The restricted range of active harm-verbs

In Roman law liability for wrongful loss turned on the words of the lex Aquilia. The details of the text of the lex are notoriously controversial, and we can omit all that is not immediately relevant to the present discussion.

It would have been tempting to say that a man bums when, having formed the intention that a thing should be damaged by fire, he executes that intention by bringing that damage about. The Philistines, for example, were sure that they knew who had burned up their fields, even though Samson had used an ingeniously indirect method of causing their crops to be burned up. He had tied torches to the tails of one hundred and fifty pairs of foxes,

And when he had set the brands on fire, he let them go into the standing corn of the Philistines, and burnt up both the shocks, and also the standing com, with the vineyards and olives. Then the Philistines said, Who hath done this? And they answered, Samson, the son in law of the Timnite, because he had taken his wife, and given her to his companion. And the Philistines came up, and burnt her and her father with fire.^[60]

The basis of the finding that Samson had burned the crops, and of the horrible revenge, is his intent to bring about that result, coupled with action, however indirect, which successfully achieved his intended goal.

It is interesting that in the delict iniuria we find no requirement of corporeal causation. Part of the reason must lie in the intentionality of that delict, part in the immateriality of the interest conceived in developed law to be protected, the integrity of the personality, not the body.^[62] However, equally important is the fact that in cases of physical iniuriae no text relevant to the edictal action required that what the victim suffered be described by a harm-verb used in the active voice. The model formula used the passive: ‘Whereas Aulus Agerius was struck on the cheek... for as much as on that account it shall seem fair and good that Numerius Negidius should be condemned, etc.’^[63] These words would reach the case both of a Numerius who struck and a Numerius who acted as decoy so that another could strike.^[64] If the ellipsis concealed the words ‘by Numer­ius Negidius’ the passive would be neutralised, since ‘by’ has equivalent effect. Lenel thought that ‘a Numerio Negidio’ must have been included.^[65] One dissents with hesitation from an unequivocal opinion of the master. But these additional words are not necessary and are not warranted by any text.

When can it be said, without reference to intention, that a defendant has damaged or injured? I knock your slave into the river. He is a poor swimmer and not blessed with Horatius’ strength.

size=2 color=black face="Times New Roman">It is not in the nature of legal authority to jump in one bound to a clear formulation of the answer to such a difficult problem. What seems to have happened is that by individual rulings the jurists pushed at the limits of the central image of a person damaging a thing,^[67] as, for instance, by including damage inflicted by throwing something hard or sharp at the thing^[68] or the thing at something hard or sharp,^[69] by rubbing, pouring in or injecting harmful drugs,^[70] by pouring things not water-resistant into muddy water,^[71] or watery mud into such things,^[72] by knocking an object from a height to the damage of something or someone below,^[73] by letting slip a load with the same effect,^[74] or by allowing a vehicle to slip down a hill crushing people below.^[75] The case-law picture extended, of course, to examples on the other side of the line, such as offering poison for the victim to take,^[76] frightening a horse which threw the victim down,^[77] sending a slave up a tree or down a well, so that he fell to his death,^[78] or watching the wind carrying your fire to a neighbour’s farm.^[79] It has been suggested that after the end of the Republic the case-law began to narrow the limits of the statutory action.^[80] That would certainly be an under­standable response to the difficulty of handling the defendant’s argument that he had not done the killing, burning, and so on,^[81] but the evidence of a stricter case-law is not secure.^[82]

A single test of doing is the culmination.

At the end of the Republic Ofilius concluded in favour of an actio in factum for the case in which the defendant had excited a horse with the effect that its rider, the plaintiff’s slave, was precipitated into a river and thus met his death. This is the earliest documented actio in factum, except the one from which, in the same text, Ofilius himself analogises: the defendant lured the plaintiff’s slave into an ambush and another dis­patched him.^{Roman",serif;color:black'>[87]} Norr must be right in saying that Ofilius cannot but have understood and been able to articulate the difference between killing and furnishing the cause of death.^[88] It would have been impossible to keep the statutory action and the actio in factum in play without being able to express that difference. On the other hand there is no reason to suppose that Ofilius could have ventured the generalisation that the limits of doing were defined by the defendant’s bodily force. More probably, the actio in factum simply began as a necessary recourse to meet the case in which the jurist found himself obliged to accept, without any stable test of the matter, that the harm-verb could not be satisfied - the defendant should answer for the damage but could not on any view be said to have done the killing, burning, breaking, and so on.

That pattern of development seems to explain the evidence in the texts. However, although it is also compatible with almost all of Professor Norr’s study of the concepts and language of indirect causation, it differs very markedly from his account in one crucial respect. The view propoun­ded with great learning and sensitivity by Norr is very much more dependent on the words of the lex itself and, in particular, on the word used for killing in the first chapter, occidere. On the view advanced above it would not have made any difference at all if the word used for killing had been, say, necare or interficere so long only as the verb appeared in the active voice, thus creating a condition of liability that the defendant must have killed: if any one shall have wrongfully killed another’s slave or four-footed beast, etc. However, Professor Norr’s view is that the first chapter was about an event specific to the word occidere, narrower than killing. It was about killing-by-striking, which is what occidere originally meant.⁵³ When it was chosen by Aquilius, or taken over by him from an earlier statute, occidere will necessarily have brought its killing-by­striking meaning into the lex.⁵⁴

There are a number of difficulties with this theory. The first is that we know that the veteres, the republican jurists, dealt vigorously with a literalist interpretation of a different aspect of another chapter of the lex. The argument was that chapter 3 in using the verbs urere, frangere, rumpere had intended to deal only with three types of damage. This had nothing to do with the defendant’s side of the matter. It was concerned not with the doing but with what was done. Literally, the plaintiff must have suffered a disaster of a particular description, loss from something ustum,fractum or ruptum.

⁵³        Norr, Causa Mortis, 4-9. See further below, text to nn. 57-9 and text thereto.

⁵⁴        Norr, Causa Mortis, 121-4, 135-8.

⁵⁵        D.9.2.27.13 (Ulpian, 18 ad edictum).

bring the plaintiff prima facie within the lex if the disaster which he had suffered was loss arising from a res corrupta, something spoiled. In view of this dynamic attitude to the triplet in chapter 3, it is difficult to believe that the old jurists would have had much time for an argument based simply on the narrowest possible meaning of the word occidere. Less violence would have been done to occidere by the assertion that it was a general word for killing than to rumpere by treating it as though it were corrumpere.

Next, Gaius tells us unambiguously that the requirement of corporeal causation was common to both chapter 1 and chapter 3.^[89] Hence, if the requirement was grounded in the literal meaning of occidere, we have to believe, not only that the jurists sustained the argument for that literal meaning, but also that they carried it over to verbs to which it was entirely alien. Urere,frangere and rumpere do not in themselves contain any hint of a narrow meaning linked to striking or bodily force. It might be objected that occidere was also relevant under chapter 3 and could, so to say, have infected the other verbs, as in the killing of birds and bees and wild beasts (non-pecudes). But the answer to that is that, under chapter 3, there was no need to conduct the debate in terms of any particular killing-verb, for in the terms of the lex a killed goose was just another res rupta. So, though killing was sometimes a chapter 3 event, occidere was not itself any more in play than interficere or necare. The application of the corpore test to all statutory actions, not just to those of chapter 1, thus strongly suggests that the narrow causal test is rooted in the general problems of the active voice, not in specific shades of meaning to be discovered in the verbs chosen to describe the damage.

The third objection is that the specially narrow meaning of occidere is nowhere evidenced outside the law. The etymology, from caedere, does not in itself indicate a powerfully restricted meaning, any more than its survival in other compounds such as the Latin homicidium and the English ‘homicide’ serves to reveal their scope at any period of their use. Julian’s appeal to that etymology is much more likely to have been made in support ex post of the narrow Aquilian interpretation than as the reason for that narrow interpretation:

D.9.2.51 pr. (Julian, 86 digestorum)

In ordinary speech a person is said to have killed (occidisse) however he has brought about the cause of death; but under the lex Aquilia a person is only held to be liable when he has furnished the cause of death by the application of force and so to say with his hand (adhibita vi et quasi manu), the interpretation of the word evidently being taken from the words for smiting and slaughter (a caedendo et caede).

This passage rather carefully abstains from a direct assertion ‘‘occidere means, or once meant, killing-by-striking’. It asserts that the word has no natural restriction to particular kinds of killing. It then invokes the etymology in support of the restrictive Aquilian understanding of the word. That etymology is drawn from Valerius Flaccus via Festus, who says that opinions differ as to whether occidere and necare should be distinguished, on the ground that the one comes a caedendo and happens ictu, the other sine ictu.⁵¹ Etymology is a guide to nuance, not to usage. There is no indication that the etymology reflected an originally narrow usage, and the quidam who favoured the narrow view of occidere need only be the lawyers driven to it by the fact of its having been employed in the active voice in the lex Aquilia. Exactly the same can be said of Labeo’s definition from his exposition of the SC. Silanianum (AD 10), which, with its brutal provisions for the household of an owner murdered by a slave, required a restrictive construction.^{[90] [91]}

Literary sources suggest that occidere was always a very general word for killing. Professor Norr himself admits that already in Plautus the word is used to denote killings other than by striking, as, for example, by starvation. The word had already broken out of its killing-by-striking meaning by, he thinks, a metaphorical extension: it was used when a writer wanted to personify hunger or poison as attacking its victim as a murderer attacks with a knife.^[92] In this way the narrow and specific sense which supposedly gripped the lex Aquilia has to be pushed back into the dark period before Plautus. More room is made for it by casting doubt even on the 287 BC terminus post quern which most scholars accept and by supposing that the word may have been carried over from earlier leges.⁶⁰ In other words the text of the lex itself has to be advanced to the earliest possible date, to accommodate the hypothesis that, when the statute was enacted, occidere meant killing-by-striking.

The one secure fact which we know about the date of the lex is that it had been passed before the jurist Brutus died, since he gave an opinion on its scope. That means that it must have been passed by about 125 BC.^[93] Though most scholars do place it earlier, a case every bit as strong can be made for a date after the third Punic war, late in the 140s.^[94] In short the lex may anyhow have been passed after Plautus; even if it was passed before Plautus, there is no hard evidence that occidere ever meant killing- by-striking.

New Roman">This touches very little of Professor Nörr’s account of the concepts and language of causation in the ancient world. It only substitutes the general problems of the active voice as the agent effecting the consequences which he derives from a rigidly narrow original sense of occidere, and it elimi­nates the question why the draftsman should have chosen or adopted that narrow word. If what is said here is correct, nobody ever chose restrictive terminology. All that happened was that the draftsman happened to express the conditions of liability using the harm-verbs in the simple active voice. Room was thus left for ingenious arguments, often but not always against the merits; and those arguments drove the lawyers to say what facts would satisfy those active verbs and to recommend actions on the case where liability needed to be extended to those who had brought about harm but had not harmed. Actions on the case were given by the praetor. And, after a long period in which the matter lay in case-law illustration, the distinction between the statutory action and the actions on the case was generalised or summed up in the test of bodily force: the statutory action lay where the defendant had done the harm corpore suo, by his bodily force.

Let us underline two aspects of this story which turn out to be inter­linked. First, the period of development to the mature position is very long. Even taking the latest possible date of the lex Aquilia, three cen­turies elapse between it and Gaius’ 'corpore' test. Secondly, supplemen­tation precedes crystallisation of the difference between doing and causing to be done. The Roman development passes through these three phases: (1) before the time of Servius Sulpicius^[95] the statutory action stood alone, with no supplementation by actiones in factum·, (2) sup­plementation began, as, for instance, against one who led a slave to an ambush to be killed by another, or frightened a horse which bolted and killed its rider,^[96] for these were fact situations in which the defendant needed to be liable but no reasonable man could say of him that he had done the harm, only that he caused it to be done; (3) first by illustration and then by abstract generalisation the law mapped the frontier between the statutory action and the actio in factum. But this sequence contains a mystery. Why should it be more pressing to locate the frontier at stage 3, when it had become a line between available actions, than at stage 1, when it marked, if it could be found, the limit of liability? The answer must be that at stage 1 the system, younger and less able to formulate and insist on difficult artificialities, relied on mechanisms of evasion and post­ponement. These are more clearly visible in the English story.

In English law the two species of liability, for doing and causing to be done, also co-existed before anybody drew a precise line between them.^[97] In trespass the plaintiff affirmed that the defendant had with force and arms and against the king’s peace done some harm: ‘Thomas Weaver brought trespass against George Ward, and counted that on 18 October 1614 in the parish of Mary-le-Bow, London, with force and arms (namely with swords, clubs and knives), the defendant beat, wounded and ill- treated the plaintiff so that his life was despaired of, against the king’s peace.’^[98] Again: ‘One Henry T. brought his writ of trespass against a woman, and supposed by his writ that she had burnt his house in G, with force and arms.’^[99]

By contrast, the supplementary action on the case typically alleged that, within the context of a given duty, the defendant had so negligently conducted himself as to cause the plaintiff to be harmed. For example: If William Beaulieu [shall make you secure of pursuing his claim], put [by gage and safe pledges] Roger Finglam [that he be before our justices of the Bench, etc. to show] why, whereas according to the law and custom of our realm of England until now obtaining everyone in the same realm should keep and is bound to keep his fire safely and securely so that no damage befall any of his neighbours in any wise through his fire, the aforesaid Roger kept his fire at Carlyon so negligently that for want of due keeping of the aforesaid fire the goods and chattels of the selfsame William to the value of £40 (being in the houses there) and the aforesaid houses were then and there burned by the fire, to the damage of the selfsame William etc.⁶⁸

As we saw in the first part of this paper, it was not until the early eighteenth century that challenges by the defendant to the form of action, usually maintaining that a plaintiff who had brought trespass should have brought case, compelled the courts to articulate a test in terms of the elusive notion of ‘directness’ or ‘immediacy’.⁶⁹ That might be said to be the equivalent of Gaius’ enunciation of the test of bodily force. Trespass was thus crippled by the same inconveniently restrictive causal test as had cut back the Aquilian action. But liability as a whole was not so restricted, since supplementation, as at Rome, had preceded definition: ‘he harmed’ (trespass) had been extended by ‘he caused harm to be done’ (case). This supplementation had happened in two stages, indeed in two long- separated generations.

Until the late seventeenth or early eighteenth century, there was no secure action on the case which could operate as a general supplement to trespass. Early supplementation, going back to the end of the fourteenth century, provided actions on the case for special instances of causing to be done, most obviously for negligently causing damage to happen within the context of a contractual or near-contractual relationship. Apart from these examples, with assumpsit-based duties, there were some others, as, for instance, the action on the case for fire exemplified above, where the duty was alleged to rest on a custom of the realm.⁷⁰ This limited sup­plementation must have taken some pressure off trespass.⁷¹ However, if we think of trespass as making liable anyone at all who wrongfully did damage, it remained true that there was no similarly general action on the case with the potential for making anyone at all liable who wrongfully caused damage to be done. The gap thus left was less pronounced than it

⁶⁸        Beaulieu v. Finglam (1401) Y.B. Pas. 2 Hen.IV, fol. 18, pl. 6, printed in Baker and Milson, Sources, 557; cf. 559, 565.

⁶⁹        M. J. Prichard, ‘Scott v. Shepherd (1773) and the Emergence of the Tort of Negligence’, Selden Society Lecture for 1973 (London, 1976), 15.

⁷⁰        Above, text to n. 68.

⁷¹        Outside the relationship cases, cases which could not be fitted within trespass would not have been common, especially in the days before the concept of ‘directness’ was insisted upon. See Baker, Introduction, 463. might have been; it was undoubtedly reduced in size by keeping the limits of trespass - the ‘he did it’ claim - unexplored, the black box phenomenon discussed immediately below.

Pronounced or not, that was the gap that began to be filled in the late seventeenth century. Actions on the case were then brought, and escaped challenge, which alleged or assumed an omnipresent duty to take care not to cause damage to be done. A good example, much abbreviated in the skeleton which follows, is provided by Browne v. Davis (1706):^{[100] [101]} whereas William [Browne]... was possessed of a certain flat-bottomed boat... and... John Davis... his barge so negligently, carelessly and unskilfully managed and steered, that the said barge, for want of good and sufficient care and management... in and upon the said boat of the said William... then and there fell foul, and the said boat broke and sank.

The full potential of this kind of claim, assuming a general duty of care, was openly and officially realised only through the great cases of Williams v. Holland’^7, and Donoghue v. Stevenson.¹⁴ But as soon as such claims could safely be advanced,^[102] the general supplementation of trespass became a tactical reality for every individual plaintiff, in the sense that for the first time it was possible for any plaintiff troubled by possible limita­tions on trespass to contemplate turning instead to an action on the case, without having to ask himself whether he fitted within the special circum­stances and relationships which had previously hedged that alternative about. Moreover, as we have seen,^[103] this general action on the case for injury and damage had, as in Rome, the potential for handling even facts within the range of trespass itself. As in Rome after Servius Sulpicius in the late Republic,^[104] the definition of a trespass became a question about the choosing between two available actions, the younger of which, unless inhibited by a doctrine of subsidiarity, was capable of gobbling up all the work done by the older.

The matter would have come more quickly to the boil but for keeping the question of liability in a black box, the jury at nisi prius.™ The jury was unlikely to be sympathetic to technical arguments against the merits. If it thought that you had negligently caused the defendant’s leg to be broken, it would not respond to your attempt to insist that liability was only for breaking. Furthermore - a peculiarity of the English story - wide disparities between stereotyped allegations and the real facts revealed in evidence were routine: no jury would expect to hear the formal allegations proved to the letter, and the court would not after­wards look for disparities ‘with eagle’s eyes’.^{[105] size=2 color=black face="Times New Roman">[106]}

In Rome there was also a black box, not a jury but, in most cases, a single iudex who gave no reasons for his decision. We must presume that he too absorbed technicalities and, in effect, postponed their resolution. However, the Roman iudex was a somewhat less efficient mechanism for suppressing nice questions. Being highly educated and bearing a respon­sibility which was not shared with others, he was more vulnerable to ingenious argument. Also, in England before the eighteenth century there were few means of bringing the real facts before the eyes of legal authority, the court itself.^[107] In Rome, by contrast, litigants could arm themselves with responsa from jurists and seek thereby to constrain the judge.

In some cases English defendants did try to take the matter out of the black box by entering a special plea. The courts did not co-operate. On the contrary, they allowed the vigilant plaintiff an opportunity to win on a technicality of his own: the defendant should have taken the general issue. In Gibbons v. Pepper (1695)^[108] the action was trespass for assaulting, beating and wounding, in the usual form. The defendant’s special plea was that the horse he had been riding took fright and ran off out of control, running down the plaintiff. It was the horse that ran down the plaintiff, not the defendant, and, further, the defendant was not to blame for it. The court took the view that this matter should have been given in evidence under the general issue, thus essentially insisting that it be kept as an issue of fact for the jury and as such kept in the black box:

For if I ride upon a horse and John Style whips the horse so that he runs away with me and runs over any other person, he who whipped the horse is guilty of the battery, and not me. But if I, by spurring, was the cause of the accident, then I am guilty. In the same manner, if A takes the hand of B and with it strikes C, A is the trespasser and not B.^[109]

Gibbons n. Pepper thus shows the law resisting pressure to disentangle the causal question from the general issue. Had the defendant not tried a special plea, he would have argued to the jury, under the general issue, that (a) it was the horse which did the injury, not him, and (b) it was not his fault that the horse bolted. Baker speaks of the law’s consistent failure to separate issues of causation from issues of fault.^[110] The reason why the causation issues never appear to have been taken separately is probably that only when mixed with assertions of blamelessness would they have enough moral weight to encourage a defendant even to try a special plea. In all other cases such matter was completely hidden from the law, lost in the arguments on evidence to the jury.

Nevertheless, it cannot be pretended that there are no elements of unsolved mystery as to how the jury could so long be used to evade and postpone the separate consideration of the causal questions. The Year­books show that early defendants were well able to challenge the allega­tions of force and arms and breach of the peace.^[111] These allegations thus lost their natural force and became decorative words of court. You could not hope to escape liability by showing that you had not in fact used weapons or other violence. When these words had been denatured, there remained the harm-verbs. The conventional counts still said, discounting the decorative words, that the defendant had beaten, burned, and so on. It is not obvious why defendants should have had difficulty in taking the causal points inherent in the verbs, and some seem to have done so. In 1353, for example, a plaintiff brought trespass for killing a horse, and the defendant answered that he had taken it damage feasant, that it was so wild that he had had to tie it up, and that, being tied up, it later struggled

so much that it strangled itself.⁸⁵ The argument having gone off on another point, the plaintiff succeeded. But, seeing the distinction between killing and furnishing the cause of death so close to the surface, we have to admit that we have not found the full explanation why these questions were suppressed till the eighteenth century.

In 1368 a defendant to trespass for burning with force and arms was found not guilty after a jury had found specially that the fire had started suddenly in the defendant’s house without his knowledge and had spread to the plaintiff’s house.⁸⁶ It seems likely that he had pleaded, as in the next four cases of alleged burning with force and arms, that he had not done the burning (which was in addition accidental). In Richard Tucker v. Philip and Eve Smith (1359) the defendants said, ‘the house of the said Philip [the defendant] which was neighbouring on the house of the same Richard caught fire by mischance and was burned down so that the fire therefrom being blown by the wind to the aforesaid house of the aforesaid Richard came down on that house and thus it was burned by mischance, etc.’.⁸⁷ This is a plea that the forces of fire and wind did the burning, not the plaintiff, mixed with insinuations of guiltlessness. In Joan Havent v. Richard Ward (1364) the defendant said, ‘And as for the burning timber etc.... a fire broke out in the houses of the adjacent neighbours from an accidental cause and thus burned the aforesaid timber... without this, that the same Richard burned the aforesaid timber as the same Joan above complains.’⁸⁸

In William Cook v. William Hasard (1387) the defendant said that he was the plaintiff’s servant ordered to light a badly built oven: ‘And so he, as servant of the aforesaid William Cook and by his order, put fire into the aforesaid oven in order to heat the oven. And the fire suddenly came out from the back of the oven, the same William Hasard being entirely unaware of this, through various holes and cracks... and burned the goods and chattels and houses aforesaid.’⁸⁹ This put the matter down to the force of fire itself. In John Ellis v. Nicholas Angwin (1390) the defendant said, ‘And as for the burning the houses, etc.... on the aforesaid Friday, unknown to the same Nicholas and against his will, a fire suddenly arose by mischance in the houses of the same Nicholas and burned down his houses there so that on account of a great gust of wind the fire spread rapidly to the houses of the said John, and thus by

⁸⁵        Anon. (1353) Y.B. 27 Lib.Ass. pl. 64, printed in Baker and Milsom, Sources, 312-13; cf.

304 and 320, where a purely causal point might have been taken, as between burning and causing to be burned.

⁸⁶        Anon. (1368) Y.B. 42 Lib.Ass. pl. 9 (Essex Assizes), Baker and Milsom, Sources, 321.

⁸⁷        Select Cases of Trespass from the King’s Courts 1307-1329, ed. M. S. Arnold (1987) 103 Selden Society 402.

⁸⁸        Ibid., 402.         ⁸⁹ Ibid., 404-5.

mischance and against the will of the same Nicholas the harm befell the same John in this behalf; without this, that he burned the aforesaid houses as the aforesaid John by his writ and count supposes.’^[112] Here the conflag­ration of which the plaintiff complains is again attributed to the wind, and misfortune; the defendant did not do the burning and, in addition, the defendant was not to blame for the force which did do it.

These all seem to be cases in which the defendant plainly denies a ‘you burned’ claim by saying that, though the burning happened, he did not bum, and, not surprisingly, he adds that the disaster happened contra voluntatem suam or words to that effect. By that addition the point ceases to be technical and unmeritorious. Nevertheless, the addition should not obscure the fact that these defendants were primarily saying that they did not do the burning. Fire, having a tendency to proceed with and without the aid of the wind, is particularly apt to provoke that answer.

In fact the making of such answers directed to the act of burning seems to have demonstrated the imperative need for supplementation of the ‘he burned the house’ claim by a claim on the other pattern, ‘he negligently caused the house to be burned’. Fire was one case where trespass was fully supplemented by case before the end of the fourteenth century. There were actions on the case alleging a general duty to keep fire safe based directly on the custom of the realm.^[113] Later there was even a second species of action on the case for fire which dispensed with the custom of the realm and simply assumed a general duty of care,^[114] in the same way as does the collision example from Browne v. Davis above.^[115] This, therefore, is one area in which we should have expected to find early attempts to draw the line between burning and causing to be burned. We do not. The supplementation was seemingly too successful. There was no embattled defence of the ‘he burned’ claim and hence no attempt to define its proper sphere.

This evidence deepens the mystery why for the next three hundred years defendants in other forms of trespass did not succeed in taking the causal question out of the general issue. It is not quite sufficient to say simply that the law was so: the question whether you had done the harm had to be considered before the jury under a general plea of not guilty. Whatever the answer may be, the eighteenth century saw both a general supplemen­tation of trespass by case and an increase in litigation about loss arising from injury and damage, chiefly because of better roads and more traf­fic.⁹⁴ At the same time there was an improvement in the procedural mechanisms for bringing the real facts before the court. Through reserva­tions of points of law arising on the evidence and by taking special verdicts reciting the facts which happened rather than the conclusion to be drawn from them, it became possible to compel the court to examine the legal inferences supported by the real facts as opposed to the facts conventionally recited in the plaintiff’s count.^[116] When the facts were at last looked at by analytically minded jurists instead of lay juries, the distinction between doing and causing to be done necessarily crystallised along the same line as it had at Rome.