9

Damnum Iniuria Datum

(Loss Wrongfully Caused)

. The Shape of the Delict

Damnum means ‘loss’, though it is often still found incorrectly translated
as ‘damage’. In fact this delict is largely about damage, but the damnum is
the loss which arises from the damage, as opposed to the killing, burning,
breaking, bursting and so on. Iniuria is used here in the ablative case and
means ‘by a wrong’ or, moving safely to an adverb, ‘wrongfully’. And
datum means ‘given’ or ‘brought about’. So at a high level of generality
the delict is simply ‘loss wrongfully caused’.
The shape of the delict is dictated by the fact that it is based on a
statute, the lex Aquilia. The juristic interpretation of the lex left the
demand for further remedies still unsatisfied. The praetor therefore
gave actions which overrode the technical limitations of the statutory
remedy. These actiones in factum can be contemplated as clustering
around the area covered by the lex itself. Thus whenever one has to
analyse a disaster which looks as though it may give rise to a liability for
loss wrongfully caused, it is necessary to ask first whether it falls within
the statutory core and then, if it does not, whether it finds a place in the
praetorian periphery of the delict.

. The Statute
The lex Aquilia was actually an enactment of the concilium plebis. So
technically it was a plebiscite and not a lex. But that strict usage was
The Roman Law of Obligations. Peter Birks.
© Oxford University Press 2014. Published 2014 by Oxford University Press.
 DAMNUM INIURIA DATUM ( LOSS WRONGFULLY CAUSED ) 

departed from. We know that plebiscites were made to bind the whole
people, rather than just the plebs, by the lex Hortensia in  BC. On
rather slender evidence it has been thought that the lex Aquilia was
passed as a result of the same social disturbances, almost immediately
after the lex Hortensia itself. More recently Professor Honoré has argued
for a slightly later date, around  BC.1 One argument in favour of that
date is that it was then that the currency got into trouble as a result of
the inflation brought about by the war with Hannibal. The relevance
of that is that one of the provisions which the lex Aquilia displaced was
the rule of the Twelve Tables relating to the fracture of a slave’s bone.
The early code provided for payment of a fixed sum,  asses. Inflation
brings down the real value of redress which is fixed in that way.
Because of the war with Hannibal there would have been a need for
reform. So  BC may be our best estimate of the date. If we were
brutally honest we would have to say that it must have been passed
after the lex Hortensia and before the middle of the second century BC
when we find Brutus commenting on it. So between  and ,
with  most probable. Some people believe that the statute as we
have it was built up over a period of time as different legislative sallies
were made into the field of loss wrongfully caused.
We know something of three main sections or chapters of the lex.
I shall briefly mention the second first, in order to get it out of the way.
Ch. II was early obsolete. We do not know its words. It was passed,
like the other, de damno, concerning loss. But the loss was of a highly
specialised kind. The chapter dealt only with the case of an adstipulator
who in fraud of the stipulator released a debt owed by the promisor. An
adstipulator is a co-promisee to a promise made by stipulation. He is in a
position to release the obligation either artificially or by accepting
payment. It is not completely clear whether the statute contemplated
only artificial release. Probably it contemplated both. Gaius tells us, at
G.., that the contract of mandate would serve to regulate relations
between the principal and the subsidiary promisee. But he notices that
the lex lay for a double penalty against one who denied liability, and
one would have thought that that would have kept the section alive.
Chapters I and III are both about loss arising from damage of various
kinds. The immediate task is just to set out their wording. Matters of
interpretation will be dealt with in the next section.

1
A. M. Honoré, ‘Linguistic and Social Context of the Lex Aquilia’,  Irish Jurist (NS)
(), , ff.
 DELICTS

Ch. I
If anyone wrongfully kills another person’s male or female slave or four-
footed herd-animal, let him be condemned to pay to the owner as much
money as at the highest that thing was worth in the preceding year.2
Ch. III
In respect of other matters [besides slaves and herd-animals killed], if
anyone causes loss to another by burning, breaking or bursting wrong-
fully, let him be condemned to pay to the owner as much money as that
matter is worth in the nearest thirty days.3

These translated versions gloss over points of detailed dispute but

nothing of substance is lost which cannot be recovered in the later
discussion. Ch. III has caused a good many headaches. It is difficult
partly because its opening words seem to attach directly to ch. I,
whereas on the day when the lex was passed ch. II had not yet fallen
into disuse and could not be ignored by the draftsman of ch. III, and
partly because the words which quantify the amount to be paid seem to
contemplate full value for small and partial losses, as where you chip my
plate or burn my slave’s arm. Apparently you might as well have killed
him, since the difference between valuation over a year and over a
month will rarely have been great and never wide enough to discrim-
inate between a burned slave and a dead one.
These considerations have led to some radical suggestions to the
effect that the text of the chapter was originally much more restricted
in scope than it now seems. Jolowicz proposed that it must have been
concerned only with total destruction of inanimate things, so that all
three chapters would have been about the extinction of items of
wealth.4 Daube on the contrary maintained that the chapter was
about wounding, rather than killing, chapter I objects, slaves and
herd-animals.5 Between these two radical approaches Daube has
much the better of the argument. However, though some difficulties
remain, the pendulum has swung back towards accepting the text as it
stands subject only to the elimination of the words in brackets which in
Latin are awkward and, in either language, read like an explanatory

2
D... pr. (Gaius,  On the Provincial Edict).
3
D.... (Ulpian,  On the Edict).
4
H. F. Jolowicz, ‘The Original Scope of the Lex Aquilia and the Question of Damages’,
 LQR (), , ff.
5
David Daube, ‘On The Third Chapter of the Lex Aquilia’,  LQR (), , ff.
 DAMNUM INIURIA DATUM ( LOSS WRONGFULLY CAUSED ) 

gloss. Professor Honoré takes this less drastic line. He thinks that ‘in
respect of other matters’, when standing alone, can follow on from
both preceding chapters; and his later date here helps since the word for
‘thing’ (res) has more time to become abstract.6 So far as the measure of
ch. III damages is concerned, it is too much in dispute to be allowed to
control one’s view of the substantive scope of the early lex.

. Interpretation in the Statutory Core

One ought to separate the treatment of the two chapters but for brevity’s
sake they can be taken together. The measure of recovery which
constitutes a major difference can be severed and dealt with later.
The matters to be considered are covered by two large questions:
Has the plaintiff suffered a disaster which is prima facie within the lex?
And, Is the defendant liable for that disaster? These two questions can
be broken into smaller ones, so that in the end one has three for the first
and two for the second. Thus: (i) Has the plaintiff suffered loss (dam-
num)? (ii) Did that loss arise from a thing spoiled (res corrupta)? (iii) Did
that spoiled thing belong to the plaintiff (res actoris)? If these are all
answered in his favour the plaintiff has suffered a disaster within the lex.
Then, (iv) Did the defendant do the spoiling with his own bodily force
(corpore suo)? and (v) Did he do it wrongfully (iniuria)? If these last two
are answered against the defendant then he must pay.
The form of some of these questions anticipates the discussion, but
they will become clear.

i. Has the plaintiff suffered loss (damnum)?

This requirement is not mentioned expressly in ch. I. It is virtually
impossible to construct a set of facts in which a plaintiff whose slave or
beast is dead will not have suffered loss. In chapter III the first condition
is, expressly, ‘If anyone causes loss to another . . .’ Not every bump or
bruise will do so. Suppose the reaction of a hypothetical buyer.
A scratch on a polished table will reduce its value, the same thing on
the outside wall of a house will not affect the price. Similarly with
living tissue. An ordinary black eye spoils a slave’s looks but unless there
is some evident risk of complications will not take down his value.

6
Honoré (n. ), –.
 DELICTS

Ulpian reports a famous example given by Vivianus.7 Successful cas-

tration of a slave-boy would actually raise his value. Hence there could
be no Aquilian liability.
All these examples assume that the verbs, bumping, bruising,
scratching, amputating, would be within, and indeed they are within,
the range of the statutory trilogy, urere, frangere, rumpere, one aspect of
which is considered next.

ii. Did that loss arise from a thing spoiled (res corrupta)?
Under the lex itself the action described by one of four verbs had to
happen. Each has an active and a passive aspect. At this point we are
concerned only with the passive aspect. This means that we are not yet
interested in the question whether the defendant did, say, the burning
but whether the plaintiff suffered a thing burned.
Under chapter I it is evident that the plaintiff had to have suffered a
slave or herd-animal killed. The words ‘thing spoiled’ do not arise under
that chapter but clearly servus etc. occisus is within the notion of res corrupta.
Under chapter III there has to be a res usta fracta or rupta, a thing
burned, broken or burst. The translation of this last word as ‘burst’ is
conventional but not very good. ‘Damaged’ would be better though
the trilogy would run less well if deprived of its alliteration. ‘Rumpere’ is
a very wide word, much less specialised than ‘burst’. ‘Frangere’ is rather
narrower, for dry breaks as of bones, plates and pencil-leads.
Defendants would naturally try to escape by arguing that the con-
sequences of their action could not be brought within one of the verbs.
Suppose that I have poured out your wine onto the floor. You sue me
and I say that this is certainly not within occidere, urere or frangere. It must
be rumpere or nothing. And what have I smashed or damaged? Ulpian
tells us that the veteres, the old jurists, put a stop to this tactic by giving
rumpere its widest possible sense. The word rumpere was understood in
the sense of corrumpere. Much as though we said ‘damage’ must be taken
as ‘spoil’. That reached the spilled wine clearly enough. One could not
argue that it was not ‘spoiled’ or ‘ruined’. This interpretation of rumpere
as corrumpere meant that urere and frangere were really redundant. Ulpian
reports Celsus as saying that he was bound to admit that the first two
verbs—fractum and ustum—were contained within corrupti appellatione,
the description of something as having been spoiled. But it was nothing

7
D.... (Ulpian,  On the Edict).
 DAMNUM INIURIA DATUM ( LOSS WRONGFULLY CAUSED ) 

new, he said, for a statute to follow special words with some general
catch-me-all. Gaius says that the interpretation as corruptum meant that
ruptum would reach not only things usta aut fracta but also scissa et collisa et
effusa et quoquo modo vitiata aut perempta atque deteriora facta: burned, broken,
torn, crushed, spilled, vitiated in any way, destroyed or made worse.8
It has long been held that there could not be said to be a res corrupta
where something was destroyed in the course of its proper use. It is
certainly true that I cannot be liable under the lex Aquilia for eating
your food or drinking your wine, but the reason is not likely to be
found under this head. It is hard to say that there is no res usta when
I burn your logs, just because logs are for burning. And when I eat your
walnuts I must frangere their shells and rumpere the nuts within. There is
no evidence that the interpretation as corrumpere did other than extend
the statutory words. To exclude these cases under this head one would
have to suppose it capable of restricting as well as extending.
The dynamic interpretation of rumpere had its limits. Moth, rust and
wrongdoers can corrupt most things, but not all. Precious stones and
metals are unaffected by immersion in the sea or in a river. Suppose
I throw your silver cup into the sea or your ring into the Tiber or
knock your coins down a drain. There is no res corrupta and so no
liability under the lex. There is one indication that Sabinus was pre-
pared to stretch this point presumably on the ground that total dis-
appearance could be regarded as a type of destruction. However, that
view, objectively incorrect, seems not to have prevailed. There are
other examples, besides precious metals and gems, of things which can
be made to vanish without damage. If you trap an animal and I let it
back into the woods you suffer damnum but there is no corpus laesum.
Perhaps nearer the line is the case in which I sow weeds in your corn.
You get less yield but no res is actually corrupta ac mutata.
It seems therefore that under the lex there had to be at least some
corporeal deterioration, some internal change for the worse. As we
shall see the praetorian actions could go further.

iii. Did the spoiled thing belong to the plaintiff (res actoris)?
In ch. I the slave is described as alienum alienamve, belonging to another.
There is nothing parallel to that in ch. III so far as concerns the words
which describe the delict. But in the words laying down the measure of

8
G...
 DELICTS

damages, in both chapters, the plaintiff is described as ‘owner’. For it is

to the owner, probably the lex used the old word ‘erus’, that the
defendant must pay the value. Hence only an owner could sue, and
only in respect of a res corrupta owned by him.
This restriction was inconvenient in two different ways, both of
which made work for the praetor. First, there were people other than
owners who suffered by the commission of the delict. The usufructu-
ary, the bona fide possessor, the pawnbroker come immediately into
view. Secondly, it meant that there was no Aquilian remedy for injuries
to a free man. For nemo dominus est membrorum suorum: no-one owns the
parts of his body. We will see that a free man could have the actio
iniuriarum, but only for intentional injury. There could also be other
problems. One’s family would refer to ‘our vault’ or ‘our tomb’ but if
people had been buried already ‘our’ would be inaccurate. Such places
belonged only to the gods below. Again fixtures could pose questions.
If I smash down your aqueduct running over my land, ‘your’ will not
stand up in court. The fixture is mine by accession to the land. The
materials which I have released are yours, but are not damaged.

iv. Did the defendant do the spoiling ‘corpore suo’?

We now turn to the defendant’s liability. The first aspect is the active
element in the verbs. He must have ‘killed’ the slave or animal or
inflicted loss ‘in that he burned, broke, burst’. As a defendant seeking to
escape, he will say that, though you have a dead slave or a spoiled
bushel of corn, yet he did not do the killing or did not do the spoiling.
The question is, When can a man be said to have ‘done’ something, as
here ‘killed’ or ‘spoiled’? And this has to be answered not just reason-
ably but in the teeth of all the unreasonableness which the defence can
muster. Suppose I put poison in your slave’s bedside drink. In the
middle of the night he wakes and, feeling thirsty, drains his glass. I shall
say that he killed himself. I did not do the killing, even though in a
sense outside the statute I did cause his death. What the lex says is ‘Si
quis servum . . . occiderit’,9 not ‘Si quis servo . . . mortis causam praestiterit’.10
This is a real problem. And there is a complicating factor. As we shall
see the lex was not confined to cases of deliberate damage. Our
question, Did he kill, burn and so on?, is more difficult to answer if it

9
i.e. ‘if anyone kills a slave’.
10
i.e. ‘if anyone provides the cause of the slave’s death’.
 DAMNUM INIURIA DATUM ( LOSS WRONGFULLY CAUSED ) 

has to be answered without reference to the defendant’s intention. In

short, it is easier to say he killed if he planned the death. In the example
just given it is because I did intend to poison the slave that my defence
looks very artificial. Suppose instead that the slave died of typhus. I was,
innocently, a carrier of the disease. When I put his drink ready
I infected it wholly without my knowledge. Unwilling cause of his
death, did I kill him? Suppose again that your field of corn has burned
down. I made a bonfire half a mile away and the wind carried the
sparks. Did I burn the corn? I never thought about it, far away from my
garden and my mind. It is difficult to say I burned it. You were my
enemy. I waited till the wind was right. And then I lit my fire, willing
the sparks to carry. Easier now to say I burned the corn. Samson tied
torches to foxes’ tails. The foxes burned the Philistines’ corn.11 But he
wanted it burned and got the credit for it.
Because the lex was understood to create a liability even for unin-
tentional harming, this problem could not be solved in terms of intent.
It had to be looked at in terms only of forces, of physical causation.
The test which was adopted was this: the defendant must have
brought about the damage corpore suo, by his body. Gaius, at G..,
says ‘But it has been held that the action based on that lex lies only
where someone corpore suo damnum dederit, caused the loss by his body.’
Three points need to be made about this solution to the problem.
First, it is not primitive. ‘Primitive’ in this kind of matter is to have
no answer at all, to leave the question to unpredictable intuition. To
say that a man only kills or spoils when he does so corpore suo, you need
both to have thought about the problem and to have mustered the
authority to impose an artificial solution.
Secondly, it certainly does not mean that the defendant must have
come into bodily contact with the res damaged. That would, at the
most extreme interpretation, rule out even wounding with a sword,
smashing with a stick. Your slave is transfixed by a javelin. I threw it
from some seventy yards’ distance. Your cart is crushed by a block of
marble. On a scaffolding fifty feet above, I, a mason making ready to
use it, let it drop. These are clear cases, centrally within the ‘with his
body’ test.
‘Corpore suo’ is better translated with the addition of an extra word,
‘by his bodily force’. The idea is easy enough to capture in general

11
Judges :–.
 DELICTS

terms. The passive aspect of the verb, which we have already looked at
(occisum, ustum and so on), must have been the immediate result of the
defendant’s bodily force. Or, the defendant’s body must have been the
last kinetic force to operate, the last source of movement. It is worth
remembering at once that in the ancient world gravity was not a force.
That is to say, if I dropped a lump of marble or took the brake off a cart
at the top of a hill it was my force which sent it into your slave’s head or
over his toe, not the unseen hand of gravity.
Thirdly, even though the general idea of the test can be made out, its
detailed application is inescapably difficult. There is nothing new in this.
Often a test which is necessary and useful is difficult to apply to the facts.
In the English case of Scott v. Shepherd12 the defendant had thrown a
firework into a market. Several stall-holders knocked it on away from
themselves until when the fuse had burned it blew up in the plaintiff ’s
eye. The action was in trespass vi et armis. English law had reached a test
remarkably similar to the Romans’ corpore suo. The question was, whether
the plaintiff ’s eye had been injured by a direct and unlawful act of force
on the part of defendant. Agreed on the test, the court divided on its
application. The majority thought that the injury had been done directly.
Quite apart from the inherent uncertainty of the formula to be applied,
opinions will be affected by extrinsic factors. For example, the colour of
the question will be different according as the choice is between liability
and no liability at all or between liability under one head and liability
under another. Substantial justice may be in issue in the first case, only
clarity in the other. Again, the hardship to be inflicted on a particular, and
especially deserving, plaintiff may enter into the decision. Among those
who decide some are willing to lean more heavily than others.
These tensions are present in the Roman texts. It is wrong to look
for perfect consistency, even in the same jurist. Consistency in appli-
cation, that is. It would be shocking to find different tests in play. In
classical law, when the praetorian supplementary actions were regularly
available, there was less pressure for a liberal interpretation of corpore
suo. The main thing then was a clear line round the action under the lex
itself, the statutory core of the delict.
Some cases are quite easy. I lock your slave up or impound your cow
and then give no food, so that death follows by starvation. There is no
bodily force involved in the death. I send your slave up a tree and he

12
 Wils KB ;  ER  (King’s Bench ).
 DAMNUM INIURIA DATUM ( LOSS WRONGFULLY CAUSED ) 

falls, or down a well and he fails to return to the surface. Again, it

would be difficult to argue that his death or injury happened corpore
meo. These examples are given in Gaius, .. My example of the
poisoned drink is clearly outside the test. If I held the slave down and
made him drink, that would be on the other side of the line. Similarly
with poisoned ointments rubbed. Or hyperdermic application. If I cut
a rope, I damage it corpore meo. If the boat attached to it then floats off
and is driven by wind and wave onto a rock, its destruction is caused by
me but not done by my bodily force. If I puncture a cistern of wine the
hole is made directly. I have damaged the container within the mean-
ing of ch. III. Suppose, however, that the wine pours or seeps out and
spills on the ground. That happens indirectly, not corpore meo. That may
be the least clear of these examples. What squeezes out the wine is
gravity, and gravity is not supposed to count. But even we think of the
weight of water as an independently identifiable force.
There are many more troublesome cases. The common character-
istic of all difficult facts is that more than one force, or more than one
body, is involved. Suppose a slave knocked into a river or into the sea.
He drowns. At G.. Gaius says this is outside the lex but he raises his
eyebrows and admits that one might as easily say the opposite. Justin-
ian, at J..., does say the opposite. There was a dispute and, strictly
but very inconveniently, the cases should have been divided. If he
bobbed around and, struggling, was eventually overcome, the river
killed him, not me. If he went down once and for all or, more clearly
still, if he broke his neck on impact, I killed him, corpore meo. Again,
suppose I set my Alsatian to bite you, adding length and strength to my
arm. Have I done it? Julian said, only if I kept hold of the dog; but a
more liberal view had prevailed earlier.13 Finally, suppose you push me
and I thus injure the plaintiff ’s slave. For example, you kick a ball. It
strikes my elbow. I, a barber, am shaving the slave. I cut his throat.
Have I wounded the slave? At one point Proculus seems to have no
difficulty in saying that I have;14 at another the mere intervention of
my body seems to make it impossible for him to say that I did the injury
corpore meo.15 That may be unfair. In this kind of issue one often needs

13
D.... (Ulpian,  On the Edict).
14
D... pr. (Ulpian,  On the Edict).
15
D.... (Ulpian,  On the Edict). Birks appears to have mixed up the parties in his
manuscript; the text of this paragraph was altered so as to reflect what he most plausibly meant
to say.
 DELICTS

more facts than the texts give us. In the second case perhaps your
intervening body was not helplessly inert, or not to the same degree as
the barber’s arm struck by a flying ball.
There is nothing mysterious about this section. The statute said that
a man was liable if he killed, if he burned, if he broke and so on. The
jurists held that he could be said to have done these things if he had
done them corpore suo, with his body. It was not to be expected that
they would all agree all of the time as to what facts satisfied the test. If
someone brought the action under the statute the defendant would
very likely run an argument under this head. The iudex would have
to decide.

v. Did the defendant do the harm wrongfully (iniuria)?

The point we have reached is this. The plaintiff has suffered a disaster
within the lex. The defendant has done it. If the lex had imposed strict
liability, that would be the end of it. But it did not. More accurately, it
left room for argument, and the decision was that it did not. The lex did
not say ‘If someone dolo malo kills another . . .’ or ‘If someone negligenter
causes loss . . .’. The word it used, in both chapters, was iniuria. This, in
the ablative case, is the noun made from the negative particle ‘in-’ and
the word for ‘right’ or ‘law’ which is ‘ius’, as though we would say ‘un-
right’. So in the case of our defendant the question after the lex was
passed was whether he had done his occidere, urere and so on
‘unrightfully’.
That wording is neutral on the question of the kind of fault to be
required, if any. All it implies is that there are some cases which are not
wrongful, some which are. The content of right and wrong is not
given. We do not know exactly the steps by which that vacuum was
filled but we can consider the choices.

A. ‘Wrongfully’ becomes ‘by malice or by fault’ It is a natural human

instinct, when faced with an accusation of damage and injury, to say,
if there is room for the plea on the facts, ‘Oh, it was only an accident.’
A child does it almost automatically when a cup is broken or milk
spilled. It is meant to avert responsibility. ‘You cannot be cross with
me’ becomes in an adult and in a court ‘You cannot punish me or make
me pay.’ An accident is like the measles. As victim you must complain
to God, not your fellow man. Insurance has been provided as a means
of material solace. After all, it is not clear why anyone should suffer any
 DAMNUM INIURIA DATUM ( LOSS WRONGFULLY CAUSED ) 

loss if a means can conveniently be found to save them. Certainly it is

not a reason against the other means of relief that no fellow human is
answerable for your disaster. Because it was an accident.
There is a trace of this plea already in the Twelve Tables: ‘If the
weapon flew from his hand rather than he threw it, let him offer a
ram.’16 The idea is evident and familiar. In the case of accident the
principal burden of responsibility is averted. But something remains,
just for causing the disaster or perhaps just for being involved in it. The
token offering puts the matter straight.
‘Accident’ is itself a notion which requires the law to make choices.
The child says, ‘It was an accident. I didn’t mean to do it.’ ‘Accident’
there is everything which happens unintentionally. But the parent may
still get angry: ‘You are old enough to know better.’ Or, ‘You could at
least have been more careful.’ Here ‘accident’ is the disaster which
happens without blameworthiness or fault, the premise being that
there can be fault without deliberate intent. Lines can be drawn at
different places, using different words and concepts, but these two are
the most easily marked. The child’s position is that liability should be
for dolus only, that is, for intentional harm, with everything else classed
as casus, accident. The parent’s view is that liability is for culpa, with casus
covering only what happens entirely without fault.
We must get back to the statute. It said the act must be done
wrongfully. Someone was bound to say sooner or later that an acci-
dental occidere or rumpere could not be regarded as wrongful. These
were the choices: accept the argument or reject it entirely; if accept,
either on the basis of the wider or of the narrower notion of accident.
The law’s choice was to accept that accident exonerated but, like the
parent above, to narrow its scope to the case in which there was no
fault at all.
That is a re-construction of the process by which a lex which
imposed liability for what was done ‘wrongfully’ came to be under-
stood as basing the liability on culpa. It is not easy to say how quickly
this interpretation was completed. It seems to be already in place in all
our sources. Alfenus Varus, pupil of Servius Sulpicius and contempor-
ary of Augustus, handles Aquilian problems in terms of culpa.17 That is
two hundred years after the lex. There is no reason to think the culpa
approach was not achieved much earlier.

16
Twelve Tables .a.
17
D.... (Ulpian,  On the Edict); D.... (Alfenus,  Digest).
 DELICTS

The developed position is summed up by Gaius at G..:

A person is understood to kill iniuria if the result is brought about by his dolus
(malice) or culpa (fault). Loss which happens sine iniuria (without wrongful-
ness) is not brought within any other statute. And so no liability is incurred
by a person who inflicts loss without fault or malice by some mischance (sine
culpa et dolo malo casu quodam damnum committit).

There is another rather different picture of the development which

I will mention in section (c). For the moment it is more convenient to
look at the content of iniuria understood as dolus or culpa. What facts
would and would not amount to ‘malice or fault’?

B. The content of iniuria (= dolo aut culpa) It is safe to contemplate this

in two parts. In the first the defendant’s claim to be exonerated is based
on casus; in the second on some other ground.
Where the defendant relies on casus, necessarily he will not have
intended harm to the plaintiff ’s property. The question is, What
exactly is an accident? We have already said that the commitment to
culpa, rather than dolus alone, is acceptance of a narrow defence of
accident as an event which happens without fault. At this point we
have to add some detail. In judging fault a high standard is taken. One
text talks of levissima culpa, the slightest fault, as being sufficient. The
expression may not be classical but the thought is right.
Suppose the defendant was not conscious of any want of care on his
part. As a matter of fact he did not appreciate that there was any risk,
was therefore not aware of taking one. He was pruning a tree in a field.
From time to time he threw down branches. Getting on with his work
he forgot there was a path below. He crushed a slave. He now says that
he not only did not intend harm but also took no risk that it might
happen. That is no answer. He ought to have foreseen the danger. It
counts as culpa under the lex Aquilia to omit precautions which might
have been taken. The question is not whether he was consciously at
fault but whether he did all that could have been done to avoid the
harm. Suppose he burned off the stubble in his field and the fire
escaped and destroyed my crops. It is not enough for him to say that
he thought he had done everything necessary. The question is whether
in fact he had done everything which ought to have been done. We
would turn immediately to the reasonable man. What is the standard
behind ‘ought to have done’ or ‘ought to have foreseen’? We say that
people should foresee harm which a reasonable man would foresee and
 DAMNUM INIURIA DATUM ( LOSS WRONGFULLY CAUSED ) 

take such precaution to avoid it as a reasonable man would take. The

Romans do not restate culpa in that way. Yet many of their conclusions
correspond to those of the reasonable man, and one text comes close to
putting the matter in our way: ‘It is culpa not to foresee what a
painstaking man (diligens) would have been able to foresee.’
Another defendant may say that not only did he not appreciate the
risk but that he personally was incapable of doing so. He did his very
best, but he could not measure up to the standards attainable by the best
of men. He too is caught by the approach described in the preceding
paragraph. Suppose his point is that he himself is not very intelligent.
He cannot foresee or judge as others do. The only allowance which the
law makes is for exceptional categories at so to say the extreme end of
the spectrum. There are people who are relieved of all responsibility,
furiosi (madmen) and young children. Everyone else is assumed to be
equally capable for the purpose of legal responsibility. That does not
mean that lawyers could not make moral distinctions, only that for
practical reasons the law did not.
A third defendant may say that he did as well as the best man possibly
could but that he lacked some special skill. His case is that people
should be judged by a standard of general knowledge without assum-
ing special training such as only some ever get. I ride my horse into
you. I do everything that could be done by a careful man. My trouble is
only that I am but a beginner in horsemanship. I bought the horse
yesterday and today when I failed to hold him back I was learning to
ride him. My argument is that I should not be judged as one who had
acquired the special skill of riding, only by the standard of an ordinary
unskilled man doing his best to avoid damage. Or, suppose your slave is
bitten by a snake. I cut into him and cause a disaster. I did what the best
of laymen would have done, but a doctor would have done something
different. He would have known that for this kind of snake a day’s rest
was all that was needed. Again I want my conduct assessed from the
standpoint of general knowledge, assuming an absence of special skills.
I am prepared to be measured by a high standard but not as a rider or a
doctor when in fact I am neither. The law’s blunt answer is: imperitia
(want of skill) counts as culpa (fault). If I engage in an activity which
requires a special skill I must answer for loss which happens because
I lack that skill. There is room for argument in cases of urgency. If the
fault consists not so much in lacking the skill as in embarking without it
on an activity which requires it, then I may be free of fault in cases in
which there was no possibility of getting in an expert at the time when
 DELICTS

I entered on the task. If the snake-bite happened miles from anywhere

and death seemed imminent, I may be free of culpa even though a
layman.
This discussion has been designed to show that culpa does not require
conscious risk-taking, it is not judged by the particular defendant’s
own, perhaps limited, ability to appreciate the consequences of his
conduct, and is not excluded by the fact that something more than
general knowledge would have been needed to avert the disaster. The
question is not, Did he foresee the damage? or even, Could he have
foreseen it? Rather, Could he have foreseen it as a paterfamilias of sound
judgement and intelligence possessing the skills necessary for the proper
conduct of the operation in question?
The second part of the inquiry into the content of iniuria here, i.e. of
dolus and culpa, concerns those cases in which the defendant did intend
the occidere or corrumpere but claims to have been justified by the particu-
lar circumstances. The word ‘justified’ exactly matches the word iniuria.
It means ‘made right’. This is about cases in which the damage is done
iure despite being done on purpose. Again I do not think there is any
conflict between the word iniuria and the notion of culpa in these cases.
What is deliberate but justified in the eye of the law is sine culpa, without
blame or fault. I shall deal only with some examples.
The simplest case is this. Under the early law, by a provision of the
Twelve Tables,18 it was lawful to kill a thief caught by night. It is
evident that this right was whittled away by juristic interpretation so as
eventually to be assimilated to self-defence. However, at all stages of its
development it provides a good model of a defence by way of justifi-
cation. The same is true of magisterial authority. Suppose that an
official acting within his jurisdiction kills or flogs your slave. Clearly
he acts iure and cannot be said to be guilty of culpa.
The most important justification is self-defence, limited to force
necessary to prevent the harm anticipated to oneself. The justification
does not extend to measures which exceed this limit of proportionality,
as for instance blows struck in revenge. Self-defence is a species within
the wider genus ‘necessity’. In the case of self-defence the danger is
generated by the threat to one’s body. There is no need to distinguish
that from danger to one’s family. But danger to property is more
problematical. Am I justified in causing injury or damage in order to

18
Twelve Tables, ..
 DAMNUM INIURIA DATUM ( LOSS WRONGFULLY CAUSED ) 

protect my own property? The answer is that in appropriate circum-

stances I am, but the details of the picture are not clear. If the wind
blows my ship into your fishing nets so that extrication is otherwise
impossible I may cut your nets. May I knock down your house to stop
fire reaching mine? The answer is not clear. Perhaps the question is
whether, like the nets, your house was already certainly lost.
Finally, one curiosity. If I eat your food or burn your logs I am not
liable for damnum iniuria datum.19 The reason is that proper user is
regarded as rightful. It is specifically the rumpere which must happen
iniuria; and when the particular rumpere is exactly what the thing is for,
the classics hold that there is no wrongful breaking etc. It may be theft
and it may be a contempt-iniuria. But these other elements of wrong-
fulness do not allow the actual chewing, burning or what have you to
be regarded as wrongful for the purpose of this delict. One way of
putting this is to say that even if there is fault there is not faulty
chewing, swallowing and so on. The chewing is impeccable, despite
being by the wrong person. There are other examples: wearing, and
wearing out, someone else’s shoes, quarrying rocks from his hillside,
pressing his grapes when ripe to make wine.
There are two more sections to be considered under ‘the content of
iniuria’: (c) an alternative picture of the development, and (d) a comparison
between culpa and negligence. Both of these must be done rather briskly.

C. An alternative picture The key according to what has been said in the
previous section is an equivalence between iniuria and culpa: What
happens ‘unrightfully’ is what happens blameworthily or faultily. And
that equivalence is seen as something initial or, more accurately, some-
thing worked out from the beginning as the interpretation of iniuria
was built up. The other picture assumes that culpa was an innovation of
early classical law into a picture first settled on different lines.
On this view iniuria was first interpreted on the basis that all occidere,
rumpere etc. was prima facie wrongful. Within that assumption were
worked out certain situations in which there was a right to do the
killing or damage. And these situations were described somewhat
stiffly, not with the sensitivity to particular facts which characterises
the approach through culpa. Thus, it was right to kill a thief by night
and right whatever his mien, whether violent or submissive. And,

19
D.... (Paul,  On the Edict).
 DELICTS

perhaps, it was right to do whatever you wanted on your own land so

that you could not be liable for dropping a branch from your tree on to
my slave unless he was on a public way overhung by your branches.
The best evidence of this stiff, typified approach actually comes from
the wrong end of the lex’s history. Justinian’s Institutes seem to use it.
Thus at J..., if you are a soldier using a javelin on a military field and
you transfix a slave then, so long as you did not do it dolo malo, you are
free of culpa; otherwise if you are a civilian, or a soldier on other land.
This is stiff in that it ignores the careless soldier and the careful athlete.
At J... a similar analysis seems to be used of the pruner’s case, which
is made to turn on presence of a road or path below and apparently
rules out liability in any case in which there was no right to walk
beneath the tree.
In such a picture the rôle of culpa would have been to bring flexibility,
sensitivity to the particular facts, now relieving a liability (as of the
careful athlete), now extending a liability (as of the careless soldier).
There are two main difficulties. First, it is a picture which can hardly
accommodate the defence of accident. A typified notion of accident
would have to be expressed in terms of superior force. You escape
liability only if your precipitation into the plaintiff ’s res was due to an
overwhelming force, a whirlwind or a flood. Not an impossible approach
but not one which is securely evidenced in the texts. It is an approach
which has an additionally complicating element in that it runs very close
to the issue of causation making ‘it was an accident’ almost the same as
‘I did not do it’. But again that is not a decisive argument against it.
The other objection is that the most convincing evidence is of the
kind given in Justinian’s Institutes from the wrong end of the devel-
opment. This evidence can be explained as giving an unintentionally
misleading picture. It can happen that when one cites facts to illustrate a
proposition which dictates a different conclusion as the facts change—
that is to say, precisely when one’s aim is to illustrate the flexibility of
the concepts in question—that one’s illustrations can seem to be too
starkly opposed. Thus, I believe that Justinian’s Institutes fall into this
trap. They intend to show the culpa conclusion varying with the facts
but give the impression of excessive typification. In favour of this way
of understanding them is the fact that in the Digest’s versions of these
same cases there is no sign of the indiscriminate approach.20

20
D.... (Ulpian,  On the Edict); D... (Paul,  On Sabinus).
 DAMNUM INIURIA DATUM ( LOSS WRONGFULLY CAUSED ) 

The culpa question is, Was the defendant at fault on the particular
facts? The possibility exists, though I am not convinced, that at some
time the Romans dealt less sensitively with typical sets of facts, not
immediately with the actual facts which happened.

D. Culpa and negligence The Roman equivalent of the modern tort of

negligence is to be found within damnum iniuria datum, and in particular
within the culpa interpretation of iniuria. One might say that nine times
out of ten ‘culpa’ means ‘negligence’ in the sense of the modern tort.
However, it is dangerous to make that translation a habit. ‘Fault’ is
better, even though the fault is nearly always negligence.
The Roman delict comprises both intended and unintended harm.
That has an important impact on the meaning of culpa. Take the case of
one who wounds in self-defence. We cannot explain his non-liability
by saying that he is not guilty of malice (dolus) or negligence, but we
can say he is not liable because neither malicious nor at fault. The point
is, he is not guilty of negligence because he meant to harm and he is not
guilty of dolus because he meant to harm for a good motive. So the first
statement lacks explanatory force. The second says that he is free of
both bases of blame, malice and fault. Suppose he exceeds the measure
of proportionality. He is liable. Unless the excess is extreme I doubt
whether we can say that he is guilty of dolus but we can say he is at fault.
He was ‘unreasonable’ but not negligent. On these facts there is no
question of unreasonable failure to foresee. We cannot explain his
liability by saying he was negligent when he lunged with his sword.
He was at fault. The difference is between unreasonableness and
unreasonable failure to foresee. Suppose the case of a cruel teacher. If
he flogs the slave-apprentice for bad learning and the slave dies—that
is, something goes wrong and there is an unintended escalation of
consequence—you might analyse his liability as based on negligence.
But for the flogging itself, always supposing it is bad enough to inflict
damnum, negligence will not do. The master is conceded a right to
chastise. The easiest analysis of his liability is to say that his excessive
punishment is unreasonable. He is guilty of ‘fault’, what good men
condemn, even though on the facts there is no question of his failing to
foresee what they foresee.
Even in cases in which culpa does refer to fault in relation to
unintentional harm and where it does bear the sense of modern
‘negligence’ there are technical differences. First, modern law is always
concerned to ask whether there was on the facts any duty of care, and
 DELICTS

Roman law does not worry about that question. The reason is
that Roman law limits itself to physical damage corpori (res corrupta) corpore
(done by his bodily force). Excursions beyond that are controlled by
praetorian discretion. The modern duty of care is a controlling mech-
anism to prevent liability running wild. It is needed because there is in
principle no limitation to damage to physical property. If there were,
the duty to take care to avoid such harm could be taken for granted in
every case. Secondly, there is no habitual reliance on the reasonable
man. Whether there was or was not culpa was a question for the iudex.
He had what might be described as negative assistance from the corpus
of juristic utterance. It was not an exonerating factor that one was
unintelligent nor uninstructed in a skill. But the texts do not contain
evidence of a positive test in habitual use. And one feels the lack.
‘Reasonableness’ though uncertain in content is an invaluable guide.

. The Praetorian Periphery

Juristic interpretation of the statute sometimes produced results which
failed to satisfy plaintiff demand. Such demand is, of course, not
everything. It does not in itself justify reform. But if in addition the
praetor could be persuaded that, judged by principle higher than the
words of the lex, the particular stopping point was or had become
unnecessary, he would allow the plaintiff a remedy. In time these
praetorian extensions of the statutory core settled down, in such a
way that a plaintiff would not have to regard himself as a pioneer
breaking new and uncertain ground.
The praetorian periphery can most easily be surveyed through the
five questions used to consider the statutory core.

i. Has the plaintiff suffered loss (damnum)?

Here there is no extension. A plaintiff who had not suffered loss would
be out of range not only of the lex but of the principle, however stated,
which the lex reflected. Praetorian extension of this liability is essen-
tially a matter of giving remedies for more kinds of damnum.

ii. Did that loss arise from a thing spoiled (res corrupta)?
Here there are extensions. It is an area which raises the issue which we
refer to as ‘economic loss’, the central problem of which is that there is
 DAMNUM INIURIA DATUM ( LOSS WRONGFULLY CAUSED ) 

a danger of ramifying and unlimited liabilities and hence a need to go

carefully.
It is necessary to distinguish two types of case. First there is the
disaster in which the absence of spoiling is no more than technical or
even freakish. There is something tantamount to damage or destruc-
tion. Coins have gone down the drain irrecoverably. A diamond has
been thrown over the side of a ship. An animal has been allowed to
escape. These cases do not really raise the problems of ‘pure economic
loss’, for the damnum still arises from a necessarily limited and finite
event, the reduction (albeit not by spoiling) of the plaintiff ’s corporeal
wealth. This type of case has to be contrasted with the other, in which
the damnum consists solely in expenditure or loss of profits. Suppose that
I negligently spread a rumour that plague has broken out in the district.
A hundred or more substantial merchants evacuate their families and
close their businesses; a thousand or more suppliers to those merchants
lose contracts; and so on. Or suppose I give bad advice on investment
opportunities. There is a bubble; and then it bursts. Thousands who
relied on my tip come knocking at my door. This is pure economic
loss, damnum not anchored to diminution of corporeal wealth.
How far did the praetor go? There is no doubt that he would give an
action in the first case. There is a hint, perhaps not reliable, that Sabinus
was even willing to fudge the issue of res corrupta and allow the action
under the lex.21 It has been thought that at least Justinian extended this
extension so as to provide a general remedy for economic loss, but that
is doubtful.22 Classical law, one may be almost certain, did not venture
beyond the first kind of case. For economic loss caused by deliberate
trickery there would be an actio doli, for fraud. Otherwise it would be a
question of asking whether the plaintiff could make out a cause of
action in contract. Did he pay the defendant for advice so as to make it
locatio-conductio operis faciendi (investment opportunity to be analysed)?
Or was he commissioned by the defendant to lay out his money to
support the now failed business, so as to give him an actio mandati?

iii. Did the spoiled thing belong to the plaintiff (res actoris)?
We saw that the lex gave the damages only to the erus, i.e. to the
dominus ex iure Quiritium. The praetor allowed claims by others with

21
D.... (Ulpian,  On the Edict).
22
J... in fine.
 DELICTS

interests in the thing, the bona fide possessor, the usufructuary and the
pledge-creditor. Also, though his obstacle was more comprehensive,
the foreigner outside the ius civile was early given an action based on a
fiction of citizenship.
The praetor also gave an action for injury to a free man’s body, a
thing not owned, also for the injury or death to members of his family
in his paternal power (in potestate). In early law dominium and potestas
may hardly have been distinguished. One writer, J. M. Kelly, believes
that free children were intended to be within the lex itself.23

iv. Did the defendant do the spoiling ‘corpore suo’?

It is difficult to see how the jurists could have done away with this test
once adopted. What could they have put in its place? Yet there was
plainly a need to reach defendants who were responsible for bringing
damage about but had not actually done it. If I cause your slave to take
poison or put a log in his way so that he rides into it I should not escape just
because the statute happened to contemplate only the case in which I did
the killing or injuring. There is no need to repeat the earlier discussion.
The praetor did give actions for bringing about these consequences.
That is not the end of the difficulties. The restriction to corporeal
causation had the effect of putting a very tight lid on questions about
the infinite chain of causes. We all know that if we journey back in
time we must at all costs not step on so much as a butterfly, lest the
whole of history change. I cause a car to slow down as I dash across
Princes Street. If I had not done so it would not two minutes later have
been in position to run over a dog. Have I caused the dog’s death? One
way of handling this problem is to ask whether the death of the dog was
within the risk which I created by my dash. That kind of question is
one which moves the issue from causation to fault. Was I blameworthy
in respect of the dog’s death? The Romans must have followed that
line, putting the burden on culpa. But there is no discussion explicitly
on this problem. That is to say the texts do not directly address the issue
of remote causation which lies behind the ‘corpore suo’ test.

v. Did the defendant do the harm wrongfully (iniuria)?

This question is of the same order as the first, about damnum. You could
not well extend the liability to instances of harm done iure, rightfully.

23
J. M. Kelly, ‘The Meaning of the Lex Aquilia’,  LQR (), , –.
 DAMNUM INIURIA DATUM ( LOSS WRONGFULLY CAUSED ) 

And in fact no sorties were made by the praetor into the field of strict
liability; that is, liability for causing loss without culpa. So far as I know
the only trace of such liability in the field of this delict is the special
vicarious liability imposed on nautae, caupones and stabularii (keepers of
ships, inns and stables) for losses inflicted on users by members of their
staff. We will encounter this figure again under the heading of quasi-
delict.
Under these five questions we have considered the range of the
praetor’s satellite actions. A word must be added on his modes of
innovation and the vocabulary used to describe them.
A plaintiff who could bring his case within the statutory core was
said to claim by the actio directa. The word ‘directa’ has nothing to do
here with direct as opposed to indirect causation; it simply means
‘directed’ or ‘laid down’ or ‘established’. An actio directa, in any field,
is ‘an established action’. A plaintiff who could not bring the actio
directa, because his facts fell outside the statutory core and would
therefore not serve to substantiate the proposition advanced by the
established action, might urge the praetor to uphold a pleading differ-
ent from the actio directa. He might ask that the pleading should be
settled in any one of the ways familiar to the praetor as means of reform
and innovation. Thus, he might ask for a fiction to be inserted to knock
out a single requirement of the established action or he might ask for a
formula which simply recited the facts which he alleged to have hap-
pened. The one case would produce a formula drafted in ius, based on
the proposition of law in the actio directa, but ficticia, with a fiction. The
other, a formula drafted on the facts, on the event which had happened.
The formulae going on to the judge would thus be differently com-
posed. But whichever adapted pleading he wanted the plaintiff ’s actio
was said to be in factum. There is a difference between actio and formula
in this respect. The plaintiff ’s actio is in factum as soon as he bases himself
on his own story as opposed to the story covered by an actio in ius. But
the business of agere in factum might lead to a praetorian formula of more
than one pattern, depending on what was thought most suitable and
convenient.
Sadly we do not know enough about the Aquilian actions. In
particular we do not know the patterns of the formulae used by the
actiones in factum which extended the statutory delict. We do know that
a fiction was used to extend the liability to and for a non-citizen.
Whether anything other than formulae in factum were used in other
cases we cannot be sure.
 DELICTS

So far at the level of actio we have only two terms, actio directa and
actio in factum. You do not need any more vocabulary to cover the facts
on the ground. Yet there is one more term, actio utilis. This adds
nothing except an element of organisation. Thus, an actio utilis is an
actio in factum re-named to express its relationship to the actio directa, in
the following way. The spirit, higher principle or policy of the lex is
called its utilitas, its social usefulness, its expediency. An actio in factum
related to an actio directa in being dictated by the latter’s policy or spirit,
though outside its letter, could be called ‘a policy action’—meaning ‘a
policy-as-opposed-to-letter action’. We turn nouns easily enough into
adjectives: a policy action, a confrontation situation and so on. It would
be risky to change to the adjective and say ‘a politic action’ because
the adjective has its own specialised senses. If we did switch to ‘politic’
we would have to keep warning ourselves that it meant ‘policy-
motivated’, not ‘astute in a manner befitting a statesman’. The Romans
did switch to the adjective. The ‘utilitas-motivated’ actions were actiones
utiles. But utilis must not be translated as ‘useful’ or ‘usable’ or ‘expe-
dient’. An actio utilis is an actio in factum upheld to implement the utilitas,
the general policy or principle, of the lex.
This discussion, which suppresses a good deal of doubt and contro-
versy, assumes that the two terms utilis and in factum are virtually
synonymous. Not quite. For it leaves the possibility of there being an
actio in factum which is so to say a spontaneous creation, not related to
the utilitas of any other action but simply needed in its own right. With
that in mind we can approach the difficult lines with which Justinian’s
Institutes close their discussion of the lex Aquilia, at J...:

But it has been held that there is only an action under the lex if someone has
caused loss immediately with his body (corpore suo). And so if one has caused
loss in another way the practice is to give utiles actiones. [Examples are then
given of loss arising other than by the defendant’s bodily force.] If the loss
was not caused corpore and in addition no corpus has been damaged but loss
has in some other way affected someone, then, since neither the directa nor
the utilis Aquilian action lies, it has been held that the person responsible is
bound by an actio in factum. For example, if someone moved by pity frees
someone else’s slave from chains so that he escapes.

Here the contrast between the actio utilis and the actio in factum is
puzzling. Nowhere else is a line drawn explicitly between them. There
are two possible explanations. One is that Justinian is using language
 DAMNUM INIURIA DATUM ( LOSS WRONGFULLY CAUSED ) 

which by his time was old-fashioned, without much regard for earlier
usage, simply to convey the idea of an extension upon an extension.
The other is that the idea is deliberately meant to be conveyed that the
sphere of the lex is damage leading to loss, not simply loss. When
phrasing the policy or utilitas of any rule one has to pick one’s level of
generality. It would be possible to say that the general principle of the
lex was that people should have a remedy for wrongful loss; but no less
sensible to affirm, at a slightly lower level, that its principle was that one
should have a remedy for wrongful damage. If one pitched the level of
generality at damage, actions for loss sine laesione corporis, without
physical harm, would be set outside the statute’s utilitas so described.
They would then be actiones in factum without being actiones utiles. That
may be what Justinian meant.
The reason for pausing on actiones in factum and actiones utiles is that,
though specially well illustrated in the Aquilian field, they have a wider
importance. They give us an opportunity to see how the praetor
worked and how the jurists organised the product of his innovative
practice. Actio in factum expresses the potentiality of the system for
growth outside and between actiones directae. Actio utilis is a system-
atising notion whose effect is to enable jurists to create compound
molecules centred on direct actions.

. The Measure of Recovery

This is the last major topic. Its place offends against the system used so
far in that we have left the statutory core and must now return to it. We
know very little about the measure of recovery under the praetorian
actions, though we are entitled to assume that it was the same as in the
actio directa, at least unless there are arguments to the contrary.

i. Lis crescit (the suit enlarges)

Under both main chapters the action gave double damages against a
defendant who contested his liability. That is a fierce rule which
probably has a historical explanation: in the legis actio system of pro-
cedure the defendant would have been subject to manus iniectio, phys-
ical seizure against which there was no defence unless a vindex came
forward to throw off the claimant’s hand. The third party defender
 DELICTS

brought the action on himself, for double the damages. Hence a vindex
would not lightly take up the defence.
If the defendant admitted liability the action was called confessoria.
You might think that there would be nothing to litigate about if the
defendant confessed. But in the actio confessoria he would still contest the
measure; and without incurring the doubling he was allowed to show
that the event had not happened at all. The slave supposed to have
been killed was actually alive and well. Confession precluded only
argument on the issues of liability, corpore suo and iniuria.
It has been suggested, though the matter is not clear, that there was
no doubling of damages in the praetorian actions.

ii. The original measure

Doubling does not tell us the unit to be doubled. What did the statute
intend to be the simplum? Under ch. I the answer is straightforward.
The original measure was the highest value of the now dead res during
the preceding year. The words of the lex are clear. The condemnation
is to be plurimi, for the greatest value, in eo anno, in that year. The
retrospective calculation is best explained as a means of overcoming
seasonal fluctuations. It was a technique which would eliminate doubts
in the plaintiff ’s favour, leaning against the defendant because of his
culpa. Sometimes the logic produced odd results, which were, how-
ever, not rejected. Suppose you killed my slave. Six months earlier he
had lost a leg. You pay the value he had before he was crippled, and
I recover more by far than you have caused me loss.
The measure under ch. III causes endless trouble. We know two
things for certain. First, there was no mention of ‘highest’ value,
because it was on the authority of Sabinus that the word ‘plurimi’ was
implied into ch. III to balance its presence in ch. I. Gaius reports this at
G... Secondly, the period of time mentioned was not a year but
thirty days.
Daube argues that the measure was meant to work quite differently
from that under ch. I. What was intended was that there should be a
delay of thirty days to see how the matter developed.24 The defendant
could get the loss as calculated after a month had passed. The point was
that wounds need to be waited for in that way, being unpredictable.
The words in Latin will more than bear that interpretation: in diebus

24
Daube (n.), ff.
 DAMNUM INIURIA DATUM ( LOSS WRONGFULLY CAUSED ) 

triginta proximis is, literally, ‘in the nearest thirty days’, which could be
prospective. And it is not impossible that the verb was erit, in the future:
‘as much as ea res shall be worth over the next XXX days.’
This is attractive. The measure is very suitable for wounds, not crazy
for other kinds of damage (though Daube himself thought inanimate
damage was not at first included).25 The trouble is that the ch. III
provision does seem to have been interpreted symmetrically with ch. I
at least from Sabinus onwards. That is, at least from the beginning of
the classical period. The XXX days were thrown backwards, and
plurimi was read in. Why would it have been turned round? It was
better as it was. If it started with a prospective delay, there was no
obvious need for the change.
Jolowicz thought the measure had from the start been substantially
the same as for ch. I, the value of the thing in the previous month.26
That really is a mad measure for wounding and other partial destruc-
tion: full value for a cut or a chip, even if the depreciation was slight. So
Jolowicz said the early substantive scope of the chapter was originally
confined to destruction of non-ch. I objects.
Kelly suggested that the XXX days was not a period for valuation
but a period for payment, later converted into a period for valuation on
the model of ch. I.27 That is very attractive. I incline to the view that it
may be right. The lex may have wanted the loss to be paid within one
month, the means being ea res, the matter, i.e. the loss-causing disaster:
let him pay its value within a month.
These difficulties are interesting and probably insoluble. It probably
remains true that all the classics took the ch. III provision to be
retrospective, working, with the help of Sabinus’s plurimi, exactly as
that of ch. I except with XXX () instead of CCCLXV () days.

iii. Full value under chapter III?

The uncertainties of the very early period oblige us to accept a base for
ch. III some  years after the enactment: highest value in XXX
retrospective days. Wounds were certainly included. Do we assume
that I recovered  per cent for a wound depreciating a slave or cow
by  per cent? Probably the ‘highest value’ calculation was only to

25
Daube (n.), –.
26
Jolowicz (n.), ff.
27
J. M. Kelly, ‘Further Reflections on the Lex Aquilia’, in Studi in Onore di Edoardo Volterra
(Milan: Giuffrè, ), , ff.
 DELICTS

provide a base from which to work. The penal nature of the lex cannot
explain such gross disparity. The safest assumption is that full value so
calculated was the prima facie measure of recovery, subject to a deduc-
tion of the injured res’s surviving value. You recover  but if he was
depreciated by only , you must be taken to have received  already.

iv. The measure in high classical law

There was evidently a move away from value and to interesse. That is,
away from asking how much the dead animal would have been worth
on the market, or the injured one depreciated in terms of its market
value before and after, and towards asking what the plaintiff ’s interest
was in the safety of his res. In the end it seems that the defendant was to
put him back so far as money could do it in statu quo ante, into his
position as it would have been without the delict. It is not easy to say
how this development was related to the words of the statute. And
there are difficulties about remote losses which the texts do not answer.
Most of what we know relates to ch. I.
There seem to have been two stages. In the first, one recovered the
basic sum, the highest value, in every case; but one might get more if
one could show that one had lost more. This is the ‘more but not less’
stage. You kill my slave. He was part of a team. I recover his value plus
the loss from the depreciation of the team as a whole. Or he was just
about to enter an inheritance which had been left to him. I can have
the value of that too, which would have been mine if he had been able
to enter. This is shortly described in G...
However, Ulpian seems to have gone further. He was apparently
willing to switch so completely to the interesse principle that he could
contemplate displacing the statutory measure altogether. If his interesse
was less than the value the plaintiff would therefore get less than his
statutory entitlement. That is a difficult way of putting it, but the steps
in the interpretation are no easier. For an example, suppose this case.
Among my slaves is X the illegitimate son of my cousin, Titius. Titius
dies. Under his will X is to be heir if I free him, in which case I shall
receive from the huge estate a legacy of ten times his market value. If
X fails to take the inheritance for any reason other than my refusal to
manumit, I am myself to be heir. You now negligently kill X after
Titius’ death but before I can manumit him. My interest in his survival
is zero since I shall acquire more by his death than the legacy, itself
greater than his market value. Retrospectively there was a time when
 DAMNUM INIURIA DATUM ( LOSS WRONGFULLY CAUSED ) 

(a) he had a calculable market value and (b) that value was enhanced by
the legacy attached to him. But Ulpian will give no award.28 A century
before, Julian would have allowed me to recover the market value as a
minimum beyond which I could not fall.
Details under ch. III are obscure but two observations are worth
making. First, consequential losses could sometimes be recovered by a
separate action. I cut a rope. The ship on the other end floats off and
hits a rock. There is an actio directa for the rope, an actio in factum (here
also undoubtedly utilis) for the ship. If you could, did you have to
divide in this way in classical law? I think the answer is that you did.
You could not recover the value of the ship just as part of your interesse
in the rope. This then raises a more alarming supplementary question:
what if the consequential loss was of a kind not independently action-
able? Because you injure my slave I fail to get my olives to market on a
day when there is still a short supply; on the next market day there is a
glut. So I lose profits. This question cannot be securely answered. It
would be odd to allow some consequential losses to be reached by the
actio directa while still insisting that those recoverable by actio in factum
should be separately pursued.
The second observation is that a measure of recovery based on
depreciation does obliquely reach many common consequential losses.
Suppose a horse injured. The value before the disaster was . A vet
will have to be called. The first item of depreciation from the stand-
point of a notional buyer contemplating the fresh injury is the vet’s bill,
say . Hence, the horse, as is, goes down to . Then there is an
imperfect recovery factor: if he gets better will he be as good as he was
at ? And a risk factor: will he pull through at all? The reductions
attributable to these factors will depend very much on the nature of the
injury. The essential point is that the vet’s bill enters into the depreci-
ation. If a plaintiff were allowed depreciation plus medical expenses he
would be getting double recovery with respect to that item. It is easier
to see with repair costs of inanimate objects. Suppose you break my
Grecian urn. One way to work out my loss is to say that I shall have to
pay out  to get it mended and that even then it will be worth  less
than the original . Total loss = . Another way is to say, looking at
the pieces, that the depreciation is , since a buyer would pay  for

28
D... pr. (Ulpian,  On the Edict).
 DELICTS

them as they are knowing that laying out a further  would give him
an object worth .
It is difficult to prove but this ‘depreciation’ approach seems to me to
fit the classical law best for ch. III. Was the res depreciated by the
disaster? You look at it from the standpoint of a notional buyer bidding
after the disaster. The plaintiff gets the difference between the highest
price in the preceding XXX days, and the price which the notional
buyer, anticipating repair work and so on, would give after the disaster
(presumably one should say, immediately after the disaster).
Notice two similar cases. I am fond of my slave and I get a doctor to
look at the black eye which you have given her. The bill is . But an
unsentimental buyer would have taken nothing off: ‘She’ll soon be
over that.’ No expenses anticipated, no risk factor, ergo no depreciation.
Hence I recover nothing. Ulpian is made to say the opposite at
D....,29 but the text survives also in the Collatio at .. There
he does deny recovery. Suppose therefore that I insist on getting
repaired something which you have broken very badly. Here there is
depreciation but I spend more than is reasonable. For the sake of the
friend who gave it I spend twice its value repairing a quite ordinary
cooking pot. What would Ulpian have awarded? The notional buyer
would have given me nothing on the ground that it would cost more
to repair than to replace. His advice would have been to throw the
pieces away. I think Ulpian and his predecessors would have given only
the market price, not the actual repair bill. In  the National Gallery
in London restored a painting by Bryan Organ of Diana Spencer. It was
said that the cost of making good the vandal’s attack was greater than
the cost of commissioning a new portrait. Let us accept the assumption
that a second commission was possible, at £x. It does not follow that
the Gallery unreasonably spent £x +  repairing the first portrait. Each
painting is unique. Getting another means getting a different item by
way of substitute. Possibly the value of the first version was £x and
the anticipated value of the repaired version still greater than £x + .
Sentimental expenses, which the market will not contemplate, are only
those which exceed the value which they can restore.

29
Ulpian,  On the Edict.