Is Possession Factual or Legal? Simon Douglas* In An Introduction to Roman Law Nicholas explains that it is both a question of fact and a question of law whether a person can be considered to be in possession of a thing: [P]ossession was not a fact if by that one means that it was unregulated by law. In the case, for example, of my taking possession of a farm, whether I have entered on the land is indeed a question of fact, but whether such an entry, assuming it to have occurred, amounts to a taking of possession is a matter governed by legal rules.1 The notion that ‘possession’ is somehow both a matter of fact and a matter of law is often repeated in respect of the common law as well. It can be seen in the common practice of qualifying the word by using terms such as ‘legal possession’, ‘constructive possession’, ‘deemed possession and ‘possession in the eyes of the law’.2 Writing in this vein, for instance, Clarke and Kohler state, ‘A person who takes physical control of land or goods, with the intention of excluding all others from it or them, acquires possession of it or them as a matter of law.’3 It can also be seen in the other common practice of distinguishing between using the word to denote a fact and using it to describe a legal concept. Pollock and Wright adopt this approach, drawing a distinction between using ‘possession’ to denote ‘… an actual relation between a person and a thing,’ and using it to describe ‘… the state of being possessor in the eyes of the law.’4 Whilst ‘possession’ may describe a factual state of affairs, these various approaches to possession each suggest that its meaning, at least in part, is defined by law. This chapter aims to explore and ultimately defend this characterisation of possession. In the first section we will see that lawyers normally use the word ‘possession’ to denote a factual state of affairs. The proof of this is that any * Jesus College, Oxford. I am very grateful to Jeffrey Hackney and Peter Mirfield for their comments on this article. 1 B Nicholas, An Introduction to Roman Law (OUP, Oxford 1962) 115. 2 Examples of these statements of legion. To give a few examples, see: S Green and J Randall, The Tort of Conversion (Hart, Oxford 2009) 86; M Wonnacott, Possession of Land (CUP, Cambridge 2006) 13; A Bell, The Modern Law of Personal Property in England and Ireland (Butterworths, London 1989) 34. For an excellent account of the different formulations of possession, see J Harris, ‘The Concept of Possession in English Law’ in A Guest (ed), Oxford Essays in Jurisprudence (Clarendon Press, Oxford 1961). 3 A Clarke & P Kohler, Property Law: Commentary and Materials (CUP, Cambridge 2009) 259 (emphasis added). See also M Bridge, Personal Property Law (OUP, Oxford 2002) 51. 4 F Pollock & R Wright, An Essay on Possession in the Common Law (Clarendon Press, Oxford 1888) 26. See also S Green and J Randall, The Tort of Conversion (n 2) and M Wonnacott, Possession of Land (n 2). litigant wishing to rely on a legal rule which makes reference to ‘possession’ will usually have to introduce evidence of possession. This requirement for evidence, it will be argued, means that ‘possession’ is being used to describe real world facts. Notwithstanding this conclusion, the second section of this chapter aims to show that ‘possession’ is also a technical term. The reason for this is that the legal usage of the word often diverges from its ordinary meaning. Consequently, one must be familiar with the relevant rules of the legal system in order accurately to state that the fact of possession exists. In this way, ‘possession’ can be understood to be both factual and legal. A. Possession as a Fact The first argument made by this chapter is that the statement, ‘X has possession of a thing’, when made by a lawyer, is normally a statement of fact. In order to make this argument it is important to begin by considering the differences between statements of fact and statements of law. 1. Statements of fact and statements of law In Definition and Theory in Jurisprudence5 HLA Hart sought to draw a distinction between legal words and statements and non-legal words and statements. Hart argued that ordinary, non-legal, words are those that have a ‘straightforward connection with counterparts in the world of fact …’6 Words such as ‘dog’, ‘cat’, ‘house’, ‘happy’, ‘sad’ etc, all have ‘persons, things, qualities, events, and processes, material or psychological’7 that directly correspond with such words. One can see and observe a dog in the real world and, pointing to it, can accurately state “That’s what a ‘dog’ is.” By contrast legal words and statements do not have factual counterparts. Take, for instance, the word ‘right’, which may be found in a statement by a judge such as, ‘X holds a right against Y’. As Hart explains: There is nothing which simply ‘corresponds’ to these legal words and when we try to define them we find the expressions we tender … are never precisely the equivalent of these legal words … Though one who has a right usually has some ‘expectation’ or ‘power’, the expression ‘a right’ is not synonymous with words like ‘expectation’ or ‘power’ …8 Unlike the word ‘dog’, there is nothing one can see and observe in the real world, and say, “That’s what X’s ‘right’ is.” Rather than denoting a fact, Hart argued that the statement ‘X holds a right against Y’ states a legal conclusion. The essence of H Hart, ‘Definition and Theory in Jurisprudence’ (1954) 70 LQR 37. ibid 38. 7 ibid. 8 ibid. 5 6 this is that the judge, in making this statement, will be applying some legal rule to the facts in the case. Let us say that it has been proved that Y had promised to pay X a sum of money in return for a service performed by X. When the judge states that ‘X holds a right against Y’, he is applying an unstated rule (in this example, a rule governing the formation of a contract) to the facts of the case. Understood in this way, the statement ‘X holds a right against Y’ is the ‘tail-end’ of this reasoning: it states the conclusion that the contractual rule applies to the facts in the case. Legal words, such as ‘right’ and ‘duty’, do not have factual counterparts, but are ways of stating these legal conclusions. Turning to the word ‘possession’, we are essentially asking whether ‘possession’ is like the word ‘dog’ or the word ‘right’: does it have a factual counterpart? When ‘possession’ is used by a non-lawyer it is clear that the word does have a factual counterpart. It is used to describe a person’s relationship with a thing, particularly where that relationship is characterised by a high degree of physical and mental control over a thing.9 The important question is whether lawyers also use it in this way. When counsel or a judge uses the word ‘possession’, do they use it, as non-lawyers do, to denote its factual counterpart; or do they use it, in Hart’s words, as a way of stating a legal conclusion? There is much to suggest either view. On the one hand, ‘possession’ is current in ordinary non-legal usage, and there may be no reason for lawyers to use the word in a different way. As Simpson warns, ‘One must not forget that lawyers are people, and legal English English.’10 On the other hand, possession has attracted a substantial amount of academic attention, with a number of full length treatises11 and many chapters in property law textbooks which appear to set out rules governing the acquisition, loss and protection of possession. When a judge states that ‘X has possession of a thing’, it may be that this is a legal conclusion reached by applying rules and doctrines found in these materials to the facts of the particular case. This ambiguity surrounding ‘possession’ means that we need a way of testing whether the statement, ‘X has possession of a thing’, when made in a legal context, is a statement of fact or a statement of legal conclusion. To this end the following test is suggested: if a judge12 reaches the conclusion that ‘X has possession of a thing’ solely by drawing inferences from evidence presented to him, then the statement is a statement of fact. The Oxford English Dictionary defines ‘possession’ as ‘the state of possessing something … visible power or control’: J Simpson and E Weiner (eds), The Oxford English Dictionary, 2nd ed (Oxford, OUP, 1989). This is similar in terms to the definition put forward by Holmes: ‘To gain possession, then, a man must stand in a certain physical relation to the object and to the rest of the world and must have a certain intent. These relations and this intent are the facts of which we are in search.’ See O W Holmes, The Common Law (Little, Brown & Co, Boston 1881) 216. 10 AWB Simpson, ‘The Analysis of Legal Concepts’ (1964) 80 LQR 535, 547. 11 The two historical examples of such treatises are F von Savingy (trans P Erskine), Treatise on Possession: Or, the Jus Possessionis of the Civil Law (6th ed, Lawbook Exchange, London 2003) and F Pollock & R Wright, An Essay on Possession in the Common Law (Clarendon Press, Oxford 1888). For more recent examples, see R Hickey, Property and the Law of Finders (Hart, Oxford 2010) and M Wonnacott, Possession of Land (n 2). 12 Because this chapter focuses on the usage of the word ‘possession’ in judgments, the focus will be on judicial usage of the word. However, the test proposed here could, in principle, be applied to any lawyer (counsel, solicitor, academic etc) using the word. 9 Central to this proposed test for whether ‘possession’ is used by lawyers to denote a fact is the relevance (or irrelevance) of evidence. ‘Evidence …’, as Dennis states, ‘… is information. It is information that provides grounds for belief that a particular fact or set of facts is true.’13 The process of collecting evidence is a process of finding out something about the world. If a crime is suspected of being committed, for example, evidence is collected when ‘… investigators go out into the field, ask questions, poke around, interview witnesses, examine physical evidence …’14 There are other common ways of collecting such information. Forensic scientists conduct experiments in the laboratory, social scientists collect and analyse empirical data, journalists and historians investigate records, transcripts, physical artefacts and any number of other materials. The purpose of collecting information in these ways is that it allows the existence or nonexistence of facts to be ‘… established by drawing conclusions (inferences) from information that the inquirer has collected for that purpose.’ 15 Evidence, therefore, is any material that makes the existence or non-existence of a fact more or less likely. It is because evidence has this function that it is central to the test proposed here. If a party involved in a dispute is required by the court to introduce evidence of possession, then the court must be asking that party to introduce material that makes the existence or non-existence of a particular fact (namely a person’s relationship with a thing) more or less likely. If the court then reaches the conclusion that ‘possession’ exists by drawing an inference from this evidence, the court must be concluding that the fact of possession exists. Conversely, if a court, in reaching the conclusion that ‘possession’ exists, does not reach this conclusion by drawing inferences from the evidence presented to it, but by some other means (such as by applying a legal rule or doctrine to the facts of the case), then they are not using the word ‘possession’ to denote a fact, but to state a legal conclusion. The best indicator of whether possession is used by lawyers to denote a fact, therefore, is whether or not a party will be called on to introduce evidence to prove its existence. 2. Examples of factual possession When the test proposed above is applied to judicial uses of the word ‘possession’, it is clear that the word is used to denote a factual state of affairs. This is evident from the numerous contexts in which a party can be called on to introduce evidence of possession. For example, in criminal law if a defendant is charged with being in possession of a banned substance, the prosecution will have to introduce evidence of the defendant’s possession of the banned substance in order to establish criminal liability.16 To this end the prosecution may call a I Dennis, The Law of Evidence (4th ed, Sweet and Maxwell, London 2010) 3. F Schauer, Thinking Like a Lawyer (HUP, Cambridge MS 2009) 206. 15 Dennis (n 13) 4. 16 See R v Lambert [2002] 2 AC 545, where the House of Lords considered the meaning of ‘possession’ under s 5 of the Misuse of Drugs Act 1971, which provides: ‘… it shall not be lawful for a person to have a controlled drug in his possession.’ 13 14 police officer to give oral evidence that he found the banned substance in the defendant’s pocket; or they may introduce the findings of a forensic analysis which found traces of the banned substance on the defendant’s clothing. The purpose of introducing such material is that it makes the existence of a certain fact, namely the defendant’s physical and mental control over the banned substance, more likely. Consequently, when the word ‘possession’ is used in this context, i.e. when a judge states, ‘the defendant had possession of a banned substance’, the word ‘possession’ is being used to denote a fact. To properly demonstrate this thesis, that ‘possession’ is used by lawyers to denote a fact, it is important to consider its usage in the law of property. There are a number of rules and doctrines in the law of property which make reference to the word ‘possession’. In particular, possession is crucial to the rules governing the creation, transfer and destruction of property rights. The important role it plays in the law of property explains why much of the legal analysis of the word has been from a property law perspective.17 What this section must examine, therefore, is how the word is used in this context. As we will see, if a party to litigation wishes to rely on one of the property law rules which makes reference to ‘possession’ it will be incumbent on him to introduce evidence of possession. Consequently, when ‘possession’ is used in the context of these rules, it is used to denote a factual state of affairs. a) Creation of property rights In the famous case of Armory v Delamirie18 a ring, which had been lost by an unknown person, was found by the claimant, a chimney sweeper’s boy, who handed it to the defendant jeweller to be valued. The jeweller subsequently refused to hand the ring back to the claimant, arguing that he was not obliged to do so as the ring belonged to a third party. Pratt CJ, finding for the claimant, held that … the finder of a jewel, though he does not by such finding acquire an absolute property or ownership, yet he has such a property as will enable him to keep it against all but the rightful owner, and subsequently may maintain trover.19 According to Pratt CJ the chimney sweeper’s boy, by taking possession of the ring when he found it, had acquired a new property right in it. This was a ‘relative’ right, in that it would not bind those with a pre-existing property right in ring (i.e. the person who had lost it), but it would bind all others, including the defendant. The rule in Armory v Delamirie is sometimes referred to as a ‘finders rule’, suggesting that it only applies to those who have found lost things, as happened in the case.20 However, the rule is wider and applies to anyone who The monographs on possession noted above (n 11) all consider possession in the context of property law. 18 (1722) 1 Stra 505, 93 ER 664. 19 ibid 505, 664. 20 eg Parker v British Airways Board [1982] QB 1004, 1008. 17 has taken possession of a thing, regardless of the circumstances in which they took possession.21 An example is Costello v Chief Constable22 where the police had seized a car from the claimant on the suspicion that it was stolen. The police were unable to trace the owner of the car but they refused to return it to the claimant. Although it was proved on the balance of probabilities that the car had been stolen, suggesting that the claimant had not ‘found’ the car but had actually stolen it from another, his possession was nonetheless sufficient for him to take advantage of the rule in Armory v Delamirie. Consequently the court found for the claimant and ordered the police to return the car. In addition, the rule in Armory v Delamirie applies to land, something which cannot, in the ordinary sense, be lost and found. In Asher v Whitlock23, the leading case involving land, Cockburn CJ said: … I take it as clearly established, that possession is good against all the world except the person who can shew a good title; and it would be mischievous to change this established doctrine.24 This is almost an exact restatement of Pratt CJ’s dictum in Armory v Delamirie. It is clearly established rule, therefore, that by taking possession of a thing, whether it be a chattel or land, one will acquire a property right in that thing. It is suggested that when the word ‘possession’ is used in the context of this rule, it is being used to denote a fact, namely a person’s physical and mental control over a thing. The reason for this is that a litigant will be required to introduce evidence of his physical and mental control over the thing (or that of someone he has acquired his title from) if he relies on this rule of acquisition. Take the recent case of Mayor of London v Hall25 where the claimant brought an action in trespass against anti-war protesters who had occupied a square outside the houses of parliament. Because freehold title to the land is vested in the Crown by statute26, the claimant argued that he had relative title to the land (entitling him to sue the protesters in trespass) under the rule in Armory v Delamirie. To establish this it was incumbent on the claimant to introduce evidence of his possession of the land. He did this by showing that his agents, such as gardeners who kept the square and policemen who patrolled it, exercised a high degree of physical and mental control over the land. The Court of Appeal’s conclusion that the claimant, prior to the defendant’s protest, had ‘possession’ of the land, was a clear inference drawn from this evidence.27 Accordingly, ‘possession’ was used by the court to denote a factual state of affairs: the claimant’s physical and mental control over the square. It is useful to contrast Mayor of London v Hall with the much older case of Young v Hichens28 cf R Hickey’s chapter in this collection. Costello v Chief Constable of Derbyshire [2001] 1 WLR 1437. 23 (1865) LR 1 QB 1. 24 ibid 5. cf Carter v Barnard (1849) 13 QB 945, 116 ER 1524. 25 [2011] 1 WLR 504. 26 Greater London Authority Act 1999, s 384(1). 27 [2011] 1 WLR 504, 515. Despite establishing possession in the case, the court did not follow Asher v Whitlock because the facts were governed by s 384(3) of the Greater London Authority Act 1999 which, it was held, gave the claimant title to sue in trespass. 28 (1844) 6 QB 606, 115 ER 228. 21 22 where the claimant fisherman sought to rely on the Armory v Delamirie rule to establish that he had title to a number of pilchards. The court refused to apply the rule because the claimant’s evidence only showed that he had partially enclosed the pilchards with his fishing net when the defendant scared the fish away. The finding that the claimant did not have possession of fish was, again, based on inferences that the court could draw from the evidence presented to it. As explained in the previous section, evidence is any material that makes the existence or non-existence of a fact more or less likely. It follows that the requirement for a claimant to introduce evidence of possession in order to establish a property right under the rule in Armory v Delamirie means that ‘possession’, when used in this context, is being used to denote a factual state of affairs. b) Transferring property rights The rule in Armory v Delamirie involves the creation of a new and original property right. Once such a right exists it can then be transferred. Historically the principal method of transferring property rights was a delivery, or traditio. In the leading case of Cochrane v Moore Fry LJ said: The law recognised seisin as the common incident of all property in corporeal things, and tradition or the delivery of that seisin from one man to another as essential to the transfer of the property in that thing, whether it were land or a horse, and whether by way of sale or of gift, and whether by word of mouth or by deed under seal.29 Although it is no longer possible to transfer property rights in land by a delivery (known as a ‘livery of seisin’)30, it remains an important method of transfer for rights in chattels. Although a delivery normally consists of a straightforward handing over of the thing, the essence of this transaction is the transfer of possession from the transferor to the transferee, and this can happen in other ways. For example, in Thomas v Times Books Co Ltd31 the poet Dylan Thomas told his producer, one Cleverdon, that he had lost his manuscript of ‘Under Milk Wood’ and told Cleverdon that if he was able to find the manuscript then it would belong to him. Cleverdon eventually found the manuscript in a pub in Soho and it was held that this was sufficient to transfer title to him. Although there had not been a straightforward handing over of the manuscript, there had been a change in ‘possession’ of it, and this was sufficient to transfer title, as Plowman J said: The fact is that Cleverdon got possession of this manuscript from the Soho public-house in which it had been left by Dylan Thomas and that he got (1890) 24 QBD 57, 65-66. Real Property Act 1845, See generally AWB Simpson, A History of the Land Law (2nd ed, Clarendon Press, Oxford 1986) 278. 31 [1966] 1 WLR 911. 29 30 that possession with the consent of Dylan Thomas. That, in my judgment, is sufficient delivery to perfect a gift in Cleverdon's favour.32 A transfer of property rights by a delivery, therefore, consists primarily of a transfer of possession. Delivery provides us with another rule in property law that includes a reference to ‘possession’. It is suggested that when used in this context the word ‘possession’ denotes a fact. Again, the reason for this is that a litigant will be required to introduce evidence that the transferor has relinquished physical and mental control over the thing and that the transferee has in turn taken physical and mental control over a thing if he is to show that there has been a successful delivery. The case of re Cole33 provides an example. A husband had bought and furnished a house and had told his wife, who lived with him in the house, that everything belonged to her. When the husband went bankrupt his wife claimed that she owned the contents of the house (thus making them unavailable to her husband’s trustee in bankruptcy) because there had been an effective delivery of them to her. The argument failed. Although the wife was able to introduce evidence which showed that she had used the chattels within the house, it also showed that the husband continued to use them as well. This meant that the evidence did not show that her husband had relinquished physical and mental control over the chattels and that she, in turn, had taken physical and mental control.34 In other words, the claim failed because the claimant was unable to introduce evidence from which the court could infer that there had been a transfer of possession from the husband to the wife. This need for evidence in order to establish a transfer of ‘possession’ demonstrates that the word, when used in the context of this rule, is denoting a factual state of affairs. c) Destruction of property rights A third context in which we can find rules which make reference to ‘possession’ are those rules which govern the destruction of property rights. If the holder of a property right in a thing has possession of the thing, his possession often prevents the destruction of his property right. Conversely, when possession is in the hands of another, the owner’s property right is sometimes liable to destruction. An example is s. 2(1) of the Factors Act 1889 which states: Where a mercantile agent is, with the consent of the owner, in possession of goods or of the documents of title to goods, any sale, pledge, or other disposition of the goods, made by him when acting in the ordinary course of business of a mercantile agent, shall … be as valid as if he were expressly authorised by the owner of the goods to make the same … ibid 919. [1964] Ch 175. 34 ibid 190. 32 33 The effect of this rule is that if an owner of a thing entrusts possession of it to a mercantile agent, his property right will be destroyed by any disposition to a third party by the mercantile agent.35 Again, if a litigant relies on this rule it will be incumbent on him to introduce evidence of possession. This can be seen in the case of Beverley Acceptances Ltd v Oakley.36 The defendant was a mercantile agent of one Green, who purported to sell Green’s Rolls Royces to the claimant without Green’s consent. The claimant relied on s. 2(1) of the Factors Act 1889, saying that the defendant was a mercantile agent in possession of the Rolls Royces and, consequently, his disposition of the cars destroyed Green’s title to them. In order to make this argument they had to prove, as Donaldson LJ said, ‘(i) That [the defendant] was a mercantile agent; [and] (ii) that he was in possession of the Rolls Royces or of documents of title to those cars …’37 The claimant failed to prove the second limb as the evidence introduced into court showed that the cars were stationed within Green’s compound, rather than being in the defendant’s control. Donaldson LJ concluded that the claimant could ‘show no such possession’38 as required by the Act. In other words, the claimant could not rely on the rule in s 2(1) of the Factors Act 1889 as it was not able to introduce evidence of the defendant’s possession at the time of the disposition. This provides another illustration of a legal rule which requires a litigant to provide evidence of possession. Consequently, when the word ‘possession’ is used in the context of this rule, it is used to denote a fact. B. Possession as a Technical Term So far we have seen that when the word ‘possession’ is used by a lawyer, particularly in the context of the rules governing the creation, transfer and destruction of property rights, it is used to denote a fact. The proof of this is that any litigant wishing to rely on these rules will have to introduce evidence into court of possession. Despite this conclusion, the present section aims to show that ‘possession’ is also a ‘technical’ term. We will see that when ‘possession’ is used in a legal context, although it is used to denote a fact, its meaning is defined by rules and, consequently, the legal usage of the word often diverges from its ordinary meaning.39 Similar rules can be found in s 24 and 25 of the Sale of Goods Act 1979. The closest analogy in land law is that of postponing an interest of someone not in actual occupation of the land under s 27 and 29 of the Land Registration Act 2002. 36 [1982] RTR 417. 37 ibid 431. 38 ibid. 39 For a full account of ‘technical’ terms in legal usage, see M Morrison, ‘Excursions into the nature of legal language’ (1989) 37 Clev St L Rev 271. 35 1. Technical words In the previous section we applied Hart’s distinction between statements of law and statements of fact. In concluding that ‘possession’ falls into the latter category, meaning that it is used by lawyers to denote a fact, the word would be classed as an ‘ordinary’, or ‘non-legal’, word by Hart.40 Such a characterisation of ‘possession’ is problematic as it begs the question of why the word has been the subject of so much legal analysis. There is certainly a temptation to think that words that denote facts should not be made the subject of legal definition or analysis. The factual counterparts of such words, the ‘persons, things, qualities, events [or] processes’41 that these words correspond with, can be seen and observed by lawyers and non-lawyers alike. One does not need any legal training to see that something is a ‘dog’, or that someone is a ‘child’. It may be thought then that the statements ‘That is a dog’, and ‘He is a child’, are not technical as they can be made by, and understood by, lawyers and non-lawyers alike. Hart certainly takes this view by labelling such words as ‘ordinary words’. So on Hart’s thesis, such words do not have any distinctive legal meaning: because the factual counterparts of such words can be seen and observed by lawyers and nonlawyers alike, these words can be used by and understood by both groups of people. Returning to ‘possession’, the factual counterpart to this word, a person’s physical and mental control over a thing, can be seen and observed by lawyers and non-lawyers alike. There is, therefore, a temptation to think that both lawyers and non-lawyers will use the word in the same way to describe these facts. However, it would be wrong to reach this conclusion. In an important criticism of Hart’s thesis, Simpson explained that lawyers frequently use words to denote facts, yet ascribe meanings to these words which diverge from their ordinary meanings.42 Take the example of the word ‘agreement’. Although this word clearly has a factual counterpart, it is still the case that when lawyers use the word its meaning is also defined by rules. There are instances in which certain facts are shown to exist which a non-lawyer would call an ‘agreement’, yet a lawyer, cautious of the legal consequences that may follow from labeling those facts an ‘agreement’, must withhold that description.43 Conversely there are instances in which facts do not amount to an ‘agreement’ according to the ordinary meaning of the word, yet a lawyer would describe them in this way.44 The consequence of this is that whilst the meaning of the word ‘agreement’, when used in a legal context, overlaps to a great extent with its ordinary meaning, it also diverges from it in some important ways. This makes Hart (n 5) 38. 38 42 AWB Simpson, ‘The Analysis of Legal Concepts’ (1964) 80 LQR 535, 546-48. 43 An example may be an ‘agreement’ made in a domestic context which does not amount to a contract: Balfour v Balfour [1919] 2 KB 571. 44 An old example of this would be the theory of ‘implied contracts’, under which the recipient of a mistaken payment was deemed to have agreed to repay the money: Sinclair v Brougham [1914] AC 398. 40 41 ‘agreement’ a technical word. One must be familiar with the rules which restrict and expand its meaning in order to properly use and understand it in a legal context. The main thrust of Simpson’s criticism is that Hart’s distinction between statements of law and statements of fact is incomplete. It fails to account for the numerous instances in which lawyers use words to describe facts, yet ascribe a meaning to those words that diverges from their ordinary meaning. In short, a word (such as ‘agreement’) can have a factual counterpart, yet still be a ‘technical’ word that has a distinctive meaning when used in a legal context. The remainder of this article aims to show that ‘possession’ is such a word. Although we have seen that the word refers to a factual counterpart, lawyers still ascribe a meaning to ‘possession’ that often diverges from its ordinary meaning. In some cases ‘possession’ has a much narrower meaning when it is used in a legal context. We will see that there are some cases where there is ample evidence that a person holds a high degree of both physical and mental control over a thing but, due to the context of his control, a lawyer would withhold the description of ‘possession’ from those facts. Conversely, there are cases where there is no evidence that a person holds any physical or mental control over a thing, yet a lawyer would still describe the facts as ‘possession’. These divergences make ‘possession’ a technical term. They explain why ‘possession’ has often been made the subject of legal definition. Lawyers cannot simply look in a normal dictionary for the meaning of the word, but must consult legal sources to learn about its meaning. In the next two sections we will consider the instances in which the legal meaning of the word ‘possession’ diverges from its ordinary meaning, beginning with cases where it takes on a much narrower meaning. 2. Narrowing possession There are a number of instances in which a certain set of facts can be shown to exist (namely a person’s physical and mental control over a thing), yet a lawyer would withhold the label ‘possession’ from those facts. The reason why this sometimes happens is that a finding that the fact of possession exists can have legal consequences that are seen as undesirable.45 A particular instance of this can be seen in relation to larceny, the precursor to modern criminal theft. There was formerly a rule that in order for a defendant to be criminally liable the prosecution had to prove that he had taken a thing from the possession of its owner.46 In this way larceny was seen as protecting an owner’s possession of his thing, rather than his ownership of it. As Pollock and Wright state: ‘[T]he proper conception of larceny is that it is a violation of a person’s possession of the thing accompanied with an intention to misappropriate it.’47 If, for example, an owner of a watch dropped the watch whilst walking in a field, and it was subsequently Simpson (n 42) 548. R v Thurborn (1849) 1 Den 387, 392; 169 ER 293, 296. The rule does not appear to be a feature of the modern crime of theft, probably due to the development of the doctrine of ‘larceny on a finding’. See Hickey (n 11) 18 for a full account. 47 Pollock & Wright (n 4) 118. 45 46 found and kept by the defendant, this would not amount to larceny as the defendant did not take the watch from the owner’s possession.48 Fletcher has sought to justify this restriction of larceny on the basis that interferences with an owner’s possession were more serious than misappropriations that did not disturb the owner’s possession.49 Whatever the justification for this restriction, it caused real problems in cases where a defendant had been entrusted with possession a thing by its owner but then subsequently misappropriated it. An example is the case of R v Walsh50 where a stockbroker, after receiving money from his principal, invested some of it and absconded with the rest. Although the defendant had clearly been dishonest, it was difficult to find him criminally liable in larceny because he already had possession of the money when he misappropriated it, meaning it could not be proved that he had interfered with his principal’s possession of the money. Although no verdict is recorded in the case, Parliament passed a law shortly after making embezzlement a crime, thus covering such facts.51 Another context in which this problem arose was that of servants who had stolen from their masters. If a master entrusted his servant with the possession of goods and the servant subsequently misappropriated those goods, it could not be said that the servant had unlawfully interfered with the master’s possession when he misappropriated it. This would lead to the undesirable outcome that the servant, despite his dishonesty, would not be liable in larceny.52 To avoid this outcome a rule was developed whereby servants were said to be incapable of possessing their master’s goods. Coke states this rule in the following passage: … there is a diversity between a possession and a charge, for when I deliver goods to a man he hath the possession of the goods, and may have an action of trespass or an appeal, if they be taken or stolen out of his possession. But my butler or cook that in my house hath charge of my vessel or plate hath no possession of them … and therefore if they steal the plate or vessel, it is larceny.53 We can see this rule being applied in a number of cases. A good example is the case of R v Bass54 where the defendant was entrusted with a package by his employer and was instructed to deliver it to a particular customer. It appears that before the defendant could deliver the package he was persuaded to visit a pub by some friends and, once inside, he proceeded to open the package and sell The example is used by Pollock & Wright (n 4) 121. See also J Edwards, ‘Possession and Larceny’ (1950) 3 CLP 127. 49 G Fletcher, ‘The Metamorphosis of Larceny’ (1976) 89 Harv LR 469, 475. 50 (1812) 2 Leach 1072, 168 ER 624. 51 52 Geo III c 63 (1812) XX. See also Fenn v Bittleston (1851) 7 Exch 152, 155 ER 895. 52 Harris (n 2) 99. 53 Coke, 3 Instit 108. See also Pollock & Wright (n 4) 26. 54 (1782) 1 Leach 251, 168 ER 228. The rule is not limited to master-servant relationships, but can be seen in many instances where a defendant has been given a licence to possess an owner’s goods or land before his wrongful interference. See, for example, R v Chisser (1678) T Ray 275, 83 ER 142, R v Pears (1779) 1 Leach 212, 168 ER 208, Bertie v Beaumont 1812 16 East 33, 104 ER 1001, Goudge v Broughton [1929] 1 KB 103 and Re Atlantic Computer Systems [1992] Ch 505. 48 the contents to his friends. In finding the defendant criminally liable for larceny, Hotham B said: [T]he [defendant], standing in the relation of a servant, the possession of the goods must be considered as remaining in the master … until and at the time of the unlawful conversion of them by the [defendant].55 It is important to stress that because the defendant had been the only person holding the package in the time leading up to his selling it in the pub, his physical relationship with the package could accurately be described as ‘possession’. However, to label the facts in this way would have meant that the defendant, despite his blatant dishonesty, would not have been liable for larceny as it would have been impossible to say that he had unlawfully interfered with his employer’s possession immediately prior to his opening and selling the package. To avoid this conclusion Hotham B deliberately withheld the label ‘possession’ from facts which, according to the ordinary meaning of the word, clearly did amount to ‘possession’. It is interesting to compare Hotham B’s judgment in R v Bass with another case he decided, R v Deakin & Smith.56 In Deakin & Smith a coach driver had also been entrusted with packages by his employer with instructions to deliver them to customers. In this case, however, the servant was the victim of theft rather than the thief himself. This meant that the problematic issue in R v Bass, the danger of a servant escaping criminal liability by showing that he had not interfered with his master’s possession, did not arise in Deakin & Smith. Consequently, there was no reason for Hotham B to withhold the label ‘possession’ from facts which clearly amounted to ‘possession’ according to the normal meaning of the word. He said: [A]lthough as against his employers the masters of the coach, he, as mere driver, can only have the bare charge of the property committed to him, and not the legal possession of it, which remains in the coachmasters, yet as against all the rest of the world … he has in fact the possession of and control over them.57 The servant’s physical and mental control over the packages meant that his relationship with them could accurately be described as ‘possession’. This is a good illustration of how the legal usage of the word coincides with its ordinary usage unless, as in the case of R v Bass, undesirable consequences would follow.58 When this happens the court will adopt a much narrower meaning of ‘possession’. (1782) 1 Leach 251, 168 ER 228. (1800) 2 Leach 862, 168 ER 530. See also Moore v Robinson (1831) 2 B & Ald 817, 109 ER 1346, cf Mayhew v Suttle (1854) 4 E & B 347, 119 ER 133. 57 ibid 875-76, 537. 58 A similar process can be seen in relation to the doctrine of seisin. See F Maitland, ‘The Mystery of Seisin’ (1886) 2 LQR 481. 55 56 3. Expanding possession So far we have seen how in some cases the legal meaning of possession can be much narrower than its ordinary meaning. The second way in which the legal meaning of possession diverges from its ordinary meaning is that it can, in some cases, be much wider. This particular form of divergence is called a legal fiction. It will be helpful to begin with an example. In the case of Dunwich (Bailiffs) v Sterry59 the claimant sued the defendant in trespass for taking away a barrel of whisky that had washed up on the shore. The claimant had a property right in the barrel as he had been granted a franchise of wreck over the shore where the barrel washed up. However there was no evidence that the claimant had either physical control over the barrel or any intention to control it because he had no physical proximity to the shore nor any knowledge of the barrel’s existence. Despite this Parke J came to the conclusion that the claimant did have ‘possession’, saying that ‘… the right to the possession draws after it a constructive possession …’60 The absence of any physical or mental control on the claimant’s part meant that the word ‘possession’ was patently an inaccurate description of the factual state of affairs. However, this inaccurate description is not the result of a misinterpretation of the evidence. Parke J cannot be accused of reaching an erroneous conclusion; rather, he reached this conclusion in full knowledge of its falsity. This is indicated by his use of the word ‘constructive’ which tells us that the court is deliberately constructing facts that do not exist. What Parke J has done is use the word ‘possession’ as a legal fiction. This means that he has deliberately adopted a description that is ‘inaccurate as an expression of a reality.’61 It was stated above that we know that the word ‘possession’ is being used to denote a fact when a litigant has to introduce evidence in order to establish its existence. The opposite prevails here. The claimant in Dunwich (Bailiffs) v Sterry was positively unable to introduce evidence which would have established the fact of possession. Yet this did not harm his case because the required ‘possession’ was not factual possession, but the fiction of possession. Fictions are typically invoked in order to expand or circumvent some legal rule. As Maitland said: When we hear that A is B in law we can generally draw an inference about past history: it has been found convenient to extend to A a rule which was once applied only to thing which were B in deed and truth …62 It is suggested that the courts have used ‘possession’ as a fiction in order to circumvent the rule, which can be found in different areas of law (including larceny, which we considered above), that an owner of a thing is only protected against interferences with the thing if he has possession of it at the moment of (1831) 1 B & Ad 831, 109 ER 995. ibid 842, 1000. 61 L Fuller, ‘Legal Fictions’ (1930) 25 Illinois LR 363, 371. 62 Maitland (n 58) 486. 59 60 interference.63 There has been constant pressure, as will be seen below, to expand the law’s protection to cases where the owner was not in possession at the time of the interference. Instead of removing the possession requirement in such cases, the courts have found it more convenient to circumvent the requirement by reaching the false conclusion that the claimant was in ‘possession’ of the thing. The main example of this can be found in the tort of trespass. It is commonly said that, like larceny, the torts of trespass to goods and trespass to land (collectively referred to as ‘trespass’ hereafter) do not protect property rights in things, but one’s factual possession of a thing.64 Pollock and Wright give the example of someone who holds a property right in a watch but is not in possession of it at the time it is taken: … where the watch is taken from the possession of a person other than its owner, the owner prima facie ought not to be able to maintain this action in his own name or to prosecute the trespass as for a theft from him, inasmuch as it was not his possession which was violated.65 The possession requirement in trespass means that the claimant, despite having a property right in the watch, is unable to sue in trespass because he cannot introduce evidence which shows that he was in possession at the time of the interference. The fact of possession is sometimes seen as an interest in itself, quite distinct from one’s property right in a thing.66 Lord Denman CJ once said that interfering with claimant’s possession of a thing is more akin to an assault on their person than an infringement of their property right. 67 The possession requirement in trespass, therefore, may be justified on the basis that the purpose of the tort is to protect a claimant’s factual possession, rather than his property right in a thing. Whatever the justification for the possession requirement in trespass, the fact is that it creates gaps in the law’s protection of property rights. If a defendant interferes with a claimant’s chattel or land, and if the claimant cannot establish the fact of possession at the moment of the interference, then there is the prospect that the claimant is left without a cause of action. This is remedied to an extent by the availability of other causes of action, such as conversion, negligence, ejectment and nuisance. However, these actions have their own stringent requirements and do not always furnish the claimant with a cause of action, particularly in cases of minor interferences. The possession requirement Although, as noted above (n 46) this no longer seems to be a feature of criminal theft. Other areas of law where we can see rule include state immunity in possession proceedings (USA v Dollfuss Mieg [1952] AC 582, 611) and insolvency proceedings (Re Atlantic Computer Systems [1992] Ch 505). 64 The similarity to criminal theft on this issue is not coincidental. There was formerly a rule that one could not be liable in theft unless they would also be liable in trespass: R v Thurborn (n 46). This explains why both the crime and the tort had (at least historically) the same requirement for an interference with possession. 65 Pollock & Wright (n 4)121. 66 Green & Randall (n 2) 50-51. 67 Rogers v Spence (1844) 13 M & W 571, 581; 153 ER 239. See also Fletcher (n 49). 63 in trespass, therefore, can result in substantive injustice68 and this has led to pressure to abolish the requirement. However, the approach of the courts has not been to abolish the requirement, but to circumvent it by developing the fiction of possession. This process can be seen in the case, considered above, of Dunwich (Bailiffs) v Sterry. The claimant sued the defendant in trespass for taking a barrel, belonging to the claimant, from a beach after it had washed up. As explained above, the absence of any physical control over the barrel or knowledge of it meant that the claimant’s relationship with the barrel, prior to the defendant’s taking of it, could not accurately be described as ‘possession’ according to the ordinary meaning of the word. Such a conclusion, however, would have led to the undesirable result of the defendant escaping liability in trespass. In order to avoid this and ensure that the claimant’s cause of action did not fail, Parke J circumvented the possession requirement in trespass by invoking a legal fiction. It will be recalled that he said: ‘… the right to the possession draws after it a constructive possession …’69 To say that the claimant had ‘possession’ of the barrel is obviously an inaccurate description of the facts, which Parke J acknowledges through his use of the word ‘constructive’. But he adopts this inaccurate description so as to furnish the claimant with a cause of action in trespass. There are numerous cases which have adopted a similar approach. For example, in Lotan v Cross70 the claimant sued the defendant in trespass for interfering with his chaise, despite the fact that at the time of the interference the claimant had lent the chaise to a friend. Lord Ellenborough said: … the property is proved to be in the plaintiff and, prima facie the thing is to be considered in his legal possession, whoever may be the actual occupier … [A] mere gratuitous possession to a third person to use a chattel does not, in contemplation of the law, take it out of the possession of the owner, and he may maintain trespass for any injury done to it whilst it is so used.71 A non-lawyer, observing the facts of this case, would not label the claimant’s relationship with the chaise as ‘possession’. However, Lord Ellenborough adopts this description (he acknowledges its inaccuracy through his use of terms such as ‘legal possession’ and possession ‘in contemplation of the law’) in order to afford the claimant’s property rights the protection of trespass. This is a clear example of the legal usage of ‘possession’ diverging from its ordinary usage: Lord Ellenborough adopts a much wider meaning of the word, applying it to facts that would not normally be described as ‘possession’, in order to furnish the claimant with a cause of action. An interesting question is whether this fiction of possession has been abolished. The normal way this happens is by reforming the rule that the fiction is designed to circumvent, which in the case of trespass is the possession eg Penfolds Wines Pty Ltd v Elliott (1946) 74 CLR 204. Dunwhich v Sterry (n 59) 842, 1000. 70 (1810) 2 Camp 464, 170 ER 1219. 71 ibid 465, 1219. 68 69 requirement. Whilst it is certainly true that a number of modern commentators have reiterated the possession requirement in trespass72, there is also much to suggest that the courts have abolished this requirement. If it has been abolished then this would remove the need for lawyers to adopt a wider meaning of the word in this context and thus realign the legal meaning of ‘possession’ with its ordinary meaning to a certain extent. One case which suggests the possession requirement in trespass was removed some time ago is White v Morris73 where it was held that a trustee of certain goods could sue in trespass even though it was the beneficiary who was in possession at the time of the interference. In holding that the trustee could sue, Jervis CJ simply said that that he had ‘a right to the possession’74 of the goods and that this gave him standing to sue in the tort. Similarly, in Nicolls v Bastard75 Parke B held that a bailor of a chattel could sue in trespass notwithstanding the fact that the bailee is in possession at the time of the interference. In neither case did the judge resort to invoking the fiction that the claimant had ‘possession’; rather, they simply held that a right to possession was sufficient to sue in the tort. The Court of Appeal recently followed these decisions in the case of White v Withers.76 The claimant discovered that his wife, who he was in the process of divorcing, had been intercepting his mail and passing it to her solicitors, the defendants, whom the claimant sued in trespass. Because the defendants did not interfere with the claimant’s possession of his letters (if the claimant ever had possession of his letters, he had lost it to his wife before the letters were handed to the defendants), the defendants argued that they could not be liable in trespass and asked for the claim to be struck out. The Court of Appeal refused to do so, stating it is arguable that the defendant still committed trespass. Ward LJ said: If it is established that the defendants took possession from Mrs White [the claimant’s wife] knowing that she was a trespasser, then their taking possession and handling the documents may be as trespassory as hers.77 Although the case law is fairly slender, it is suggested that there is authority for the view that possession at the moment of the interference is not a requirement for liability in trespass. If we have removed the possession requirement in trespass, as it is suggested we have, then there is no longer a need to circumvent this rule by using possession as a legal fiction. This would remove one of the pressures on the courts to adopt a meaning of ‘possession’ that diverges from its ordinary meaning. Conclusion Clerk and Lindsell, for instance, states that ‘Trespass is actionable at the suit of the person in possession of land …’: M Jones (et al) (ed), Clerk & Lindsell on Torts, 20th edn (London, Sweet & Maxwell, 2010) para [19-10]. See also Star Energy Weald Basin Ltd v Bocardo SA [2010] UKSC 35, [6]. 73 (1852) 11 CB 1015, 138 ER 778. See also Barker v Furlong [1891] 2 Ch 172. 74 ibid 1028, 784. 75 (1835) 2 CM & R 659, 150 ER 279 76 [2009] EWCA Civ 1122. 77 ibid [49]. 72 This chapter has reached two conclusions. First, we have seen that when the word ‘possession’ is used in a legal context, it is used to denote a fact. The proof of this is that a litigant wishing to rely on some rule that makes reference to the word ‘possession’ will have to introduce evidence of possession. The statement that, ‘X has possession of a thing’, when found in a legal context, therefore, is a statement of fact. It was noted that this first conclusion may lead to the temptation to see ‘possession’ as an ordinary, non-legal, term. The second part of this chapter has shown that this is not the case. The legal usage of the word often diverges from its ordinary meaning. We saw that there are pressures on the courts to adopt a narrow meaning of ‘possession’, so as to prevent certain classes of defendants from escaping criminal liability in theft. Conversely, we also saw that in certain cases there are pressures on the courts to adopt an extremely wide, or fictional, meaning of ‘possession’ in order to furnish particular types of claimants with the protection of the tort of trespass. The second conclusion reached by this chapter is that, as a result of these divergences in meaning, possession has become a technical term. Although its legal meaning overlaps to a great extent with its normal meaning, it is also defined by rules which, as we have said, both narrow and expand the meaning of the word. What makes the term ‘technical’ is that one must be familiar with these rules in order to properly use or understand ‘possession’ when it is used in a legal context. This explains why ‘possession’, despite being a common fact, observable by lawyers and non-lawyers alike, continues to be the subject of legal analysis and debate.