The physical element
The most obvious difficulty stemming from the adoption of a definition of furtum based on contrectatio, assuming that we should take this to connote an improper handling, was that there may have been cases which were generally regarded as furtum but where there was no physical touching.[136] What of the hypothetical case put by Buckland of the chasing of your hen into my hen-house?[137] What of the refusal to return a pledge or to hand over another’s swine which have been rescued from wolves?[138] What of the person who lends a seller false weights with the result that he obtains a lesser price for his goods?[139] In all of these situations the owner is deprived of his property, and it can hardly have made a difference whether or not the wrongdoer or his accomplice had fortuitously touched it.
If the essence of furtum had been simply a wrongful interference, it might have been possible to hold that there could be furtum of land.
This was Sabinus’ view[140] and others may have agreed with him, but the popular opinion that there could not be furtum of land came to prevail in classical law.[141] It might have been possible to explain this away on the grounds that the wrongful occupier of land[142] could not in the nature of things ever have had an animus furandi, assuming that this connoted an intention to take away. What, though, of the analogous case of the person who handles a chest with the intention of carrying it off but finds he is incapable of moving it? This, by definition, cannot be explained away on the grounds of lack of intention, yet Ulpian is clear that it cannot be furtum.[143]* Nor can it be explained as a crude rule that a man cannot be liable for the furtum of a thing too heavy for him to lift.[144] Apparently, and quite logically, he is not guilty of furtum of the chest on the straightforward grounds that he was not stealing it.The case of the heavy chest is just one instance of a more general problem raised by the Roman definition of contrectatio, the problem of the measure of damages for the theft of part of a whole.
In English law if I handle a purse with the intention of taking away some of its contents it does not really matter whether I am treated as stealing the purse-plus- contents or merely the money which it contained;[145] it is simply a question of how the charge is drafted. In Roman law the problem was acute, though, since the damages recoverable were assessed as twice the value of the thing stolen. It was in this context that the tension between handling and carrying off, contrectatio and asportatio, received its most substantial consideration.The question is discussed at some length by Ulpian, in D.47.2.21 pr.-8,61 a passage of no little complexity. The text may simply be irredeemably corrupt, but the fact that it is extremely difficult to make sense of the highly condensed argument is not in itself a reason for so concluding; what follows is based on the assumption that the text is substantially genuine, and that it makes substantial sense. If I follow Ulpian’s reasoning, his argument takes the form of an extended critique in four sections (pr., 1-4, 5-6, 7-8) of the notion (of Sabinus?) that the simple handling of the thing with the appropriate mental state is sufficient in itself to constitute furtum.
The first section - 21 pr. - introduces the problem, drawing attention to the tension between the idea of furtum as contrectatio and furtum as subreptiof2 The conflict is stated starkly. The man who takes a measure out of a whole heap of corn or a small quantity of wine from a barrel is held - presumably by Sabinus63 - to commit furtum of the whole, for he has committed a contrectatio of the whole; on the reasonable assumption that the heap is a universitas, this seems an unavoidable conclusion, indistinguishable from the analogy (cited by Ulpian) that a person who touches a slave’s ear must be regarded as touching the slave.
On the other hand, says Ulpian, the damages should be assessed solely by reference to the amount that he took.64 The conclusion that it is the taking away that is relevant is then supported by two analogous situations (‘Nam et... Sed etthat I am stili guilty of the offence: Smith, The Law of Theft (6th edn, London, 1989), para. 136.
61 Ulpian, 40 ad Sabinum. The inscription attributes the passage to Paul, but this seems an obvious mistake, and the emendation to Ulpian is almost universally admitted. 21.9 and 10 constitute a brief coda to what I take to be the main argument. For discussion of the text, see in particular K. Olivecrona, Three Essays in Roman Law (Lund, 1949), 43-51 and Thomas, Synteleia Arangio-Ruiz, 11.609-15.
62 ‘Volgaris est quaestio an is qui ex acervo frumenti modium sustulit totius rei furtum faciat an verum eius tantum quod abstulit. Ofilius totius acervi furem esse putat; nam et qui aurem alicuius tetigit inquit Trebatius totum eum videri tetigisse; proinde et qui dolium aperuit et inde parvum vini abstulit non tantum eius quod abstulit verum totius videtur fur esse. Sed verum est in tantum eos furti actione teneri quantum abstulerunt. Nam et si quis armarium quod tollere non poterat aperuerit, et omnes res quae in eo erant contrectaverit, atque ita discesserit, deinde reversus unam ex his abstulerit, et antequam se reciperet quo destinaverat deprehensus fuerit, eiusdem rei et manifestus et nec manifestus fur erit. Sed et qui segetem luce secat et contrectat eius quod sequente nocte asportans deprehenditur manifestus et nec manifestus fur est.’
I have followed Mommsen’s generally accepted emendation of the last sentence; for a different view, see MacCormack [1977] Acta Juridica 129, 135.
63 Albanese, 23 Annali Palermo 140 n.
126.64 I cannot accept Olivecrona’s view that this sentence (‘sed verum... abstulerunt’) is interpolated (Three Essays in Roman Law, 44 n. 3). On the contrary, the opposition between it and the preceding part of the text is essential to the logic of the argument...First, the person who handles all the contents of a chest[146] and goes away, who then returns and is caught in the act of carrying off one individual item; here he is a fur manifestus of the item he actually takes. But this might be distinguished from the initial situation of taking part of a heap, for here the separate items are clearly individual res. Hence the second, stronger, analogy: the person who cuts corn by day and goes away, who then returns by night and is caught carrying off part of it; again he is a fur manifestus of the part he carries off.[147] In both cases it is the carrying off rather than the simple handling that determines the liability of the wrongdoer.
The second section, 21.1-4,[148] explores the same problem from a different angle. A man deposits a bag containing twenty coins and then, by mistake of the depositee, receives back a bag containing thirty; on the assumption that he knows that the bag contains thirty, but believes that twenty of these are the coins originally deposited, he is held liable[149] for furtum of only ten.[150] We must take it that this is a straightforward deposit and not depositum irregulare, and that the depositor is wrong in believing that his own twenty are contained in the bag.[151] Here, then, is the case of the subreptio of the whole, but with the animus furandi only as to a part; liability is determined by the animus. Ulpian immediately puts the contrasting case: if I take copper believing it to be gold, for example, I am liable for what I actually take, and not for what I believe I am taking.
The problem is to reconcile these two decisions; this is done by considering closely the nature of the universitas. If I take two purses, one containing ten and the other twenty coins, believing that one belongs to me, then I am only liable for the one which I knew was not mine.[152] Hence, applying the reasoning back to 21.1, the bag containing thirty has to be treated as two logically distinct parts: the twenty to which I think I am entitled and the ten to which I do not.[153] In 21.4 the contrast is finally underlined between the universitas and the single res: if I take a cup, believing that the handle which is attached to it belongs to me,[154]1 am liable for the furtum of the whole cup.The function of 21.1-4 in the context of Ulpian’s argument can thus be seen. When dealing with furtum as a universitas, logic demands that it be treated not as a single entity, but in terms of its component parts. The implication for the resolution of the initial problem, taking the measure of com from the heap, is obvious: liability cannot simply be determined by considering the heap as a single thing.
The next section - 21,5-6[155] - returns to the initial problem, applying this reasoning to it. What if a man removes a sextarius of corn from a shipload, or a small quantity of wine from a ship’s hold? Working from the simple case of the removal of goods from a warehouse (horreum), which (he says) could hardly be furtum of the whole of the contents of the warehouse, Ulpian concludes that we should similarly not say that the abstracter of the corn or wine was liable for the furtum of the whole cargo.[156] We must see here three distinct questions, in increasing order of difficulty:
(a) the simple case of taking goods from a horreum·, this is not furtum of the whole, for it is easy to see that the individual items are obviously separate res.
(b) the removal of corn from the shipload;[157] following Pomponius, Ulpian might allow that the ownership of the corn remained vested in its original owners, if the mixing had been done without their consent,[158] so that the taking of the sextarius could be logically indistinguishable from the easy case of the removal of goods from the horreum.
(c) the removal of wine from the cistern; here the mixing of different persons’ wine would have destroyed the individual ownership, and the resultant mass would have been subject to common ownership even if the mixing had been done without the owners’ consent.[159] It could not therefore be treated as individual res.
Still, though, we would not treat this as furtum of the whole.It is this last case which marks the true equivalent of the original situation of 21 pr., and which is the logical parallel with the bag containing thirty coins dealt with in 21.1. Strengthening this parallel, Ulpian now uses exactly similar reasoning as that used to link 21.1 and 21.3 to analogise the wine in the hold of a ship to amphorae of wine in a farmer’s warehouse (apotheca[160]); just as a bag of thirty coins can be treated as two distinct entities of twenty and ten, so a shipful of wine can be treated as if it were distinct amphorae. Even if what is taken is part of a universitas, the wrongdoer’s liability is determined by what he removes.[161]
This leads on to the final section of Ulpian’s argument, 21.7-8.[162] He begins, summing up what he has said about the horreum and the apotheca, by making the seductively obvious point that entering a room with the intention of stealing does not in itself constitute furtumf2 Equally, if someone breaks open a chest which is too heavy to lift, he is liable for the furtum of such of the contents as he carries off, but not of the whole;83 breaking into a chest is essentially the same as breaking into a room. Similarly, though slightly less obviously, if he unstraps a bundle which is too heavy to lift, intentionally handles the contents and carries some of them off, this is furtum only of those that he takes;84 his liability is determined by what he removes.
With the final sentence of 21.8 Ulpian ties together his whole argument, although not in a way to make his meaning immediately clear:
Quod si totum vas tollere potuit, dicimus eum totius esse furem, licet solverit ut singulas vel quasdam tolleret: et ita et Sabinus ait.
But if he could take the whole receptacle, we say that he is thief of all, even though he opened it to take one or some items; and so says Sabinus.85
This seems logically indistinguishable from the two cases just dealt with, yet the opposite result is reached; in consequence even such a conservative scholar as Buckland has agreed in the conclusion that the text has suffered badly at the hands of compilers who misunderstood what Ulpian was saying.86 If we think simply in terms of 21.8 itself, there is no doubt at all that this last sentence looks very odd indeed. But if we consider it in terms of the whole argument beginning at 21 pr. then its force becomes clearer.
tamen totum involucrum tollere non potuerit, singularum rerum quas tulerit fur est, ceterarum non est. Quod si totum vas tollere potuit, dicimus eum totius esse furem, licet solverit ut singulas vel quasdam tolleret: et ita et Sabinus ait.’
82 See too Ulpian in D.47.2.39 pr. (41 ad Sabinum). The same point is made by Paul in P.S.2.31.35 (FIRA 2.356) and D.47.2.54 pr. (35 ad edictum). The text had been treated by several commentators as having been introduced from a wholly different context, but it is vigorously and convincingly defended by F. Raber, Grundlagen klassischer Injurien- anspriiche (Vienna, 1969), 152-9.
83 21 pr. (supra, p. 63). Thomas (Synteleia Arangio-Ruiz, 11.613) is clearly right to suggest that what is envisaged is the person who breaks open the box intending to ransack it, but I do not think it is necessary (or possible) to assume from the text that the box is so heavy that he could not have contemplated moving it.
84 Compare the analysis of D.47.2.22.1 (Paul, 9 ad Sabinum): if he handles the other things in the box with the intent to steal them, even if he does not carry them off, he commits furtum of them; if he handles them only to move them out of the way of the object of his theft, it is not furtum. Ulpian would presumably not disagree even with the first case: the difference is that for Paul there is liability if the things were handled furti faciendi causa, while for Ulpian the bundle was unstrapped simply ut contrectet the contents. If I understand him rightly, his whole point is that contrectatio alone is insufficient.
class=31 style='margin-left:0cm;text-indent:0cm'>85 The translation is from the Pennsylvania edition of the Digest.86 ‘Digest XLVII.2 and the Methods of the Compilers’ (1930) 10 T.v.R. 117, 131-2. To the same effect: Albanese, 23 Annali Palermo 156-9; Thomas, Synteleia Arangio-Ruiz, 11.613.
The starting-point is Sabinus’ assertion that the person who takes part of a universitas is liable for the furtum of the whole. Ulpian has cast doubt on this first (21.1-4) by showing that sometimes we must take into account the animus of the wrongdoer as determining not simply whether he stole but what he stole, and then (21.5-7) by showing that we cannot coherently conclude that a person who breaks into or breaks up a collection of objects which have been gathered together as a unity commits furtum of the whole collection. Having thus weakened Sabinus’ position by these flanking arguments, in 21.8 he attacks Sabinus head-on. The first two cases lead inexorably to the conclusion that the person who ransacks a container which he could carry off, but which he has no intention of taking, should only be liable for what he does in fact carry off. In starkly stating the contrary conclusion as Sabinus’ opinion, he is not politely referring to the view of an earlier jurist but brutally if implicitly pointing out that it is wrong. ‘Et ita et Sabinus ait’, with its emphatically repeated ‘et... et’, should be translated not simply as ‘Sabinus says the same’[163] or ‘And so says Sabinus’,[164] but rather as ‘But this is what Sabinus says.’ Modo suo, he teases out the weak point in the opponents’ position, and rips it to pieces.[165]
Even here, Ulpian does not resolve the tension between handling and carrying away. It cannot be resolved, since it is implicit in the opposition between the definition of furtum as contrectatio and the conception of furtum as taking away. Similarly, we can only point to the existence of the same tension in English law.
The English world ‘appropriation’ does not have the overtones of physical touching which are associated with the Roman contrectatio, and in consequence English law has avoided some of the problems which had to be faced by the Romans. It is able, for example, to impose liability for theft in a case where one person holds himself out as owner of another’s property and purports to sell it to a third party.[166] In addition, since English law does not have to calculate damages by reference to the value of the thing stolen, it has not had to deal with the difficult problem of the theft or part of a universitas. Notwithstanding this, it is generally recognised that the courts’ exegesis of the concept of appropriation has been regrettably unclear and that the law today is something of an impenetrable forest.[167]
At the root of the difficulties lies the apparent conflict between the decisions of the House of Lords in Lawrence v. Metropolitan Police Commissioner[168] and Burnside and Morris.[169] In the former case D, a taxi-driver, was held to have been rightly convicted of theft when he had taken money in excess of the lawful fare from the wallet of a passenger; it did not matter that the passenger had proffered the wallet and hence consented to the taking. In the latter, D was held to have been rightly convicted of the theft of a joint of meat in a supermarket when he replaced its original label with one showing a lesser price, and then tendered this price at the supermarket check-out. In the course of his elaborate discussion of the concept of appropriation, Lord Roskill stressed that it involved an act which was not expressly or impliedly authorised by the owner, ‘an act by way of adverse interference with or usurpation of [the owner’s] rights’.[170] If there is a single concept of ‘appropriation’,[171] then there is clear conflict between these two decisions. In Lawrence the conviction was upheld notwithstanding that the owner consented to the taking; in Morris an act was said to constitute an appropriation if and only if the owner did not consent to it.
Subsequent cases have striven to reconcile these decisions, but without signal success.[172] No attempt will be made to do so here; the logicchopping involved should have no part to play in the criminal law. What is more interesting from a comparative perspective is why English law has managed to get itself into such a mess. Behind the conflict, it is argued, there is the same tension as we have seen underlying the Roman law, the tension between the definition of the minimal conditions which need to be satisfied in order to ground a conviction and the basic conception of the offence.
Morris is clearly a case of minimal conditions. The question was not what offence had been committed, but whether an offence had (yet) been committed. In this context appropriation has the same function for the English lawyers as contrectatio had for the Romans. This question can arise in a range of situations: when D has been interrupted in an apparently theftuous course of conduct;[173] when D has undoubtedly stolen the goods by a series of acts beginning in England but only completed in a foreign jurisdiction;[174] when the question is of the liability of secondary parties who only became involved at a late stage of the theftuous course of conduct;[175] when D is charged with dealing with property which he has already dishonestly come by;[176] or where there is a question whether D has committed one single theft or a number of distinct thefts which are required to be charged separately.[177]
In this context there are a number of good reasons why appropriation should be limited to acts which are not authorised by the owner. If appropriation involves ‘an assumption of the rights of an owner’[178] it is difficult to give this any real meaning when D’s acts are precisely those which the owner has permitted; such acts do more to confirm the owner’s rights then to assume them.[179] If a theftuous course of conduct is held to involve a single act of appropriation, rather than a continuing act or a series of acts,[180] then it makes good sense to ensure that so far as possible the appropriation should be held to occur at a time when it is possible to prove the other elements of the offence; this is best achieved by requiring that D’s behaviour on its face raises the inference that he is stealing.[181] Moreover, as a matter of reality we do not normally regard someone as stealing (as opposed to preparing to steal) whose acts are apparently perfectly innocent.
A case like Lawrence is very different. Where D has actually taken P.s property (assuming that he has acted dishonestly and with intent permanently to deprive P), there is a clear case of criminality. There may, legitimately, be a question of what offence he has committed, but there is no room for any doubt whether he has committed an offence. It is necessary therefore to face up to two distinct questions: (a) whether D.s conduct prima facie falls within the ambit of the offence of theft; (b) if so, whether the fact that he was clearly guilty of the offence of obtaining property by deception106 should preclude a conviction of theft. Both the Court of Appeal and the House of Lords were (rightly) of the opinion that the answer to the second question was in the negative.107 Hence the only question for resolution is the first; and it is hard to disagree with the conclusion that this is a perfectly clear case of appropriation, of taking another person’s property, and that there is no warrant for reading into the statute a wholly artificial requirement that the taking be without the consent of the owner. The case is the English analogue of Buckland’s hypothetical Roman question of inducing the hen to enter the henhouse;108 if we are satisfied that D has wrongfully obtained P’s property, we do not need to look a great deal further into the definition of the offence. The same point is visible even more strongly in Dobson v. General Accident.109 Here the wrongdoer had deceived the owner of a watch and a diamond ring into parting with them in exchange for a worthless cheque; the question was whether this loss fell within the clause in P’s insurance policy covering ‘loss or damage caused by theft’. The obvious answer, reached by the Court of Appeal, is that it did; as a matter of the construction of an insurance contract, anything else would have been absurd. Similarly in two further recent cases, Shuck110 and Rader,111 the Court of Appeal has upheld convictions of theft where D has obtained -
that D’s state of mind is one factor which determines whether his external acts constitute an appropriation.
106 Theft Act 1968, s. 15(1).
[1971] 1 Q.B. 373 at 377-8; [1972] A.C. 626 at 633. Equally the fact that he was no doubt guilty of a number of statutory offences was irrelevant. The Theft Acts are full of apparently overlapping offences; if they were held to be wholly distinct it would be all too easy for a person to escape conviction for one offence by arguing that he was guilty of another, perhaps more serious, one. See P. R. Glazebrook, ‘Thief or Swindler: Who Cares?’ [1991] C.L.J. 389.
108 Supra, n. 52.
109 [1990] 1 Q.B. 274.
110 [1992] Crim.L.R. 209.
[1992] Crim.L.R. 209.
taken - P’s property; it has been held not to matter that D’s taking has merely been the opposite side of the coin to P’s giving.[182]
The courts have in general attempted to analyse all these cases as if they dealt with the same problem, whether an appropriation can be said to have occurred when the owner of the goods consented to D’s acts. When faced with a statutory definition of theft couched in terms of appropriation it is hard to see that they could have done othewise; yet in truth the two questions - whether an offence has been committed and what offence has been committed - are distinct, and it is by no means obvious that the answer to the first question should also provide the answer to the second. Roman law provides no easy answer to the English lawyers’ problem. Nor should we expect it to do so. What can be learned, though, is that the problem stems not from the nature of the definition of theft but from the fact that it is being used to fulfil two functions. Ulpian’s demonstration of the inadequacy of framing the question of liability simply in terms of contrectatio should lead us to recognise the parallel inadequacy of asking simply whether and when there has been an appropriation. The apparently conflicting answers of Lawrence and Morris can be reconciled; but only if we realise that they are answers to different questions.