19771
COMMENTS – COMMENTAIRES
One Thousand Years of Arra
Introduction
Roman arra has been, and still is, susceptible of great controversy.
Some jurists insist that the arra of Roman law was a continuously
uniform legal concept. Intransigent, they cannot envision an imported
Greek arra, capable of being totally romanized at one period and
yet at another succumbing to the potent customary forces of the
East. No jurist of modern-day civil law has any difficulty in appre-
ciating that a single legal principle may mean something quite dif-
ferent in one century than in another. Nor should any jurist of the
Roman law. Although no startling and innovative theory will assault
the reader of this paper, an attempt will be made to provide the
reader with an historical perspective of Roman arra.
What could be more acceptable, although manifestly less exciting,
than the consideration of one of the great conundra of Roman legal
science as being without solution! Indeed, any definitive solution
would be a travesty of Roman arra. However, among the possible
interpretations, some appear easier to accept than others.
The arra of Roman law differs in no great way from the earnest
of the law of sale in modem civil law systems. For most of the one-
thousand-year history of Roman law, arra was a sum of money ad-
vanced by the purchaser to the vendor in order to both ensure the
fulfilment of the sale and secure the payment of the balance of price.
Both parties, therefore, had a considerable interest in the giving of
arra. Yet, the modalities of its operation varied with the particular
rules governing the law of sale in each of the Roman law’s formative
periods. Whereas some civil law systems restrict the giving of
earnest to promises of sale, the Roman law, in its maturity, allowed
arra to operate both prior to and at the moment of a binding con-
tract.
The early period
The early period in the development of arra occurs before Plautus,
for it is Plautus who indicates a metamorphosis in the nature of
Roman arra.1 Since the advent of consensualism is generally consider-
ed to be the dividing point between the early law and the classical
1 Titus Maccius Plautus was a Roman comic dramatist who lived from 254
to 184 B.C.
McGILL LAW JOURNAL
[Vol. 23
law, the early period is thus characterized by the absence of con-
sensual contracts. This period also witnessed both the development
of the autochthonous forces of contract and the acceptance of the
arra of Greek sale.
Unrefined systems of contract have difficulty in separating con-
tract as a personal obligation from contract as a physical bond. The
human body was looked upon as the general pledge of creditors:
A son point d’irruption l’engagement contractuel d’une chose se confond
pour ainsi dire avec l’engagement contractuel de la personne du d6biteurn2
Consequently, the non-payment of a debt led to dire results under
the law of the XII Tables; 3 a creditor could call for the sale of his
defaulting debtor. To ensure that the contractual duty was fulfilled,.
there was the giving of arra. The human hostages of primitive law
later became small objects intimately connected with the debtor’s
body: clothes, arms, gloves, etc. If the debtor defaulted, he exposed
himself to bad luck, the belief being that there was, between a man
and his possessions, a certain and ill-defined magical link. Elsewhere
in the Roman law, the giving of a ring in promise of marriage, the
arra sponsalicia, was indicative of this very personal pledge.
There was nothing unusual in this sort of arrangement in very
early law. Indeed, such arra-giving was common to the formative
stages of all legal systems. For example, the Semites and the Baby-
lonians had also developed a similar concept of arra. However, the
most direct of all influences on the young Roman legal system came
from Greece.
The principal distinguishing feature of the Greek law was that,
throughout its history, consensual contracts did not exist. The trans-
fer of ownership was dependent upon payment and not, as in the
Roman law, upon delivery. However, prior to the payment of the
price, the buyer was protected, to a certain extent, by registration,
cooperation of neighbours, public announcements and public docu-
ments.4 Although the Greeks lacked a binding contract of sale creat-
ing an obligation to convey the thing sold and thereby pass title, an
6 preceded each cash sale.7 This trans-
6vf
– 7r.paai; 5 or simply cbv
2 Cornil, Ancien Droit Romain (1930), 96.
3 The XII Tables, the earliest Roman code of laws, were drawn up in the
middle of the fifth century B.C. and were described by the historian Livy
as the foundation of all public and private law. The XII Tables codified the
city’s customary law.
4 Pringsheim, The Greek Law of Sale (1950), 91.
5 O-prasis: sale-purchase agreement.
6 nM: sale.
7 See supra, note 4.
1977]
COMMENTS – COMMENTAIRES
action would be accompanied by the giving of arra. Since there was
nothing substantial, save a non-binding promise, the arra could not
be evidence of a pre-existing contract. Under a law which provided
no actions on mere promises, the buyer was penalized by the loss
of the arra if he failed to pay the price. The defaulting seller restored
the buyer’s arra and paid him as much in addition. This was the
-r~3v &ppa6&bvwov,8 described as the “real sanction of sale in
v6poi
the Greek East throughout the Hellenistic period”.9
It is not known how Greek arra was received by Roman law, only
that it was current practice in the time of Plautus, probably imported
into Rome in understandable recognition of an already sophisticated
.legal system. Even before the advent of classical law and its doctrine
of consensual contracts, some liability must have seemed desirable.
Not unlike other civilizations, certain aspects of Roman life were
best revealed (and ridiculed) by their playwrights. Plautus, in parti-
cular, through the medium of his somewhat bawdy plays, has been
regarded as accurately representing the language and mores of his
times.’ Although most of Plautus’s plays can be traced back to Greek
comedies on similar themes, the dramatist adapted his Greek models
to Roman taste. The extent to which Plautus was affected by his
Greek originals is difficult to determine. Unlike Terence,” who
strived to preserve the atmosphere of Hellenistic Athens and con-
ceded little to popular taste, Plautus romanized his originals to suit
his comic purpose. Was Plautus, therefore, presenting Greek arra
to his audience and counting on their familiarity with this aspect of
the Greek law of sale or was he faithfully portraying an already
romanized adoption of the Greek law?
Academic opinion is varied. Pringsheim is not much concerned
“[w]hether it was the original Greek arra of Greek New Comedy or
an imported, but not yet adopted, Hellenistic arra drawn from the
Roman law of his time”. 2 Nevertheless, he says, the Plautine arra
“would in substance be the pure Greek arra”‘u and continues:
SNomos ton arrabonon: the law of arra.
9 De Zulueta, The Roman Law of Sale (1945), 22.
‘0 Duckworth, The Nature of Roman Comedy (1952), 14346, 33140, 384-85.
11 Publius Terentius Afer lived from c.190 to c.159 B.C.
12 Supra, note 4, 429. Greek New Comedy, some say, was the last creative
literature of Athens. It belongs to the last quarter of the fourth century and
the third century B.C. Its principal representatives are Philemon, Diphilus
and Menander. The models for Plautus’s plays came from this literary period.
Is Ibid.
McGILL LAW JOURNAL
[Vol. 23
In this way the old dispute as to whether Plautian comedy should be
used as an illustration of Roman or of Greek law can perhaps be disposed
of.14
However, Pringsheim has already suggested that the legislation of
the emperor Justinian allowed the Greek concept of arra to “again
prevail”’15 which leaves his suggestion that much the weaker for lack
of precision. But Plautine sales, almost invariably, deal with arra
intricately; consequently, it seems unlikely that a Greek-type arra
was not already Roman law because constant repetition of a foreign
institution would have little, if any, comic effect.””
Watson 17 denies that the arra of Plautus’s plays is necessarily
always Greek arra, and insists that it is, at times, not of a legally
penal nature, being both Roman and in harmony with consensualism.
He points to the plays of Mostellaria and Rudens as containing exam-
ples of non-Greek, non-penal arra. Since Watson maintains that at
the time of Plautus consensual contracts were actionable, arra would
have no particular legal significance.17a The recourse open to the
aggrieved party would be an action on the contract itself, not an
action for return of the arra. However, Watson must account for the
multiplicity of references to arra:
The stress in them on arra would, of course, derive from Greek law, but
it is equally understandable on a purely social plane and may have been
seen by Plautus in that light. In the mind of the man in the street, a
consensual contract is that much more binding where arra has been given
even where he is aware that legally it makes no difference to the con-
–
tract.’8
14 Ibid.
15 Ibid., 416.
‘1 In the Curculio, line 618 et seq., Plautus has the boastful soldier claim
fourfold and not double damages for a free girl whom he claims to have
bought by mancipatio, thinking she was a slave. Such comic exaggeration
would only be of immediate relevance if penal arra were current practice in
Rome. Watson, infra, note 17, 46-47, seems to think that the Roman audiences
were capable of understanding Greek law.
17 Watson, The Law of Obligations in the Later Roman Republic (1965), 46-57.
17a See Watson, The Origins of Consensual Sale: A Hypothesis (1964)
32
Tijdschrift voor Rechtsgeschiedenis 245, 253-54 where he suggests the middle
of the third century for the advent of consensual contracts between cives,
citizens, there being no reason to ascribe their introduction to the peregrine
praetor alone. Crook, Law and Life of Rome (1967), 220, counters Watson
and has strong doubts about the date of consensualism. He conjectures that
Plautus is full of arra because his plays were written before consensual sale
had become fully recognized. Likewise Cornil, supra, note 2, 97, believes that
because in Plautus reciprocal promises of sale and purchase were not binding,
the giving of arra took the place of the later vinculum iuris and acted as a
pledge of the buyer executing his promise.
18 Supra, note 17, 51.
19771
COMMENTS – COMMENTAIRES
The “social usefulness” theory of Watson19 creates a demi-monde, an
alternative justice unrecognized by the law. Such a theory is the
unhappy consequence of reconciling consensualism with the texts.
For example, it is no doubt quite possible, although not convincing,
to consider the following arra as of no inherent legal validity, but
merely of social value:
nunc si me adulescens Plesidippus viderit quo ab arrabonem pro Palaestra
acceperam, Jam is exhibebit hic mihi negotium. (Rudens, 1, 554-56).
now if that young lad Plesidippus catches sight of me –
for he’s the one
I accepted arra from for Palaestra –
then he’ll hold me to the bargain,
he will.
Certainly, this text would indicate consequences more penal than
social.
The Greek law of sale was not ambiguous; it obligated the seller,
on receipt of the arra, to pay a penalty in the event of non-delivery of
the object sold. As such, judgment could be given for double the
amount. Thus arra was legally enforceable. In the Mostellaria, Theo-
propides hopes to have judgment against the seller of the house in
question and says:
tanto apud iudicem hunc argenti condemnabo facilius (1.1099).
so much the easier will I be able to obtain judgment for the money in
court.
Such a statement is susceptible of two interpretations: the seller
might be condemned to pay a penalty (double the arra) or the judge
would have him hand over the house to the buyer. The mention of
argenti -leads most authors2 to believe the anticipated judgment to
be financially penal. Watson says this is a judgment in a Greek case
of arra.2 ‘ Other judgments produce similar problems.’ Watson
19 Ibid., 49, where Watson says: “But the social value of arra must not be
ignored. Even if it had no legal effect whatsoever, sellers would still want
to extract substantial arra from the buyer.”
20 MacCormack, A Note on Arra in Plautus (1971) 6 Iriur.(n.s.) 360, 363, n.13.
21 Supra, note 17, 56.
22 In the Rudens, Labrax announces the judgment which Plesidippus has
obtained against him at lines 1281-83:
“Quis me est mortalls miserior qui vivat alter hodie,
quem ad recuperatores modo damnavit Plesidippus?
abiudicata a me modo est Palaestra.”
Name me one man who has had a more miserable day than I? To begin
with, Plesidippus has got a verdict in court against me and Palaestra has
been snatched right out from under me.
Pringsheim, supra, note 4, 423, believes that Labrax was not condemned
to perform a contract but to pay back the arra and being unable to do so,
handed over Palaestra to the plaintiff (contra, Watson, supra, note 17, 52-53).
However, MacCormack, supra, note 20, 364, thinks it strange that if the judg-
McGILL LAW JOURNAL
[Vol. 23
argues22a that (1) arra was not legally sanctioned but socially current,
(2) adjudications were for the object sold and never for monetary
compensation, and (3)
in cases where arra forms part of a sales
transaction, the arra was Greek arra of Greek law, easily comprehend-
ed by the audience. 3 Those texts which clearly provide a choice
between producing the object sold or returning the arra given24 are
not, Watson says, examples of the classical Roman law of consensual
contracts.
Perhaps there is an easier solution: the arra of Greek law was
adopted in Rome and legally sanctioned in the pre-consensual or
transitional period, a two-sided liability being therefore enforced.
A judicial ruling, in the classical law, would not condemn the vendor
to forfeit the arra, let alone double its amount. However, the Greek
law would allow such a condemnation since, as has been mentioned
above, the contract would not be complete until full payment of the
price. To dichotomize Plautus’s plays into those where arra is Greek
and those where arra is Roman is a delicate operation indeed.
The classical law
After consensual sale became fully actionable, there is no doubt
the arra was suppletive, a security device additional to the action on
the sale. The nature of arra changed, incurring liability only on the
part of the giver. In the event of non-performance, the arra was
either lost by the giver or returned by the receiver, according to the
nature of the default. Before Gaius espoused the classical doctrine,
Varro, the first century B.C. scholar, defined arra in his treatise
De lingua latina:
Arrabo sic data ut reliquum reddatur; reliquum, quod ex eo, quod debitum,
reliquum (5.175).
Thus, arra is given so that the balance might be paid; the balance is
that which remains from the amount owing.
and
ment had been for money, the amount should not have been stated –
exploited – by Plautus.
22a Supra, note 17.
2 The following comment of Duckworth, supra, note 10, 81-82, may help
temper Watson’s high regard for the Roman audience when he says: “The
prologues of Terence, as we have already seen, give us an interesting insight
into the nature of the audience which viewed Roman comedy in the second
century B.C. The playwright’s constant plea for silence and a fair hearing
implies that the spectators were a noisy and unruly lot, and if a play had
less appeal than some nearby attraction, such as a boxing match or a
gladiatorial combat, they might rush from the theatre, as in fact happened
in the case of the first two presentations of the Hecyra.”
2 4 Curculio, line 612; Pseudolus, line 1183.
1977]
COMMENTS – COMMENTAIRES
The text is not informative. Furthermore, there seems to have been
an association of arra with pignus, pledge, so that in Apuleius,2 gold
and silver objects were given to secure a loan, to be forfeited if the
money were not paid on time.20 Nothing much more is known of
arra before Gaius:
This first period ended with the recognition of the actionability of the
consensual contract. In the next period arra could still be used, but had to
be adapted to the new legal situation ….
If [the references to the pignus-arra in the early classical law] show a
one-sided liability that does not necessarily mean that arra, when imported,
was at once confined to the task of securing the vendor’s claim to the
price. It is more probable that the two-sided arra was at first used as
such by the Romans and only later restricted, when Roman sale created
full liability. The purchaser’s liability could still be increased by the
addition of a pledge-like arra.
The role of arra in the classical period, at least in theory, is ex-
pressed in Gaius:
Emptio venditio contrahitur cum de pretio convenerit, quamvis nondum
pretium numeratum sit, ac ne arra quidem data fuerit. nam quod arrae
nomine datur, argumentum est emptionis et venditionis contractae (3.139).
A contract of sale is concluded when the price has been agreed, although
it have not yet been paid and even no earnest have been given. For what
is given by way of earnest is evidence of a contract of sale having been
concluded.28
This arra confirmatoria was evidence 9 of a binding contract, a mode
of proof by which it was recognized that the sale had been concluded
by the mere consent of the parties3 Gaius may have been warning
against adopting the view prevalent especially in the East, that arra
was penal in nature 8 ‘ His rule lasted, despite vulgarization, in the
Western Roman Empire until the fifth century A.D. and was applied
by Roman jurists and emperors in Hellenistic arra cases.
25 Apuleius was a second century A.D. Roman satirist-novelist.
26 Supra, note 4, 417.
27 Ibid., 418.
28Text and translation of de Zulueta, The Institutes of Gaius, Part 1 (1946).
1P See Didier-Pailhd, Cours Eldmentaire de Droit Romain 4th ed. (1895), vo12,
for a possibility of a contrary stipulation.
3 0 Cuq, Les Institutions Juridiques des Romains (1902), vol.2, 403, gathers
from Varro and Isidorus that arra was also a guarantee that the balance of
the price would be paid.
31 Crook, supra, note 17a, 221, adds that Roman Egypt had a law of arrhabo
which was penal. Also of the same opinion is Taubenschlag, The Law of Greco-
Roman Egypt in the Light of the Papyri 332 B.C. –
640 A.D. (1955), 407-409.
3 Supra, note 4, 344.
McGILL LAW JOURNAL
[Vol. 23
Hellenistic and Byzantine arra
The Hellenistic peoples, though subjugated by the Romans, did
not cease giving arra of a significant amount to secure the bargain
before a written contract: 3
The arra system, coming from the orient, seems as characteristic of the
Greeks as the system of consensual contracts of the Romans.3 4
The transactions of arra and sale are distinct: a Ptolemaic draft
letter of the third century B.C. speaks first of arra and then of
V 16V Xv Kzrpzrcov TrpitV, “the purchase of the goods.”‘ m A letter of
the same period speaks of an anticipated sale of poppy: 8o6; -r6v
ppa6ova… o6K TI Aiy6paKa; “although you have given arra, you
have not yet purchased”20 Pringsheim has an impressive list of arra
transactions spanning the whole period of Hellenistic influence until
the Byzantine era. They demonstrate: (1) that an arra transaction
was separated from the sale, (2) that arra regularly represented a
part of the price, and that ownership passed on the payment of the
full price, (3) that arra consisted of money and was to be set off
against the price, (4) that arra could not simply be returned by the
receiver without penalty, (5)
that the giving and taking of arra
created liabilitiesssa (6) that arra must be given immediately, and
(7) that arra was a first step to and security for a future sale.
Justinian in 528 A.D., in Codex 4.21.17.37 was to allow the parties
33 Ibid., 344 et seq. and passim. See also Comil in M9langes P.F. Girard (1912),
vol.1, 254 et seq.
3 Supra, note 4, 376-77.
35 Ibid., 384 tan tun ktumatun prasin.
36 Ibid., dous ton arrab~na … ouketf igorakas. Pringsheim adds that the
contrast between the giving of arra and the future sale is evident also in
Byzantine commercial transactions.
36a Ibid., 399-400: “For the act of the giving Si56vat &ppa6cva is used by
Aristotle, Theophrastus, Isaeus and Harpokration, in the Ptolemaic papyri
and the New Testament and in the papyri of the Roman epoch. In corres-
pondence with this term the comedies of Plautus, Varro and the classical
jurists, and even Justinian speak of ‘dare arrabonem'” ‘dare arrae nomine’, ‘datio
arrarum’. The act of the receiver, on the other hand, is called
.ppa67vq by Theophrastus, Harpokration and Ptolemaic and later papyi.
Plautus, the Roman jurists and the Codex Euricianus use the corresponding
term ‘accipere arrabonem (arram)’. What renders this terminology so charac-
teristic is that there are no exceptions. In all epochs everywhere not an
agreement, but the giving and taking is decisive.”
kacA6d(vEv
37 This is the date assigned by de Zulueta, supra, note 9, 23-4, as being that
of the first version of the Codex. Marasinghe, Arra – Not in Dispute (1973)
20 R.I.D.A. 3e sdrie, 349, 353, n.23, agrees. Watson, Arra in the Law of Justinian
(1959) 6 R.I.D.A. 385, says that the communis opinio is that the Codex is
earlier in date than the Institutes.
1977]
COMMENTS – COMMENTAIRES
at their own initiative to enter into an effective arra transaction. The
text of the Codex demonstrates that all the above characteristics, save
the second, of Hellenistic arra were precisely those encompassed by
this imperial constitution. 38
The Corpus Iuris Civilis
Justinian’s codification, illustrative of the mature Roman law,
has given rise to disputes both textual and substantive in nature,
involving Institute 3.23.pr. and Codex 4.21.17. The older text will
be dealt with first. 9
38 This is not to suggest or to articulate the fear of Thomson, Arra in Sale
in Justinian’s Law (1970) 5 IrJur.(n.s.) 179, 183: “I cannot accept, for example,
the view that the Hellenistic law of sale was so prevalent that the con-
sensual Roman contract was quite unknown in the period immediately prior
to Justinian.”
39 The text of the Codex is Krueger’s reading (Krueger, Corpus luris Civilis,
Codex Justinianus (1959), vol.2). The translation of the Codex is Scott’s
(Corpus luris Civilis, The Civil Law (1973), vol.xii). The text and translation
of the Institute are Thomas’s (Thomas, The Institutes of Justinian (1975)).
Both translations adeptly avoid the neat problems of interpretation, some of
which are outlined in the text of this paper (see infra, “The Corpus Iuris
Civilis”).
C.4.21.17 Contractus venditionum vel permutationum vel donationum,
quas intimari non est necessarium, dationis etiam arrarum vel alterius
cuiuscumque causae, illos tamen, quos in scriptis fieri placuit, transac-
tionum etiam, quas instrumento recipi convenit, non aliter vires habere
in mundum recepta subscriptionibusque par-
sancimus, nisi instrumenta
tium confirmata et, si per tabellionem conscribantur, etiam ab ipso corn-
pleta et postremo partibus absoluta sint, ut nulli liceat prius, quam haec
ita processerint, vel a scheda conscripta, licet litteras unius partis vel ambarum
habeat, vel ab ipso mundo, quod necdum est impleturn et absoluturn, aliquod
ius sibi ex eodem contractu vel transactione vindicare: adeo ut nec illud in
huiusmodi venditionibus liceat dicere, quod pretio statuto necessitas venditori
imponitur vel contracturn venditionis perficere vel id quod emptoris interest
ei persolvere … 2. Illud etiam adicientes, ut et in posterum, si quae arrae
super facienda emptione cuiuscumque rei datae sunt sive in scriptis sive
sine scriptis, licet non sit specialiter adiectum, quid super isdem arris non
procedente contractu fieri oporteat, tamen et qui vendere pollicitus est
venditionern recusans in duplum eas reddere cogatur, et qui emere pactus est,
ab emptione recedens datis a se arris cadat, repetitione earum deneganda.
We order that contracts of sale, exchange, or donation, registry of which is
not necessary, gifts of earnest money, or those made for any other reason
which are required to be in writing, and also such as relate to compromise,
shall not have any force unless evidenced by written documents and con-
firmed by the signature of those who execute them; and if they have been
drawn up by a notary, they must be completed by him, and finally acknow-
ledged by the parties interested, so that, where these formalities have not
to claim any right for
been complied with, no one will be permitted
McGILL LAW JOURNAL
[Vol. 23
The principium of C.4.21.17 explains that the validity of those
transactions which the parties have agreed to be executed in writing
will depend on the writing itself. The execution of the document,
(even
himself growing out of a contract or compromise based upon a written
memorandum
though it be signed by one or both the parties),
whether it has not yet been carried out, or is complete; in order that in
transactions of this kind it cannot be said that the vendor was required to
sell the property at a certain price; or that the contract of sale was per-
fected; or that the purchaser should be compelled to make payment … 2.
We also add that, hereafter, where earnest money has been given for the
purpose of making a sale of any kind of property whatsoever, whether the
contract is in writing or not, even though it may not have been expressly
stated what disposition must be made of the earnest money in case the
contract was not carried out, he who promised to sell the property, and
then refuses to do so, shall be compelled to pay double the amount of the
deposit; and he who agreed to purchase it, and refuses to do so, shall lose
the sum which was given, and shall be denied the right to recover it.
Institute 3.23.pr.
Emptio et venditio contrahitur simulatque de pretio convenerit, quamvis
nondum pretium numeratum sit ac ne arra quidem data fuerit. nam quod
arrae nomine datur argumentum est emptionis et venditionis contractae.
sed haec quidem de emptionibus et venditionibus quae sine scriptura consistunt
obtinere oportet: nam nihil a nobis in huiusmodi venditionibus innovaturn
est. in his autem quae scriptura conficiuntur non aliter perfectarn esse emp-
tionem et venditionemn constituimus, nisi et instrumenta emptionis fuerint
conscripta vel manu propria contrahentium, vel ab alio quidern scripta, a
contrahente autem subscripta et, si per tabellionem fiunt, nisi et completiones
acceperint et fuerint partibus absoluta. donec enim aliquid ex his deest, et
poenitentiae locus est et potest emptor vel venditor sine poena recedere ab
emptione. ita tamen impune recedere eis concedimus, nisi iam arrarum nomine
aliquid fuerit datum: hoc etenim subsecuto, sive in scriptis sive sine scriptis
venditio celebrata est, is qui recusat adimplere contractum si quidem emptor
est, perdit quod dedit, si vero venditor, duplum restituere compellitur, licet
nihil super arris expressum est.
Sale is contracted as soon as there is agreement on the price, even
though it be not paid nor earnest (arra) given in respect of it; for what is
given by way of earnest is but evidence that the sale is concluded. This,
however, applies to sales which are made without writing; for we have made
no innovation in respect of such sales. But, for sales effected in writing, we
ordain that the sale shall not be complete unless the document of sale be
completed in the very hand of the contracting parties or- drafted by another
but signed by the parties; and, if it be prepared by a scribe, only when it be
completed and executed by the parties. So long as any of these requirements
be lacking, there is room for reconsideration and either vendor or purchaser
may resile from the contract with impunity. However, we allow them to with-
draw with impunity only if no earnest has been given: for in that event,
whether the sale be in or without writing, the party who refuses to implement
the contract loses what he has given, if it be the purchaser and, if it be the
vendor, he must restore two-fold: and this, although nothing be expressed
about the earnest.
1977]
COMMENTS – COMMENTAIRES
duly signed, was, in such cases, essential.40 Until that moment either
party could withdraw but not without penalty. The penalty was the
forfeiture of the arra for the prospective buyer and the double
indemnity for the vendor. The great controversy with this text lies
in the words “si quae arrae super facienda emptione cuiuscumque
rei datae sunt sive in scriptis sive sine scriptis . . .”. Does the phrase
“sive … sive” refer to the datio arrarum or to the future sale? The
natural construction has “sive… sive” clearly belonging to the giving
of the arra.4′ The preponderance of arra documents and indeed the
principium (which speaks of written memoranda not fulfilling the
requirements) indicate that a written document of arra is neither
particularly unusual to the Byzantine mind nor, necessarily, a fund-
amental breach of Roman consensualism. If this is so, then the arra
of C.4.21.17.2 is penal since there is no agreement for it either to
confirm or evidence. Honore4 2 and Thomas both remark that the
language of the constitution: facienda emptione not contractum
emptionis, [is] qui vendere pollicitus est not venditor, qui emere
pactus est not emptor, indicates a future sale in writing. This would
be in harmony with the remarks on the validity of a written sale
made in the principium. What conclusions can be made of this
passage? Thomas has one:
In the light of what has been seen earlier, it would clearly appear that
Justinian is indeed applying to the written sale the regime which charac-
terised the Hellenistic sale. Just as in the latter, the agreement to buy
and sell did not of itself bind the parties, so now the agreement of sale
does not bind them before the completion of the document: but if arra
4o Cornil, supra, note 33, 259 remarks on the deep-rooted belief in the East
that it was impossible to conceive of a contract which was not executed in
writing and at p.260 he says: “D’ailleurs la ddcision de Justinien relative aux
contrats h conclure par dcrit … ne s’inspire-t-elle pas des pratiques usuelles
dans le bas-empire, bien plut6t que des principes du droit classique?”
41 Contra, Honor6, Arra As You Were (1961) 77 L.Q.R. 172, 174 who says
that arra cannot take place in writing. In favorem, Thomas, Arra in Sale in
Justinian’s Law (1956) 24 Tijdschrift voor Rechtsgeschiedenis 253, 266-67, who
says that such a conclusion is envisaged by the principium and evidenced by the
papyri. He adds at p.267: “From the point of view of the sale, as the constitu-
tion is concerned with written sales, it is at least improbable that a reference
to unwritten transactions wotld have been here introduced without more, by
a simple sive … sive: the improbability becomes the more manifest when one
remembers the full distinction taken between the written sale and the normal
consensual regime in the principium. From a purely linguistic and syntactical
standpoint, the phrase clearly goes more easily with the arra than with sale.”
42 Honor6, supra, note 41, is quick to point out that such a conclusion does
not mean he accepts “sive … sive” as going with the giving of arra.,
McGILL LAW JOURNAL
[Vol. 23
-TCov &ppa6&v’)v
has been given, the parties are subjected to the v6j.o
in the event of their not proceeding to the appropriate completion. 43
The younger text presents as many problems. The first portion
of the Institute states the general rule of consensual sale, to wit, that
arra is merely evidentiary, as in Gaius. The next part is a simplifica-
tion for handbook purposes of C.4.21.17.pr., that is, the law of written
contracts.” Finally the Institute seems to redirect itself toward a
general discussion of arra. The troublesome “nam nihil a nobis
in huiusmodi venditionibus innovatum est” does not purport to say
that the Institute should be interpreted in accordance with classical
Roman law but rather that the introduction of such a concept should
not shock those used to the laws of Byzantine sale and Hellenistic
arra, so prevalent in the sixth century. Indeed, Justinian is archaizing
here. He is reintroducing the evidentiary arra to sales where no
writing was envisioned.4V 5 The Institute continues to speak of the
applicability of arra to sales both written and unwritten. Clearly
Justinian was allowing for penal arra in a complete unwritten sale.
Thus, by extending the rules of arra covering written contracts of
the Hellenistic tradition to cases where the parties required no form-
alities of each other, Justinian gave a new right.40 Arra, therefore,
tuzioni di Diritto Romano 2d ed. (1927), 315-16.
43 Thomas, supra, note 41, 266, a conclusion supported by Arangio-Ruiz, Isti.
44 Marasinghe, supra, note 37, 351 says: “One gets the impression that the
Codex was a development from the Institute.” Such a conclusion is not
necessary given the different purposes of Justinian’s tomes. Thomas, supra,
note 41, 264-65 distinguishes.
45 Levy, Western Roman Vulgar Law The Law of Property (1951). Levy says
at p.149: “Nothing could more graphically characterize the spirit of renaissance
permeating the codification”, and again at p.150: “The parties were at liberty
to choose between the old and the new with the legislator regarding as the new
what to the practicing lawyer was the inveterate thing.”
46 Watson, supra, note 37, presents the contrary arguments. According to his
analysis, Justinian did not modify the law of unwritten sales, that is to say
that arra could serve only an evidentiary function and not a penal one. Con-
sequently, the “sive in scriptis sive sine scriptis venditio celebrata est” must
not be made to refer to complete written sales and complete unwritten sales. It
refers to sales where there is writing which does not fill all the formal re-
quirements of C.4.21.17, and to sales where there is no writing but where
writing is clearly envisioned. To this end Watson considers venditio to mean
“an imperfect sale” and celebrata est to mean “was made”, so that the whole
reads: “where an imperfect sale was made in writing or without writing”.
However, even if venditio means an imperfect sale, celebrata est cannot be
easily diluted. Celebrare is a formal verb with strong overtones of solemniz-
ation. Honor6,.supra, note 41, demonstrates that the celebrata est must refer
to a complete sale which is, so to speak, signed, sealed and delivered. Further-
more, references to celebrare throughout classical and post- classical literature,
19771
COMMENTS – COMMENTAIRES
fulfilled in a complete unwritten sale both an evidentiary and a penal
function. This combination appears less bizarre when consideration
is given to the amount of the arra – often a substantial proportion 4 7
There may not have been any need for ordinary judicial damages
since de facto damages, achieved by the giving of arra, would have
been determined by the parties at a preliminary stage. Yet there
existed the possibility of a suit for damages in addition to the
liquidated damages of arra. The two are not incompatible48 In cases
where no arra was given, the aggrieved party had at his disposal
the judicial recourse of an action on the sale, or an action for
recovery of the thing sold would lie.49
Conclusion
An absolute freedom of contract may be defined, for these pur-
poses, as the capacity to determine the moment when the contract
becomes legally sanctionable, its form and substance, the extent of
the legal sanction, and its effect on pre-contractual agreements.
Those who hold that consensualism precludes the application of
penal arra as being unnecessary, since contractual liability would
arise from the contract itself and not additional arrangements of a
penal nature, would do well to remember that contractual freedom
also embraces those instances where the mere will of the parties de-
termines the contract. Consequently, parties desirous of unwritten
agreements need not extend the arra beyond its evidentiary functions,
should that be their intention. However, it is submitted that con-
although not in legal writing, would militate against Watson’s conclusions.
Watson’s hypothesis is based on a separate appreciation of each word. It
would appear more sensible to analyze venditio celebrata est as a whole,
especially since both words, separately considered, are susceptible of different
meanings in different contexts. A natural reading therefore, would be “whether
the sale had been completed in writing or without writing”. See also Moyle,
The Contract of Sale in the Civil Law (1892), 43, and Tylor, Writing and Arra
in Sale under the Corpus luris (1961) 77 L.Q.R. 77. Tylor comments that the
real contrast should not be between written and unwritten sales, but between
sales with a fixed price and those void until the formalities of writing were
entered into. He notes that agreement as to a fixed price was crucial to the
consensual contract of sale.
7 Cornil, supra, note 33, 261-62 says that the amount of damages would
4
generally be inferior to the arra given.
4 8 Contra, Buckland, Elementary Principles of Roman Private Law (1912), 274.
In favorem, Honor6, supra, note 41, 174, explains that since sale was ex fide
bona any forfeiture of arra would be taken into account in an action for
damages.
49 D.19.1.11.6 (Ulpian) where it is revealed that Julian favoured the actio ex
empto while Ulpian himself preferred the condictio or action for recovery.
McGILL LAW JOURNAL
[Vol. 23
sensualism is only a stepping stone to the larger, more generous
principle of freedom, and that such a freedom is contained in
Justinian legislation on arra when the modalities of the sales trans-
action were entirely subject to the will of the parties. What could
be a more liberal attitude than to allow written or unwritten arra
agreements in written or unwritten contracts of sale with evidentiary
and penal function? All was regulated. Everyone was satisfied.
Michael McAuley*
* BA. (Hons.) (London), A.K.C. (King’s College London), B.C.L. (McGill),
Macdonald Travelling Scholar.
