Actions of Covenant, 1200–1330

JOSEPH BIANCALANA

The following abbreviations and abbreviated citations are used in this essay:

Bedfordshire Eyre--Bedfordshire Eyre 1247, ed. G. H. Fowler, Bedfordshire Historical Record Society, vol. 21 (1939).
Berkshire Eyre--The Roll and Writ File of the Berkshire Eyre of 1248, ed. M. T. Clanchy, Publications of the Selden Society, vol. 90 (1972–73).
BL--British Library
BNB--Bracton's Notebook, ed. F. W. Maitland, 3 vols. (London: C. J. Clay, 1887).
CP40--Common Plea Rolls, Public Record Office
CRR--Curia Regis Rolls (1922– )
Earliest English Law Reports--Earliest English Law Reports, ed. Paul Brand, Publications of the Selden Society, vols. 111, 112 (1995, 1996).
Gloucs.--Rolls of the Justices in Eyre for Gloucestershire, Warwickshire and Staffordshire [Shropshire] 1221, 1222, ed. D. M. Stenton, Publications of the Selden Society, vol. 59 (1940).
JUST 1--Itinerant Justice Rolls, Public Record Office
KB26--Curia Regis Rolls, Public Record Office
Kent Eyre--The Eyre of Kent, 6 and 7 Edward II, a.d. 1313–1314, ed. F. W. Maitland, L. Harcourt, and W. Bolland, Publications of the Selden Society, vols. 24, 27, 29 (1909, 1912, 1913).
LI--Lincoln's Inn
London Eyre--The Eyre of London, 14 Edward II, a.d. 1321, ed. H. M. Cam, Publications of the Selden Society, vols. 85, 86 (1968, 1969).
RCR--Rotuli Curiae Regis, ed. Francis Palgrave, 2 vols. (London: Eyre and Spottiswoode, 1835).
RS--Rolls Society
Shropshire Eyre--The Shropshire Eyre Roll of 1256, ed. Alan Harding, Publications of the Selden Society, vol. 96 (London, 1980).
SS--Selden Society
YB--Yearbook
This article tries to answer two questions about the writ of covenant that have perplexed legal historians. Maitland noticed that the common law courts adopted a requirement that plaintiff produce a writing of the covenant under defendant's seal as proof of the covenant.1 When and why royal officials adopted this requirement continue to be the chief unanswered questions about covenant. Second, there is the matter of remedy. Some legal historians believe that the remedy in covenant changed from specific performance to damages.2 But what was the remedy in covenant? Did it change? If so, when and why did it change?	1
Legal historians have had difficulty with the writing requirement and the question of remedy because they have thought of covenant as if it were a monolithic legal entity--a cause of action. This assumption has shaped the questions of when did the royal courts adopt a requirement of a writing in covenant and when did the remedy in covenant change? These questions presuppose that the royal courts adopted a writing requirement for all cases of covenant at a single time for a single confluence of reasons and that the remedy changed in all cases of covenant at a single time for a single confluence of reasons. But there was no cause of action in covenant in the thirteenth century.3 There were writs of covenant or, more accurately still, there were claims made by writs of covenant, claims arising out of particular transactions that allegedly went bad for plaintiffs. The anachronistic assumption that there was a cause of action in covenant is understandable not only because the abstraction, cause of action, was familiar to common law historians but also because using the abstraction seemed to organize and simplify the evidence. For once examine the claims made by writ of covenant and one can quickly become lost in a welter of individual cases. The author of the Statute of Wales (1284) suffered that plight.4 He said that the variety of actions of covenant was infinite. For that reason, he continued, no rule can be set forth for all actions of covenant. Rather, litigation proceeds and the law varies according to the nature of each covenant. The author's despair at stating general rules for cases of covenant wins the sympathy of anyone who has surveyed more than a handful of covenant cases.	2
Yet any attempt to understand the actions of covenant in the thirteenth century requires that individual cases be grouped into types or categories. Any attempt to construct types or categories of covenant cases will be governed by the purpose for which the types or categories are constructed. This article seeks to answer two legal questions about cases of covenant--the law governing proof and the law governing remedies. For this purpose, claims made by writs of covenant should be put into types or categories according to the legally significant differences among claims. Claims differ in a legally significant way when the principle of defendant's liability differs, the legal idea in virtue of which defendant's conduct is a wrong to plaintiff and not merely a misfortune for plaintiff.	3
For many of the legal historians who have struggled with covenant, the idea that there was more than one principle or basis for defendant's liability seems odd. In the accepted view, a covenant (conventio) was simply an agreement.5 The writ of covenant enforced the performance of agreements.6 Covenant was thus a contractual action.7 The writing requirement was a requirement that plaintiff produce a written contract. Because covenant was a contractual action, a remedial order to perform the covenant was specific performance, until the remedy changed to damages.	4
Covenant (conventio) indeed meant agreement. But that is no reason to suppose that all actions of covenant in the thirteenth century were contractual actions in our sense of the term. One does better to look at the cases and listen to how plaintiffs made out their claims in their pleadings. Plaintiffs made three types of claim. In most cases, plaintiff was a termor who claimed that defendant had demised to him certain land for a term of years, that plaintiff had been seised of the land under that demise, and that defendant had ousted him. In this type of case, plaintiff was making a proprietary claim, a claim, essentially, that he had greater right to hold the land as defendant's termor or lessee than defendant had to hold the land free of plaintiff. In other words, plaintiff claimed that defendant had made a grant of land to him and was barred from taking back the grant. In a second type of case, plaintiff asserted that he and defendant had entered into an agreement and that defendant had breached the agreement. Plaintiff thus made a contractual claim, a claim that defendant had undertaken a special obligation to plaintiff and had breached that obligation. The obligation was special, as opposed to general, because absent the agreement defendant had no duty to perform the actions as alleged by plaintiff.	5
The writ of covenant thus had a dual nature.8 It could be used to initiate a proprietary, real action or it could be used to initiate a contractual, personal action. Until the last decades of the thirteenth century, plaintiff's claim was more important than the fact that he used a writ of covenant. The dual nature of covenant is revealed by the way clerks described the writ in plea roll entries. The writ of covenant was a precipe writ, which commanded defendant to hold to his covenant with plaintiff.9Precipe writs ordinarily initiated proprietary or real actions. In the earlier cases the plea roll clerks entered the writ on the rolls as a precipe writ.10 Beginning in the 1220s, however, they entered the writ as if it were an ostensurus quare writ.11Ostensurus quare writs ordinarily initiated personal actions. In the late 1230s plea roll clerks adopted a hybrid form for entering covenant writs, one that began as if the writ were an ostensurus quare writ and ended as if it were a precipe writ.12 The hybrid style of entry reflects the dual nature of the writ.	6
There was also a third type of covenant case. In this type of case, plaintiff claimed that he and defendant had entered into an agreement that called for defendant to convey to him land in freehold or fee but that defendant had failed to make the conveyance. This type of case and claim presented judges and lawyers with great difficulty. There was no consensus until perhaps the first decade of the fourteenth century whether this type of case presented a proprietary claim or a claim of obligation. If one focused on plaintiff's claim to recover land in freehold or fee, this type of claim looked like a proprietary claim. But if one focused on the agreement and its breach, this type of claim looked like a claim of special obligation.	7
Putting cases and claims of covenant into these three categories takes one a long way toward understanding the writing requirement and the remedies in covenant cases. The proof requirement for the peculiar kind of proprietary claim made by writ of covenant differed from the rules of proof that governed claims of special obligation. The remedy in real actions was to put plaintiff in seisin, a remedy resembling specific performance, but the remedy in cases of obligation was damages. In cases involving claims of freehold, where judges and lawyers were uncertain whether plaintiff made a proprietary claim or a claim of obligation, there was uncertainty over what proof plaintiff had to produce and what remedy, if any, plaintiff could have by writ of covenant.	8
I. The Requirement of a Writing
The first unanswered questions about covenant are when and why royal officials adopted a requirement that plaintiff produce a writing of the covenant under defendant's seal. It is useful to consider the three types of covenant cases separately. In actions by lessees, the requirement of a writing was probably as old as the writ, which was probably invented as a remedy for a lessee ejected by his lessor. In this case, an action on the writ was a real action based on a grant. A written lease was necessary to make the lessor warrant his grant. Other claims made by writ of covenant, except claims to freehold or fee, were based on defendant's special obligation to plaintiff. In these cases, plaintiff had to produce suit as witnesses or a writing of the special obligation defendant had undertaken to plaintiff. In the late 1270s, suit and wager of law were eliminated as modes of proof in actions of covenant. From this decision, there emerged the requirement that plaintiff produce a written covenant.13 Yet, earlier, a minority view had held that where plaintiff claimed land in freehold or in fee, suit alone was not sufficient. He had to produce a writing. There was disagreement whether plaintiff's writing could be a charter of feoffment or had to be a chirograph. The minority view that plaintiff had to produce a writing succeeded when as a general matter suit and wager of law were eliminated. There continued to be debate whether plaintiff's writing could be a charter of feoffment.	9
A. Leases and Other Grants of Less than Freehold
Legal historians have long been aware that most actions of covenant involved leases.14 Only David Ibbetson, however, has treated actions on leases separately from other actions of covenant when considering the requirement of a writing.15 He argued that the requirement of written leases was transplanted from Roman law, an argument he has since abandoned.16 But he was right in that a requirement of written leases preceded the general requirement of a writing. Indeed, that a lessee had to produce a written lease was a requirement probably as old as the writ.	10
The evidence that a lessee suing on a lease had to produce a written lease is simply the fact that plaintiff-lessees almost always produced written leases. In the comparatively few cases in which the plea roll entry does not say that plaintiff produced a written lease, defendants never took issue with plaintiff's description of the lease. Most often in these cases, defendants admitted the covenant, denied the breach, and put themselves on a jury. In one case, defendant claimed that plaintiff had never had seisin under the lease, but the jury found otherwise.17 The plea roll clerk might well have omitted mention of a written lease in these cases because they did not turn on plaintiff's description of the terms of the lease. None of the cases in which defendant waged his law to deny plaintiff's description of the covenant involved leases.18 Where defendants waged their law to deny a breach of the lease, plaintiffs' claim, under written leases produced in court, was that defendant had failed to acquit the tenements of services owed a superior lord, had committed waste, or had denied plaintiff estovers in woods appurtenant to his lease hold.19	11
Very few cases raised the issue of plaintiff's proof of the lease. In 1256, defendant asked plaintiff to produce a writing, charter, or other muniment of the lease.20 Significantly, his request was for a writing, not suit. But plaintiff had neither suit nor a writing. He offered a mark for a jury trial. No judgment is recorded. In 1277, a lessee sued his lessor in covenant for ejectment and the defendant asked what proof plaintiff had of the lease. Plaintiff said he had nothing, because defendant had taken his charters in the ejectment.21 The case went to a jury. Plaintiff assumed that he had to produce a writing or had to explain his inability to do so. In 1284, when plaintiff produced a written lease, defendant, arguing that the writing was the foundation of plaintiff's writ, claimed that the writing varied from the writ.22 Plaintiff admitted the variance and was given license to withdraw. Because these latter two cases date from the period during which a writing requirement was emerging generally in covenant, one cannot be certain that they are evidence of a requirement peculiar to leases. Yet in earlier as well as later cases, lessees almost invariably produced written leases. One infers that they were not merely following a rule but were being guided by one. The question is not whether there was a requirement but the reason for the requirement.	12
Maitland and Simpson, perhaps following what seems to have been Britton's suggestion, have surmised that the writ of covenant was probably invented in order to provide an ejected lessee a remedy analogous to that provided a freeholder by the assize of novel disseisin.23 By thinking through the implications of this suggestion, one can deduce the requirement of written leases. In the thirteenth century, a term of years could be a farm lease under which the tenant farmer rendered a periodic rent or, more frequently, it was security for a loan hardly distinguishable from a gage of land.24 A creditor loaned money and his debtor secured and frequently repaid the loan by granting the creditor a term of years.25Glanvill, a treatise written in the 1180s, said that if a creditor received land in gage and was ousted by his debtor, the creditor did not have an action in the royal courts to recover his gage.26Glanvill made a point of saying that the ousted gagee did not have an assize of novel disseisin. He rationalized this rule by saying that the creditor had his action to recover his debt. Although Glanvill speaks of the gagee being seised of the land held in gage, one wonders what kind of seisin the gagee enjoyed if it was not protected by novel disseisin. The gagee and the lessee did not have novel disseisin because they had not been seised of a free tenement, of freehold.27	13
But not all gagees were without a remedy. As early as the 1190s creditors could purchase royal assistance to acquire land gaged to them.28 In these transactions the debtor designated land as gage or security for his loan and his creditor could take possession of that land in gage should the debtor default. The agreement could be called a covenant.29 Perhaps from the 1190s a Jewish creditor ousted by his debtor from land held in gage could recover the land by judgment of the Exchequer of the Jews.30 Christian creditors who took land in gage or for a term of years had no remedy. The writ of covenant gave the Christian creditor a remedy in the common law courts analogous to the remedy enjoyed by Jewish creditors in the Exchequer of the Jews.	14
But why did they have to produce a written lease? In credit transactions substantial enough to be secured by leases or gages of land the prudent creditor might well require the transaction to be written down for his own protection. The evidence of a writing requirement--that plaintiffs-lessees always produced written leases--might only reflect the practical prudence of creditors or of other lessees for that matter. But there were also legal reasons requiring plaintiff-lessees to produce written grants of their terms of years. The legal requirement can be reconstructed by comparing covenant to recover a term of years with novel disseisin or mort d'ancestor to recover land in freehold or in fee.	15
A lease for a term of years required a grant of land for the term.31 A lessee who sued his lessor in covenant for ejectment said in his pleadings that defendant had conceded or demised the land to him for a term of years, that he had been seised under the concession or demise, and that defendant had ejected him.32 Plaintiff thus set forth a grant and said how he had been seised in virtue of the grant until defendant, the grantor, ejected him. Plaintiff was demanding that defendant-grantor maintain his grant and not take it back. Ordinarily what barred a grantor from taking back a grant of land was that he had taken the grantee's homage or had made a written commitment to warrant the grant. What was peculiar about the lessee's grant was that the grantor had not taken the grantee's homage and there had been no livery of seisin in the usual sense. Nor could the lessor take the lessee's homage without turning the grant into a grant of freehold or fee. Without homage there was no implicit warranty of the grant. If there were to be warranty, it would have to be by charter, by writing. This was because the only two bases for raising a warranty bar were homage or a writing.33	16
Claims in covenant of ejectment from a term of years were of three types. First, the lessee might claim that his lessor ejected him.34 In this type of case, covenant served as the lessee's novel disseisin. If, however, a grantor of land in freehold or fee ejected his grantee, that would be a disseisin only if the grantor had taken the grantee's homage and had put him in seisin.35 In order for the lessee to raise a warranty bar against his lessor taking back his grant, he needed a written lease because there had been no homage. Second, the original lessee's heir might claim that his ancestor's lessor had excluded him.36 The plaintiff's ancestor was most often his father or, less frequently, his brother or uncle. In this type of case, the writ of covenant served as mort d'ancestor for the lessee's heir. In two cases, the lessor answered that his lessee had died seised and had left the term of years to a third party by testament.37 The lessor had accepted the legatee to continue the term of years. Mort d'ancestor used the homage or warranty bar to enforce rules of inheritance within a limited range of possible heirs.38 The lord's warranty to his tenant descended to the tenant's heir. The lord was thus prevented from taking back the land upon his tenant's death. An important requirement of mort d'ancestor was that the tenant have held of the lord in fee and inheritance. A lessee did not, of course, so hold his term of years. The problem was to create the lessor's warranty and get it to descend to the lessee's heir where there had been no homage and could have been no homage. A written commitment by the lessor to his lessee and his heir was necessary.	17
Difficulty could also arise upon the death of the lessor. If the lessor's heir was of age, he might not recognize the lessee.39 Indeed, lessees complain that their lessor's heir ejected them. The analogous action for a freehold tenant was novel disseisin. If the lessor's heir were underage, the lease was suspended during the wardship of the heir. Upon coming of age, the lessor's heir might not admit the lessee to the remainder of his term of years.40 The analogous action for a tenant in fee was a writ of right. In the actions by tenants in freehold or fee analogous to covenant for the termor the lord's heir was bound only if the lord had taken homage or had warranted the grant. In order to protect a lessee, it was necessary to create the lessor's warranty and get it to descend to his heirs. Absent homage, the lessee needed a writing to create the warranty and have it descend to the lessor's heir.	18
This analysis helps to explain the 1256 case noted above.41 In turn, this case elucidates why there was a requirement of a written lease in covenant. A lessee claimed that defendant's father had given him a lease of certain land and that defendant, soon after his father's death, had ejected plaintiff. Defendant asked to see a writing of the lease. When plaintiff conceded that he had only his own bare word, defendant asked for judgment whether that was sufficient to put him to his law or to a jury. Defendant might have wanted to deny the ejectment or the lease itself. Suit alone might have been sufficient to put him to his law to prove his denial of the ejectment but not of the lease itself. Defendant could not wage his law to deny that his father had entered into a lease. One could not wage law for another.42 Suit alone would not be sufficient against a lessor's heir. Furthermore, defendant would not be bound by his father's lease unless the lease bound the lessor and his heirs.43 Homage would bind a grantor's heirs to warranty, but there had been no homage. Without homage, a lessee needed a writing in order to bar his lessor's heir from taking back the lessor's grant. Defendant's demand for a writing, not merely suit, was right on point.	19
A remedy for a lessee ejected by his lessor analogous to novel disseisin or mort d'ancestor could not have been invented without a requirement that the lessee produce a written lease. Conversely, if royal officials had decided that grantees of less than freehold or fee did not have to produce a written grant but could nevertheless bind their grantor and his heirs without homage or livery of seisin, they would have greatly expanded novel disseisin and greatly contracted mort d'ancestor. Novel disseisin would no longer have been limited to freehold tenants but would have been available to anyone in possession under any type of grant or permission. That would have opened novel disseisin to villeins holding by customary tenure. If any sort of oral grant or permission without homage or livery of seisin would have been sufficient to stop an assize of mort d'ancestor, any last will of land or death-bed transfer, under which the grantee gained some sort of possession, would have withstood mort d'ancestor.44 That would have destroyed the royal, common law policy against last wills of land. Royal officials had good reasons to maintain the integrity of novel disseisin and mort d'ancestor. In order to fit a remedy for lessees into the legal framework of novel disseisin and mort d'ancestor there had to be a requirement that the lessee produce a written lease. Little wonder, then, that lessees who sued in covenant for ejectment by their lessors regularly presented written leases.	20
The writ of covenant also served as the lessee's writ of mesne.45 In an action of mesne, plaintiff's claim had four parts: defendant was his lord for certain land held of the defendant, defendant was obliged to perform the services owed a superior lord for that land and thus to acquit the plaintiff, defendant had failed to do so, and, as a consequence, the superior lord had distrained plaintiff for the services. Although there were plaints sounding like actions of mesne as early as 1200 and some evidence of a writ of mesne as early as the first decade of the thirteenth century, the writ does not appear to have been in regular use until the later 1220s.46 Tenants in fee, however, could use either a writ of warranty of charter or a writ de fine facto to lodge their complaint that their lord had failed to acquit them of services owed the superior lord.47 Plaintiffs to these writs had to produce the charter or the final concord. As early as the 1250s, defendants to actions of mesne began demanding that plaintiff produce a writing as proof that defendant had undertaken the obligation to acquit the tenant of services owed a superior lord.48 In the 1280s the king's court gestured toward requiring plaintiff to produce a writing of defendant's obligation to acquit.49 But the prevailing rule was that plaintiff could produce a writing or could plead either that defendant had been seised of his homage and the same services as those owed the superior lord or that defendant or his ancestor had acquitted the tenant or his ancestor in the past.50 This last alternative, a sort of prescriptive acquittance, was a development of the later thirteenth century. Earlier in the century, plaintiff had to produce a writing or plead that defendant was seised of his homage. A lessee, unable to plead that his lessor had taken his homage, had to produce a written lease.	21
The reason for requiring lessees to produce written leases also applied to other grantees of less than freehold or fee. Plaintiffs in these less frequent cases also based their claims on a concession or demise by defendant under which they were seised until defendant ejected them. They almost always produced a written covenant. Defendants never waged their law to deny the covenant.51 The written grant was necessary to bar defendant from taking back the grant.	22
For the most part lessors used covenant either to claim rent in arrears or to recover for the lessee's failure to return the property in the condition required by the lease.52 They almost always produced a written lease, but we do not know whether they were required to do so. The reasons why a plaintiff-lessee had to produce a written lease--to bar the defendant-lessor from taking back his grant--would not apply to a lessor. A desire for symmetry between lessee and lessor and the desire for a uniform rule for leases could have extended the requirement beyond its supporting reasons. It is hard to find a case by a plaintiff-lessor without a written lease. Given that plaintiff-lessors produced written leases, not surprisingly, one does not find defendants waging their law to deny the lease. Until the later 1270s, defendants could wage their law to deny the breach. As one might expect, defendants waged their law to deny that they had failed to return the property in good condition.53 Given the written lease, a defendant could not wage his law to deny that he was in arrears in his rent. He had to produce suit or a written acquittance.54	23
Two actions by lessors dating from 1304 have been read to mean that there was no writing requirement in covenant in the early fourteenth century.55 What the cases show at best, however, is that the court was willing to make discrete exceptions to a requirement of a written lease. The first 1304 case presents the difficulty that the plea roll record of the case stops after defendant's standard, opening denial of force and wrong.56 The plea roll record thus differs from three reports of the case, which show the parties pleading to an issue of fact.57 From the plea roll record one infers that plaintiff was not successful.58 The reason for the plaintiff's failure probably had more to do with the fact that plaintiff claimed freehold where a writ other than covenant was available for that purpose than that he did not have a written lease. Plaintiff claimed that he had leased land to defendant for a two-year term, that defendant had agreed to make a writing of the lease, that defendant never did so, and that defendant had stayed beyond his two-year term. Defendant demanded that plaintiff produce a written lease. According to two reports, plaintiff responded that he had good suit; according to another, he responded that he had alleged defendant's broken promise to make a writing and that whether defendant had stayed beyond the term was cognizable by a jury.59 The court let the pleading continue. According to two reports, defendant then argued that one could not recover freehold by writ of covenant without at least showing title or specialty.60 At this point, according to all the reports, defendant claimed the freehold. The court then put the burden of proving freehold on defendant.61 He asserted that an earlier assize of novel disseisin plaintiff had brought against him had found that he, defendant, had freehold. That, according to all three reports, was the issue of fact. But the plea roll record stops well short of an issue of fact.	24
According to the reports, the court treated plaintiff's claim as if he had brought a writ of entry ad terminum qui preteriit, which fit the case of a lessee holding over his term and for which plaintiff need not produce a written grant of the term to defendant. But in an action on the writ of entry, plaintiff would have had to plead his right to the land based on his own or his ancestor's seisin and could, on that basis, recover the freehold. It is odd in the 1304 case that the court put the burden of proving freehold on defendant. In at least three much earlier cases, lessors had used covenant to claim that their lessees had held over their terms.62 But by 1300, the court had severely restricted the use of covenant to recover freehold.63 In 1292, for example, when a widow, for reasons best known to her and her husband, tried to use covenant to recover her dower, the court threw out the case.64 She had a writ of dower. In the 1304 example, the court probably, on second thought, also threw out the case. Plaintiff had his writ of entry. The justices also might well have reconsidered making defendant, rather than plaintiff, prove freehold. It is at any rate worth noting that plaintiff claimed that defendant had promised to make a written lease, a claim with little purpose unless there was a requirement of a written lease.	25
In the second 1304 case, plaintiff, a dowager, had granted her life estate to defendant.65 She alleged that defendant had committed waste and the heir had recovered against her. The heir could not have proceeded against her grantee.66 In that action brought under Chapter Five of the Statute of Gloucester she had lost her life estate and had paid triple damages.67 She now wanted to recover over against the wrongdoer. The problem was how to do so. She could not sue under Chapter Five of the Statute of Gloucester, because in an action under the statute she would have to claim that defendant had committed waste to her disinheritance.68 The claim of disinheritance would be the basis for recovering the land. As a dowager, she could neither claim inheritance nor recover the land in fee.69 She could sue in covenant if the lease prohibited waste or required the tenant either to maintain the property or to return the property in as good a condition as he had received it.70 These contractual obligations, as defendant pointed out in one report, required a specialty.71 Plaintiff was careful not to allege that a covenant with defendant had prohibited waste. Rather, she alleged that she had made the grant on the condition that ("ita quod") defendant not commit waste.72 Nevertheless, because she brought a writ of covenant and did not have a written lease, Justice Bereford had a difficult case before him.	26
If he enforced the writing requirement in covenant, plaintiff would have no other means of recovery and the wrongdoers would escape.73 He could prevent the innocent parties from bearing the loss and could do the least damage to the law by making an exception to the requirement of a writing. That is what Bereford does. He never denies that there is a requirement of a writing. Rather, he says that "it would be great hardship if they [plaintiff] were not to be answered on the writ without specialty."74 Bereford also says that since waste is against the law, if defendant had been given permission to commit waste, he should produce a specialty to prove the permission. And he excuses plaintiffs for not having a specialty. In one report he says the defendant has the specialty and in another he says that a lease ordinarily stays in the possession of the tenant, here the defendant.75 Bereford is unwilling to allow defendants to escape because of the writing requirement. In effect, Bereford allows dowagers to use covenant to bring an action under the statute without claiming that the waste is to their disinheritance, at least where the heir has already recovered against her. A very narrow exception to the writing requirement.	27
B. Other Covenants
The requirement of a writing developed differently for actions that were not based on a lease. Actions to recover land or rents in freehold or in fee presented special difficulties. These can better be appreciated after the pattern for non-lease covenants is established. Covenants that were neither leases nor agreements calling for conveyances of land in freehold or fee were of various kinds. In return for land received, the recipient might grant back an annual supply of food and clothing. The grant of necessities, a type of corrody, was analogous to a grant of an annuity. Covenants regulated access to and enjoyment of commons in pasture or in woods. The agreement might be to provide access to commons, to maintain an enclosure of commons, or to apportion improved and enclosed commons between the parties.76 There were various indemnity agreements: to indemnify a surety, to hold a sheriff harmless for releasing a prisoner to defendants, to hold harmless a coparticipant in a disseisin, to indemnify a grantee of land if he is penalized for not maintaining a sea wall.77 An indemnity agreement could operate as a warranty of title in a sale of chattels.78 And there were agreements to perform services such as building a mill or a bell tower, transporting a cartload of hay, or healing a wound.79 Very few cases involved indemnity or service agreements.	28
The Emergence of the Writing Requirement. Before the late 1270s plaintiff to a writ of covenant had to produce suit or a writing as proof of the covenant. At the least plaintiff had to produce suit. In at least eleven cases, plaintiff lost because he failed to do so. Some plea roll entries say that plaintiff lost because he had only his bare word. Other plea roll entries say that plaintiff lost because he had neither suit nor a writing, which indicates that either would have been sufficient. Yet others indicate that plaintiff had nothing or no suit.80 Suit was required in many types of action in the first half or so of the thirteenth century. They were witnesses ready to support crucial points of plaintiff's claim.81 Where a plaintiff in covenant produced both suit and writing, and most plaintiffs did, the suit were probably witnesses to the breach. Where plaintiff had only suit, his suit had to serve as witnesses to the making of the agreement and to the breach. In one case, plaintiff lost because he had no proof of the covenant or the breach.82 In another, defendant admitted the covenant but won because plaintiff did not have suit to prove the breach.83	29
If plaintiff produced suit, but no writing, a defendant who denied plaintiff's description of the covenant could put himself on a jury or wage his law to prove his denial. In at least ten cases defendant waged his law in denial of the covenant.84 Where the plea roll entry does not mention that plaintiff produced a writing, a defendant who neither conceded the action nor waged his law almost always admitted the covenant, denied the breach, and either waged his law in denial of the breach or put himself on a jury. The only cases I could find in which defendant denied the covenant and put himself on a jury were in 1280 and 1292.85 By then defendants could no longer wage their law. This pattern suggests that even when the plea roll entry does not say so plaintiffs probably had a writing. If plaintiff produced a writing of the covenant, defendant could not wage his law against his own or his ancestor's deed. He could claim, however, that the writing was not authentic. The authenticity of the writing would be tried by means similar to those used to try the authenticity of charters. The seal on the proffered writing could be compared to other seals known to be valid.86 More frequently, the issue was put to a jury.87 Witnesses named in the specialty, if any, were added to the twelve jurors to make up the jury.88 In either case, suit or writing, a defendant who denied the breach could wage his law or put himself on a jury. In at least thirteen cases, defendant waged his law in denial of the breach.89	30
Wager of law was removed from cases of covenant in the later 1270s. The latest case I could find in which defendant waged his law in denial of the covenant occurred on the Lincolnshire Eyre of 1271–72.90 The latest case I could find in which defendant waged his law in denial of the breach occurred in 1275.91 The loss of wager of law in the later 1270s was not peculiar to actions of covenant. In actions of mesne, wager of law to deny a failure to acquit plaintiff ended in the late 1280s.92 In actions to enforce final concords, wager of law to deny a breach of the final concord did not survive much beyond the Yorkshire Eyre of 1279–81.93 In actions of replevin, wager of law did not appear much beyond the Lincolnshire Eyre of 1271–72.94 In these other actions, wager of law was analogous to wager of law to deny a breach of covenant rather than wager of law to deny plaintiff's description of the covenant. For example, in an action to enforce a final concord, defendant could deny that he acted contrary to the final concord and prove his denial by wager of law. His answer is similar to a defendant's denial in covenant that he went against the covenant.	31
In covenant, defendants also lost their ability to wage law in denial of the covenant. The loss of wager of law for defendants meant that suit was no longer sufficient proof for plaintiffs. They had to produce a writing of the covenant. In two other actions, the elimination of suit and wager of law in the 1270s resembled the loss of suit and wager of law as a mode of trying plaintiff's description of a covenant. There is some evidence that in an action for an annual rent plaintiff had three choices. He could produce a written grant of the annual rent, he could claim that he had been seised of the rent, or he could produce suit.95 If he produced suit, defendant could wage his law. Wager of law went to whether defendant had an obligation to pay an annuity to plaintiff. For in 1256, the court asked plaintiff whether he had a writing of the annual rent.96 When plaintiff said he had no charter or writing, the court permitted defendant to wage his law. A writing could only show whether defendant was obligated to plaintiff, not whether he was in arrears. Defendant probably could not admit the obligation and wage his law to deny that he was in arrears. He would have to show a written acquittance. This much was clear by the 1270s.97 At some point after 1256, perhaps in the later 1270s, wager of law was eliminated from actions for an annuity. From then on, plaintiff was left with the choice between producing a writing or alleging seisin.	32
Before the later 1270s, a defendant to an action of debt on a written obligation who admitted the debt but wanted to plead part or full payment had to produce suit or a written acquittance as proof of payment.98 He could produce suit to prove payment whether plaintiff had produced suit or a writing in proof of the debt. If defendant produced suit, plaintiff could wage his law to deny receipt of the payment.99 The elimination of suit for defendant and wager of law for plaintiff began in about 1272. In the Hampshire and Cambridgeshire eyres of that year the justices began to enforce the new rule: where plaintiff had written evidence of the debt, defendant had to produce a written acquittance of the debt.100 The old and the new rules coexisted until shortly after 1278. Defendants to debt on an obligation thus suffered the same fate as plaintiffs in covenant. They once could choose between suit and writing, but the elimination of suit and wager of law meant they had to produce a written acquittance as proof of payment. Before wager of law was eliminated, defendants in covenant could also plead that they had complied with the covenant and produce suit in support of their answer.101 Plaintiff could then wage his law.102 There are, however, very few examples of this practice on the plea rolls. The latest example I found was in 1276.103	33
The removal of suit for defendants and wager of law for plaintiff as the means of proving payment affected wager of law in debt where plaintiff produced only suit. A defendant who had paid part or all of the debt had three choices: he could produce an acquittance, he could allege payment or enter a general denial and put himself on a jury, or he could enter a denial and try to wage his law. Defendant was unlikely to have an acquittance. Some defendants in debt alleged payments and put themselves on a jury.104 There was some effort toward not permitting defendants who had paid to wage their law. After the 1270s, in most cases of debt in which defendant waged his law, he denied having incurred the debt by specifically denying the transaction alleged by plaintiff.105 Specific denials of the debt-creating transaction are found earlier, but they appear as a smaller proportion of cases in which defendant waged his law.106 The move toward limiting wager of law to denials of having incurred the debt was also and by the same token a move toward not permitting defendants who had paid to wage their law. As Chief Justice Scrope explained on the Northamptonshire Eyre of 1329–1330, an allegation of payment is an affirmative statement unsuited to wager of law.107 Wager of law was to prove a negative statement. But defendants could no longer produce suit as proof of the affirmative allegation of having paid the debt. This change left defendants in a hard place. Some of the denials on which defendants waged their law sound as if they are denials on the grounds of having paid. A defendant who answers that he is not bound by "the aforesaid debt" ("predicto debito") sounds as if he is not denying having incurred the aforesaid debt but is denying that it is still owing.108 But the phrases used were ambiguous, making it as hard then as now to tell whether defendant denied incurring the debt or any longer owing the debt.109 A plaintiff on the Kent Eyre of 1313–1314, aware of the ambiguity of a general denial, tried with unknown results to pin defendant down to a specific denial of having incurred the debt or an allegation of payment.110 The effort would have been pointless unless defendant was thought to be unable to wage law for an allegation of payment.	34
After defendants in covenant lost wager of law, there were few cases in which plaintiff produced only suit and defendant denied the covenant. I have found only two possible cases, one in 1280 and another in 1292.111 The plea roll entries do not mention that plaintiff produced a writing. Defendants denied the covenant and put themselves on juries. Without reports of the cases one cannot be certain whether plaintiff lacked a writing or whether he produced one at defendant's demand. The issue whether suit alone was any longer sufficient was almost raised in a case in 1283.112 Defendant asked what plaintiff had to show of the covenant. The plea roll reads that plaintiff had suit but no writing. The parties did not debate whether suit was sufficient; they instead settled their dispute.	35
As one might expect, once suit and wager of law were eliminated, judges and lawyers paid greater attention to the writing a plaintiff produced as proof of the covenant. In 1284 plaintiff lost because of a variance between his writ and his writing.113 In 1279, when plaintiff claimed that defendant had agreed that she and her late husband were to receive an annual rent from a certain tenant, the only writing she could produce authorized her and her husband to collect the rent from the tenant for defendant.114 Because defendant was the mother of plaintiff's late husband, one suspects that, while her son was alive, defendant in fact permitted him to keep the rent so collected but stopped the practice at his death. Perhaps a decade earlier plaintiff could have produced suit to prove the arrangement. Now, however, she was caught because no writing recorded the practice of the parties. In 1292, a plaintiff lost because the writing he produced was illegible to the court.115	36
The argument that the writing requirement in covenant emerged in the later 1270s from the elimination of wager of law runs counter to the position taken by those legal historians who, unaware of the changes in proof, have believed that the requirement was not in place until the early fourteenth century.116 This position is based largely on three cases, one decided in 1292 and two in 1304, in which the court permitted plaintiff to proceed without producing a written agreement. The two 1304 cases, brought by lessors, were discussed earlier.117 Once the report and record of the 1292 case are sorted out, it becomes clear that the case was not one of covenant.	37
The confusion arises from two printed reports of a case in the Shropshire Eyre of 1292. The two reports have been assigned to the wrong plea roll entry. The assignment of reports to record seems to have been made on the basis of the names of the parties rather than on the basis of the substantive issues. There was, however, a second similar case in the Shropshire Eyre. The plea roll record of this second case fits the reports better than the record of the case to which the reports have been assigned.118 When the reports are assigned to the right record, the record reveals that the case was treated as one of detinue or debt. Not surprisingly, plaintiff did not have to produce a writing and defendant waged his law.	38
One report describes Thomas Corbet bringing a bill of covenant against B complaining that Thomas and B had covenanted that Thomas would bail his horse worth £20 to B for a joust on the condition that if the horse were maimed or killed, B would pay Thomas £20.119 Thomas bailed the horse to B, the horse was maimed and then died while in B's keeping, but B refused to pay the £20. Thomas sued. B demanded judgment when Thomas could produce only suit. No judgment is reported. A second report describes Thomas Corbet bringing a bill against Richard Scurrye and Richard Fitz Pride for £20.120 This report treats the bill as a case of debt. Defendant first argues unsuccessfully that plaintiff must proceed by writ, not bill, because his claim exceeds 40 shillings. He then asks for proof of the debt and plaintiff says he has good suit. Defendant then moves on to a jurisdictional point that the case is one for the law of arms but is overruled on the grounds that the law of arms applies only where both parties participated in the joust. Having exhausted his preliminary points, defendant answered that he had indeed within three weeks returned the horse in good condition. The parties go to a jury on this issue.	39
These reports have been assigned to the record of a case between Thomas Corbet and Richard Scurrye.121 The record of the case has Thomas claiming that on 23 January 1292 he delivered a horse to Scurrye and to one Richard Pryde on the condition that if the horse was maimed or died in their custody, they were to pay him 30 marks on 2 February 1292. The horse died and Thomas demanded the 30 marks. They refused and he claimed damages of £40. The parties settled: plaintiff accepted 20 marks. Four things about this record are worth noting. The record does not mention a joust or how the horse was injured, but the reports mention a joust. The record names the sum of money as 30 marks, not as £20, although the two locutions name the same value. The record names a date when payment was due, the report does not. The record does not mention when, if ever, the horse was returned to plaintiff, but in one report defendant says he returned the horse within three weeks.	40
There was another case on the Shropshire Eyre about a horse.122 The record of this case says that one Richard Sturmy made a plaint against Roger Folyot that on 23 January 1292 they covenanted that if one of them injured the other's horse, he would compensate the other for the injury. Roger injured Richard's horse by hitting him with a lance to the head. Richard brought the horse to Roger because of the injury and Roger accepted him on the condition that if he could not make the horse well he would pay Richard £20, the price of the horse, and if he could heal the horse, he would return it. Richard kept the horse for one month, after which the horse died. Richard claimed damages of 20 marks and produced suit. Roger denied that he had received the horse or that they had agreed ("inter eos convenit") that he would pay Richard £20. Defendant made his law and plaintiff was amerced for a false claim.	41
The reports fit this record better than they fit the record of Corbet v. Scurrye. The two reports and this record name the price of the horse by the words twenty pounds, not thirty marks. The reports mention a joust and this record describes an injury to the horse of the sort a horse might well suffer in a joust--a blow to the head by a lance. The record also describes an indemnity agreement that cautious participants in a joust might find prudent. The second report has defendant invoke the law of arms, which is plausible if one infers from the indemnity agreement that both parties participated in the joust. The confusion of the reporter in using the names of the parties of a similar case is understandable.123 Both agreements were made on the same date. Thirty marks is equal to twenty pounds. Both cases involved the bailment of a horse. The amount claimed in damages in the second case is the same as the amount of the settlement in the first case. The confusion between characterizing the plaint as covenant or as debt is also understandable. The indemnity agreement in the second case sounds like a covenant, although it has nothing to do with plaintiff's ultimate claim. In the record of the case, defendant denied receipt of the horse, appropriate in an action of detinue, and denied the agreement to pay £20, appropriate in an action of debt. Viewing the plaint as detinue or debt, the justices accepted plaintiff's production of suit and defendant's wager of law.	42
The Reasons for the Writing Requirement. Because the writing requirement emerged from the removal of suit and wager of law in the late 1270s, the reason for the writing requirement has two parts. First, why did plaintiffs before the late 1270s have to produce suit or a writing in covenant? Second, why was suit for plaintiffs and wager of law for defendants removed in the late 1270s? An answer to this second question should also explain why suit and wager of law were removed from actions of covenant but not from actions of debt or detinue.124	43
Suit or a writing proved that defendant had undertaken a special obligation to plaintiff, an obligation not imposed generally by common law or statute.125 This reason for suit or a writing in covenant was also at work in debt and detinue, where plaintiff had also to produce suit or a writing. The writ of debt is almost as confusing as the writ of covenant and for much the same reason: at least at its beginning debt was a hybrid action. Some claims of debt, those arising from transactions such as sales and loans, were based on the idea of obligation.126 Other claims of debt were more like claims to property based on a grant. The debtor granted a debt to his creditor, frequently, but certainly not always, by means of a writing.127 The introduction of trespass writs in the 1220s and 1230s to enforce a general obligation not to breach the king's peace changed the conceptual landscape.128 Although the new trespass writs were introduced piecemeal, they nevertheless were based on the idea of a general obligation enforceable in the royal courts. Their introduction could clarify the concept that debt and covenant were based on a special obligation. In trespass, plaintiff did not need to prove the general obligation not to breach the king's peace. The very fact that he obtained a writ meant that there was such a general obligation. But in debt and covenant, plaintiff needed suit or a writing as proof that defendant had indeed undertaken a special obligation to plaintiff.	44
A similar idea was at work in cases of nominated dower and dower assensu patris. In such cases, suit or a writing was required as proof of a grant that gave rise to a special entitlement, one not created by the common law rules of dower. Although in these dower cases it was a matter of proving a special entitlement under a grant rather than a special obligation, the opposition between special and general entitlements was analogous to the opposition between special and general obligations. The king's court thought of dower as a grant made by a groom to his bride at the marriage ceremony.129 After Magna Carta, however, as long as there had been a marriage ceremony at the church door, a widow was entitled to one third of the lands of which her husband had been seised in fee at any time during the marriage.130 That was her general entitlement at common law. A groom, however, could designate at marriage specific lands as his bride's dower. If she later claimed this nominated dower, she had to produce suit or a charter.131 Although there were some arguments in the last three decades of the thirteenth century that a widow who claimed nominated dower had to produce a writing, they were not successful. The possibility of producing suit was not eliminated from cases of nominated dower as it was from cases of covenant. If a groom had no land when he married, he could endow his bride with lands held by his father with his father's consent. The endowment could be either of common law or nominated dower. This arrangement, known as dower assensu patris raised two issues: whether the widow's husband had endowed her at the marriage ceremony and whether her husband had authority to make a grant of land that was not his own but his father's. A widow had to produce suit or a writing of the endowment and of the father's consent.132	45
That suit or a writing was required to prove a special obligation or a special entitlement comes close to the explanation of the writing requirement in covenant offered by Patrick Philbin.133 He argued that a writing was required in covenant because a covenant was thought to be created by the wills of the parties to be bound to the covenant, the will of a party was thought to be internal to him, and a writing was thought to be necessary to render manifest his internal will. In a case of covenant on the London Eyre of 1321 Serjeant Burton argued that covenant is the will of another and the will of another could only be proved by specialty.134 In another case on the same eyre, Justice Herle spoke of covenant as being a matter of assent that lay in specialty.135 Justice Herle's statement need not rest upon Burton's explanation. Twenty years earlier, when he was a serjeant, Herle argued for defendant in a case of dower assensu patris that plaintiff must produce a writing of the father's assent, because his assent was internal to him and could be made external and evident only by a writing.136 Going the other way, from Herle's argument in assensu patris to Burton's argument in covenant, Philbin reaches the position that a writing was required in covenant because the will to be bound is internal and can be made external only by a writing.	46
There are difficulties with this argument. Plaintiff in dower assensu patris need not produce a charter. In the 1301 case of dower assensu patris cited by Philbin the argument that a writing was necessary to make the father's assent external was countered by Warrewick, who observed that one could render internal assent external by spoken as well as written words.137 Witnesses, that is, suit, were as good as a specialty. Justice Hengham agreed. At that point, Herle, who had insisted on a writing, gave up his point and offered to produce witnesses that the father had not assented. But Mutford pointed out that proof against proof ("prove contre prove") was the procedure in ecclesiastical, not common law, courts. The case went to a jury with plaintiff having produced suit but no writing. By 1301, the practice in cases of assensu patris of accepting plaintiff's production of suit and sometimes adding the suitors to the jury was almost a century old.138 And the practice continued in cases in 1308 and 1315.139	47
A more serious difficulty with Philbin's argument is that he sets up the wrong oppositions. The relevant opposition is between special and general obligations, not between internal will and its external manifestation. In assensu patris, the issue was not whether the father entered into an obligation but whether he authorized his son to make a grant to his bride. If the son made an endowment of less land than that to which his father had assented, his son's widow did not have an action for the larger amount of land but only for the endowment actually made. The converse, however, was true. If the endowment was greater than that authorized by the father, the son's widow could recover only the endowment authorized by her father-in-law.140 In dower assensu patris as in nominated dower plaintiff had to produce suit or a writing because she was claiming a special entitlement, not a general entitlement provided by common law. In covenant, according to one note, plaintiff, "allegat speciale factum vel contractum."141 After the later 1270s, plaintiff needed a writing as proof of the special obligation.	48
The second part of the reason for the writing requirement is why suit and wager of law were removed from covenant in the later 1270s. Two different lines of thought converged to make the writing requirement a conscious change in the law.142 One had to do with the relationship between central and local courts. The other, with when trial by jury was appropriate.	49
Suit and wager of law were removed from covenant as part of a fairly extensive revision of the modes of proof used in the central courts. At the same time as royal officials removed wager of law from covenant, they removed wager of law from mesne, replevin, actions on writs of annuity, and actions to enforce final concords.143 At this time, too, suit and wager of law as the means of trying whether defendant had paid the debt were removed from actions of debt and there was some effort to restrict wager of law in debt to denials of having incurred the debt.144 These revisions in modes of proof were themselves part of wider efforts by royal officials to restructure the relationship of central to local courts. In the later 1270s royal officials took steps to tie central and local courts more closely together. From 1274 to 1281 the writ of pone, used to remove cases from county to central court, reached its peak.145 In 1275 recordari writs began to be used frequently to remove plaints of trespass, replevin, and debt from county to royal court.146	50
The moves toward greater integration of local with central courts were reversed and the dominant theme soon became to keep the two kinds of court clearly separate from each other. In the 1280s the increased use of pone subsided. By the 1290s the expansive use of the recordari writ had been cut back so that the writ was restricted to the removal of replevin cases.147 Royal officials also designed amount-in-controversy requirements for cases in the central courts. Chapter Eight of the Statute of Gloucester in 1278 confirmed and perhaps stiffened a requirement that trespass cases in the central courts involve at least forty shillings in damages.148 This floor for trespass actions in the central courts was extended without statute to cases of debt and detinue and in these cases became a ceiling for litigation in county court by the 1290s.149 The adoption of the writing requirement in covenant, through the elimination of suit and wager of law, might be a reason why the forty-shilling rule was not extended to covenant.150 At any rate, the revision of proof in covenant and in other cases helped to keep central and local courts separate and distinct from each other. Wager of law was a traditional mode of proof in local courts. By severely restricting suit and wager of law in the central courts, royal officials expelled from the central court modes of proof traditional in local courts. This was an effort to make central royal court procedures different and distinct from what went on in local courts.151 A late thirteenth-century treatise argued that claims based on specialties, that is to say written instruments, could not be tried in courts baron because in those courts defendant had to be able to wage his law.152 This view might have been the result of removing suit and wager of law from a number of actions, including covenant, in the royal courts.	51
This institutional reason for removing suit and wager of law from covenant converged with new thinking by judges and lawyers as to when trial by jury was appropriate. The underlying idea was that where alleged facts were sufficiently public to be cognizable by a jury, a party should not be permitted to wage his law. The idea can be seen at work precociously in a case of covenant as early as 1248.153 Plaintiff, a lessee, claimed that defendant had leased lands to him for a term of twenty years but had ejected him during the term. Defendant answered that plaintiff had assigned the balance of the term to a third person who had bequeathed it to his widow who had sold it to defendant, the original lessor. Plaintiff denied that he had assigned the entire balance of his term to the third party: he had assigned only part of the outstanding term. The court asked plaintiff whether he accepted trial by jury. Plaintiff did so. Then the court ruled that plaintiff could not wage his law in such a case. The parties settled. Plaintiff could not wage his law, not because wager of law by plaintiffs in covenant was impermissible, but because the involvement of a third party and his widow made the facts of plaintiff's denial sufficiently public to preclude wager of law.154 This idea also informs a case of debt in 1275.155 Defendant said that plaintiff's wife in her testament and with plaintiff's consent had authorized defendant to pay certain legacies to third parties in satisfaction of his debt to plaintiff. The written authorization had been destroyed by fire. Defendant produced suit and put himself on a jury. Plaintiff was ousted from waging his law. Here, too, the involvement of third parties assured the requisite publicity for trial by jury. The idea that public facts should be proved by trial by jury was applied sporadically before the later 1270s. The decision to apply the idea systematically was not taken until the 1270s and 1280s. Arguments articulating the idea clearly appear in the 1290s.156	52
This idea would apply most strongly to wager of law to deny the breach of a covenant. A denial of a breach of covenant was frequently a denial of facts cognizable by a jury because of their publicity. Did the defendant commit waste? Did he prevent plaintiff from taking estovers? Did he convey the land to plaintiff and put him in seisin? Did defendant build the bell tower? In a number of cases of covenant defendant waged law to deny a breach where plaintiff's claim mimicked a claim on a writ of mesne.157 It would have been strange to eliminate wager of law in mesne but not in actions of covenant that mimicked mesne. The idea that trial by jury was the mode of proof appropriate for proving public facts applied less strongly to the making of covenants. But once the decision was taken to remove wager of law from denials of a breach, extending the decision to wager of law to deny plaintiff's description of covenant could have appeared necessary for the sake of uniformity in actions of covenant. The value of having uniform rules for the writ was further secured in that, once wager of law was removed, all actions of covenants would be treated in the same manner as actions by lessees, the type of covenant case in which a writing was already required.	53
The reason why suit and wager of law were removed from covenant but not from debt lies in the difference between wager of law in debt and in covenant. Wager of law in debt was a more entrenched practice than was wager of law in covenant. For the period from 1220 to 1275, I have been able to find only twenty-three cases in which defendants waged their law in covenant.158 Instances of wager of law in debt were far more frequent. The overwhelming majority of plaintiffs in covenant already had written evidence of their covenants.159 Eliminating wager of law in covenant extended the practice of producing written covenants to all cases of covenant and thus brought all cases of covenant on a par with actions by lessees. The same cannot be said for debt. Until perhaps the very late thirteenth century, plaintiffs in debt who produced a writing of the debt were probably not even in the majority. Nor were there claims in debt for which plaintiff was required to have a writing. That wager of law in debt was more deeply entrenched made the introduction of a writing requirement in debt a more radical change in the law. Denials of a breach of covenant fit under the idea that public facts should be proven by trial by jury. Royal officials could move from eliminating wager of law to prove a denial of the breach to eliminating wager of law to prove a denial of the covenant. They could thus substitute trial by jury for wager of law. But neither part of a debt transaction--incurring the debt or paying it--need be public acts cognizable by a jury. Royal officials could not substitute trial by jury for wager of law in debt as they did in covenant, mesne, replevin, and actions on final concords. Where they could remove older modes of proof in debt, they did so. Royal officials eliminated suit and wager of law as the mode of proving payment. The requirement of an acquittance where plaintiff produced a writing was established because trial by jury was inappropriate under prevailing ideas. And there was the appeal of symmetry to require defendant to produce a written acquittance where plaintiff produced a written debt.	54
C. Actions to Recover Freehold or Fee
Actions of covenant to recover land in freehold or in fee presented special difficulties because there was no consensus until perhaps the first decade of the fourteenth century on whether plaintiff's claim in this type of covenant case should be treated as a proprietary claim, in which case plaintiff had to produce a specialty, or as a claim of obligation, in which case until the 1270s plaintiff need produce only suit. Some judges and lawyers who believed that plaintiff's claim was one of obligation could not understand how plaintiff could recover land in freehold or fee in an action of covenant. They took the position that the writ of covenant simply could not be used to recover freehold or fee.	55
If plaintiff's claim were based on the special obligation defendant had undertaken to convey land in freehold or fee, plaintiff should be able to produce either suit or a writing of the covenant. If plaintiff produced suit, defendant should be able to wage his law to deny the covenant. That is what one finds until the later 1270s.160 In the 1280s, lawyers turned the view of covenant as based on a special obligation against the recovery of freehold or fee. They said that because covenant was a personal action that led to the recovery of damages or chattels, the writ could not be used to recover land in freehold or fee.161 The underlying reason against using covenant to recover land in freehold or in fee was that a plaintiff who claimed freehold or fee made a proprietary claim that in most cases had to be based in part on the fact that he or his ancestor had been seised of the land. In cases in 1236 and 1285 defendant argued that covenant should not be used to recover freehold or fee because plaintiff could not plead his own or an ancestor's seisin.162 In cases in 1243–44 and 1247, the court rejected plaintiff's claim because he had only a charter of feoffment and defendant was still seised of the land.163