed out above, instrumentum is not the Latin
equivalent for the Greek opyavov in this context. On the other, as Perl has
noticed, Aristotle's opyava £|xi|)uxa encompassed both men and animals
together, whereas according to Varro, they belonged to different catego-
ries.36' Instead, Perl adds, Varro was merely applying the threefold distinc-
tion between voiced, semi-voiced and voiceless used in other areas of
ancient scholarship such as grammar to the world of farming.^^' Thus, he
was capable of elevating the discussion of rural matters to the same level as
other fields of knowledge.
Whom Varro had in mind in this passage were two Latin authorities on
farming matters that he not only quoted extensively, but also named (1.18).
One of them is Saserna, an agronomist and member of the gens Hostilia
from the second century BC whose treatise on agriculture has not survived.
The other is Gato, who wrote the oldest piece of Latin prose that has come
down to us. Agricultura. If we are to believe Plutarch, Cato's stinginess and

^^' It is true that Varro discussed the purchase and reproduction of slave shepherds and
animals together. He used the concept/eiura for livestock as well (both livestock and human
beings are mammals after all), but he was careful to distinguish the offspring of slaves as
human.
^*' E.g. Steinwenter 1942,27.
^^' Gentilli (1903) has collected Varro's sources.
^^' Perl 1977, 425. He does not give references, see e.g. Aristotle De anima 4iia20 (for ani-
mals) and PoL i253b29 (for men).
^'" Perl 1977, 425-26. On the use of vocale, semivocale and mutum in other areas of knowl-
edge, namely music, astrology and warfare, see Hübner 1984.
644 JP- Lewis/Mnemosyne 66(2013)634-648

brutality towards his slaves were proverbial [CatMa. 5.1). Nonetheless, he


never called his slaves tools. When he discussed how a 240-iugum (i.e. 115-
acre) olive farm had to be 'furnished' {instruere oportet),firsthe gave a list
of the workers needed, who added up to thirteen men {summa homines
XIII, Agr. io.i):38'

Quo modo oletum agri iugera CCXL instruere oporteat. vilicum, vilicam, oper-
arios quinqué, bubulcos III, asinarium I, subulcum I, opilionem I: summa h.
XIII. boves trinos, asinos ornatos clitellarios, qui stercus vectent, tris, asinum
<molarium> I, oves C. vasa olearia instructa iuga quinqué...
How a 115-acre olive farm should be furnished; one bailiff, his companion,^^'
five workers, three cattle herders, one donkey groom, one swine herder, one
shepherd: total of men thirteen. Three oxen each, three saddled donkeys, used
for dung, one mill-donkey, a hundred sheep. Ten oil vessels for each fanned
acre...

It is not clear whether all the people listed here were free or slaves. They
were probably mostly slaves, but they could be a mixture of both, as Varro's
genus vocale was.**"' Immediately, Cato listed the number of beasts of bur-
den and cattle needed, and after that farming utensils, vessels, and other
inanimate items ofthe equipment.*" Cato's style is much coarser than Var-
ro's. Unlike the latter, he did not present his topic as a philosophical dia-
logue, but as a list of apodictic instructions that had to be followed. He did
not mention the^e«era ofthe instrumentum. He merely provided random

38' A 57-acre vineyard was to be worked with fourteen men {Agr. 13.1).
3^' Traditionally, it has been believed that the viiica was the companion ofthe viticus (bai-
liff). Roth (2004) has challenged this view and has argued that the viiica was a job title in
itself with specific responsibilities and that not all vilicae were necessarily married to a bai-
liff. For the purpose of this article, however, the role ofthe viiica and her relationship to the
vilicus are irrelevant.
40) ji^g yiiiçug and the viiica were usually but not always of servile status. Operarii were usu-
ally poor free peasants. Other workers such as the bubulci (cattle hearders), the asinarius
(donkey groom), the subulcus (swine hearder) and the opilio (shepherd) could be of either
status. For differing views on the status of some agricultural labourers, see Scheidel 1990 and
Schumacher 2010.
•*" Ulpian's definition of what constitutes the instrumentum resembles this one as well (see
above).
J.P. Lewis /Mnemosyne 66 (2013) 634-648 645

lists of some of the elements of each subset, but the three subsets of the
instrumentum are easy to identify.'*2'
Skydsgaard has also pointed out that there is nothing particularly degrad-
ing or dehumanizing about slaves in this passage, and "indignation
expressed by scholars commenting on [it] is out of place".'*^' If anything,
Varro seems to have been more hostile towards free agricultural workers
than to slaves. He reminded his audience that it was convenient to allocate
the hardest tasks to free workers. He also added that it was more profitable
to make free workers cultivate unwholesome land, and by unwholesome
he possibly meant lands infested with malaria (1.17.3).'^' This favouritism
towards slaves is unsurprising, however. Slaves had to be bought and fed.
They were probably more expensive than free workers, who did not have to
be provided for and were hired only temporarily to perform specific
duties."*^' He may have done so by expedience and self-interest rather than
by humanitarianism, but Varro certainly held his slaves dearly.
It could always be argued that, even though Varro did not regard slaves
as mere objects with the capacity to speak, nonetheless the whole passage
still portrays a very dehumanizing picture of slaves. After all, Varro was
lumping slaves with cattle and other dumb utensils. Ulrike Roth, for exam-
ple, has ventured that "the only distinction made by Varro between man and
animal is that symbolized by the labels attached to either", and that what
they have in common is that "they are the means by which the soil is
cultivated".'*^' As pointed out above, the capacity to speak is an essential
characteristic that distinguishes human beings from animals. It is not a
simple label, but a significant difference. It does not carry any necessary
moral implications, but the a priori assumptions of the modern scholar.
The fact that both Varro and Cato saw slaves as clearly distinct from ani-
mals can always be construed as a proof of their humanitarian views about
slavery, but there will not be any reason to support this conclusion either.

*^' Notice that Cato used the verb instruere without nominalizing it as instrumentum.
*^^ Skydsgaard 1980,68.
'"*' Sallares 2002, 61. On the spread of malaria across the Mediterranean in Antiquity, see
Sallares, Bouwman and Änderung 2004.
*^' On the price of slave labour compared to free workers in the late Republic and the
Empire, see Scheidel 2005.
*^' Roth 2005,312, my emphasis; cf. Tilly 1973, 56.
646 J.P. Lewis /Mnemosyne 66 (2013) 634-648

On the other hand, that Varro lists slaves and free workers together with
bellowing animals and dumb wagons is not surprising at all. Running a
farmstead requires farming tools, the animals that are to be farmed, and
the necessary labour force. This is just an uncontroversial statement of fact.
Similar distinctions could be applied to the administration of any unit
incorporating animals.*^* As shown above, Varro was not even being origi-
nal, but just doing what Cato had done before him in a less elegant manner.
As should now be clear, there are no conceptual differences between Cato
and Varro, only stylistic ones. Nobody, however, has ever used the former
to argue that the Romans saw slaves as tools.
To recapitulate, Varro did not think that slaves were tools, let alone 'talk-
ing tools'. Rather than a value judgement full of moral connotations. Res
Rustica 1.17 draws upon the work of other ancient agronomists and recasts
them using new language that has all the trappings of scholarly discourse.
What Varro did with his definitions of the three,genera of the instrumen-
tum was merely to wrap up Cato's random lists of what was needed to fur-
nish afiindus in a language that sounded academic. Thus, he attempted to
make a pedestrian claim—namely, that farming uses human labour, beasts
of burden and other type of hardware—sound like a philosophical state-
ment, either because he was taking himself too seriously or because he was
just having a laugh.

Bibliography

Barrow, R.H. 1928. Slavery in the Roman Empire (London)


Bradley, K. 1994. Slavery and Society at Rome (Cambridge)
Brunt, P.A. 1972. Cn. TremeUius Scrofa the Agronomist, CQ 22.3,304-8
Cambiano, G. 1987. Aristotle and the Anonymous Opponents of Slavery, in: Finley, M.I. (ed.)
Classical Slavery (London), 28-52
Carandini, A. 1988. Schiavi in Italia. Gii strumentipensanti dei romanifra tarda Repubblica e
medio impero (Rome)
Cardauns, B. 2001. Marcus Terentius Varro:Einführung in sein Werk (Heidelberg)

*'•' Discussing the administration of a modern laboratory could be just one example. The
scientific staff needed would be the genus vocale. Mice and other animals on which experi-
ments are done would form part of the genus semivocaie. The rest of the laboratory equip-
ment would be Ûie genus mutum. Accountants would lump them all together in the annual
balance sheet.
J.P. Lewis /Mnemosyne 66 (2013) 634-648 647

Carlsen, j . 1995. Vilici and Roman Estate Managers until AD 284 (Rome)
Conte, G.B. 1994. Latin Literature:A History (Baltimore)
Flacb, D. 1996. Marcus Terentius Varro. Gespräche über die Landwirtschaß, Buch I (Darmstadt)
2006. Marcus Terentius Varro. Über die Landwirtschaft (Darmstadt)
Gentilli, G. 1903. De Varronis in libris rerum rusticarum auctoribus, Studi Italiani di filologia
classica 9, 99-163
Halmos, P. 1974. Naive Set Theory (New York)
Heatb, J. 2005. The Talking Greeks: Speech, Animals, and the Other in Homer, Aeschylus, and
Plato (Cambridge)
Heurgon, j . 1978. Varron. Économie rurale, livre I {Paris)
Holmes, M.R. 1998. Elementary Set Theory with a UniversaiSet (Louvain-la-Neuve)
Hooper, W.D. 1979. Marcus Terentius Varro. On Agriculture (Cambridge, MA)
Hopkins, K. 1978. Conquerors and Slaves (Cambridge)
Hübner, H. 1984. Varros instrumentum vocale im Kontext der antiken Fachwissenschaften
(Wiesbaden)
Kronenberg, L. 2009. Allegories of Farming ftom Greece and Rome: Philosophical Satire in
Xenophon, Varro and Virgil (Cambridge)
Kübner, R., Stegmann, C. 1912 ('1877). Ausführliche Grammatik der lateinischen Sprache, 2;
Satzlehre, 1 (Hannover)
Lauffer, S. 1961. Die Sklaverei in der griechisch-römischen Welt, Gymnasium 68,370-95
Lindberg, D.C. 2007. The Beginnings of Western Science: The European Scientific Tradition in
Phiiosophical, Religious, and Institutional Context—Prehistory to A.D. 1450 (Cbicago)
Martin, R. 1971. Recherches sur les agronomes iatins et leurs conceptions économiques et
sociales (Paris)
1995- Ars an quid aliud?La conception varroniennc de l'agriculture, REL 73,80-91
Marx, K. 1968 ('1867). Das Kapital, I (Berlin)
McCartby, K. 2004. Slaves, Masters, and the Art ofAuthority in Plautine Comeífy (Princeton, NJ)
Milani, P.A. 1972. La schiavitù net pensiero politico: dai Greci ai basso Medio Evo (Milan)
Münzer, F. 1937. Tremeiius, RE VI A.2,2286-9
Pagani, G. 1854. Deli'agricoltura iibri tre dl Marco Terenzio Varone (Milan)
Patterson, 0.1982. Stavery and Sociai Death (Cambridge, MA)
Perl, G. 1977. Zu Varros instrumentum vocale, Klio 59.2,423-9
Rotb, U. 2004. Inscribed Meaning: The viiica and the Villa Economy, PBSR 72,101-24
2005. No More Slave Gangs: Varro, De re rustica 1.2.2, CQ 55.1, 310-5
Sallares, R. 2002. Malaria and Rome: A History of Malaria in Ancient Itaty (Oxford)
Sallares, R., Bouwman, A., Änderung, C. 2004. The Spread ofMaiaria to Southern Europe in
Antiquity: New Approaches to Otd Problems, MedHist 48.3, 311-28
Scbeidel, W. 1990. Free-Bom and Manumitted Baiiijfs in the Graeco-Roman World, CQ 40.2,
591-3
2005. Real Stave Prices and the Relative Cost of Slave Labor in the Greco-Roman World,
Ancient Society 35,1-17
Scbumacber, L. 2001. Sklaverei in der Antike. Alltag und Schicksal der Unfteien (Municb)
648 f.p. Lewis /Mnemosyne 66 (2013) 634-648

2010. On the Status of Private actores, dispensatores, andvilici, in: Roth, U. (ed.) By
the Sweat of your Brow: Roman Slavery in its Socio-economic Setting (London), 31-47
Skydsgaard, j.E. 1968. Varro the Scholar: Studies in the First Book of Varro's De re rustica
(Copenhagen)
1980. Non-slave Labour in Rural Italy during the Late Republic, in: Garnsey, P. (ed.)
Non-slave Labour in the Greco-Roman World (Cambridge), 65-71
Staerman, E. 1969. Die Blütezeit der Sklavenwirtschaft in derrömischen Republik (Wiesbaden)
Steinwenter, A. 1942. Fundus cum instrumento. Eine agrar- und rechtsgeschichtliche Studie
(Vienna/Leipzig)
Tilly, B. 1973. Varro the Farmer: A Selectionfrom the Res Rustica (London)
Tirado Bendi, D. 1992. Marco Terencio Varrán. De las cosas del campo (Mexico)
Traglia, A. 1979. Varrone M. Terenzio. Opere (Turin)
Wallon, H. 1847. Histoire de l'esclavage dans l'antiquité {Vans)
Weber, M. 1924. Agrarverhättnisse im Altertum, in: Weber, M. (ed.) Gesammelte Aufsätze zur
Sozial- und Wirtschaftsgeschichte (Tübingen), 1-288
Westermann, W.L. 1955. The Stave Systems of Greek and Roman Antiquity (Philadelphia)
White, K.D. 1970. Roman Farming (London)
Wiedemann, T. 1981. Greek and Roman Slavery (London)
Copyright of Mnemosyne is the property of Brill Academic Publishers and its content may
not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.

You might also like