2. Commodatum (Loan for Use)

First a general comment, on leaving mutuum behind. The remaining three real contracts have much in common and stand somewhat apart from mutuum. Mutuum is stricti iuris, unilateral, entails transfer of owner­ship, and torn from the condictio is the founder member of the category of contracts re, its only member in Gaius's Institutes.

The basic idea of commodatum has already been introduced: the loan of a pen, a knife, a horse, something to be returned in specie. The word itself is quite slippery. The simplest thing is to say that the noun commodum means �an advantage, a profit, a favourable opportunity'. Then, the adjective commodus means �pleasing, advantageous, suit­able'; our �commodious' is horrible now, abused by estate agents. The verb commodare thus means �to put (something) at (someone's) con­venience' and hence �to lend for his use'. The main action (actio commodati directa) lay against the borrower to enforce his obligations to keep safe the res and to return it. In its bonae fidei version the action had this formula:

Whereas Aulus Agerius commodated the... to Numerius Negidius, which matter is the subject of this action,

whatever on that account Numerius Negidius ought to give to or do for Aulus Agerius ex fide bona,

for the value of that let the judge condemn Numerius Negidius to Aulus Agerius; if it does not appear, let him absolve.^[34]

However, Gaius tells us (G.4.47) that commodatum is one of the events for which the praetor published model pleadings both in ius concepta (framed on the law: the one just given) and in factum concepta, framed so as to do more than tell the story, without asserting any legal inference about an oportere (an �ought’).

If it appears that Aulus Agerius commodated the... which is the subject of this action to Numerius Negidius and that it has not been returned to Aulus Agerius,

for as much as it shall be worth for so much let the judge condemn Numerius Negidius to Aulus Agerius; if it does not appear let him absolve.

The existence of both types offormulae and the active survival of the almost certainly older actio in factum is precious evidence for and, at the same time, a puzzle in the story of the development. It also raises interesting questions about the relationship between law resting on praetorian imperium and the non-praetorian ius civile. However these large questions have to be laid aside in favour of a brief account of the contract itself. I shall use the formula in ius concepta and shall stick to the method which deals separately with demonstratio and intentio.

i. The demonstratio

�Whereas AA commodated the...’: There is not much that needs to be said. Ulpian thought the same. Expounding the words of the edict, not here the formula, he observes �Huius edicti interpretatio non est difficilis: the interpretation of this edict is not difficult’ (D.13.6.1.1 = Ulpian, 28 On the Edict). And the one point he feels constrained to take is that the word commodatum is strictly apt only for movable things. He says Labeo’s view was that utendum dare (to give for use) was the genus with commodare the species peculiar to movable goods. If that was right as a matter of nice linguistic usage the edict was given the generic scope, to include land too. Ulpian cites Cassius and Vivianus for that.

The exercise of differentiation has mostly been done under other contracts. If the recipient gives a reward (merces) for the benefit of borrowing the plough the contract is hire, not commodatum.

ii. The intentio

What obligations follow? The lender's main concern is to get the thing back at the end of the loan. This shows up well in the formula in factum concepta in which the complaint is precisely that the res has not been given back: eam rem redditam non esse. Return is thus the borrower's main obligation.

But there is also the question of care during the period of user. What if the thing is given back but in a bad state? And this bears back on to the obligation to return. For what if the res has been destroyed or taken away by theft and cannot be returned for that reason?

This is a difficult matter. It turns on that warranty liability for safe­keeping called in Latin the liability custodiam praestare: to answer for security (but both words are difficult). There are two questions about this security-liability: What did it entail? and, When did it apply? The texts have been interfered with to a degree which remains exception­ally controversial.

It is a good policy when one sees a horrid instability approaching to say first whatever can be said with certainty. That way the doubt can be penned in. Otherwise more seems unintelligible than really is.

2. Now we must ask what is caught by the action on commodatum. Is it dolus (hostile intent), culpa (fault) or omne periculum (insurance liability for every disaster)? Under contracts we sometimes have to answer for dolus only and sometimes for culpa too. In the case of deposit, it is dolus only. That is quite right, because no advantage (utilitas) accrues to the depositee... But where advantage accrues to both sides, as in sale, hire, dower, pledge and partner­ship, there the liability is for both dolus and culpa. 3. In commodatum, however, the advantage generally accrues solely to the borrower. Which is why the opinion of Quintus Mercius is the more correct. His view was that the borrower must answer for culpa and must therefore show diligentia (care, attention); and further that in a case where the res is handed over subject to a money valuation he must also answer for omne periculum in that he will have undertaken to make good the money value at all events. 4. On the other hand nothing is to be put down to the borrower’s account where the res is overtaken by old age or by disease or dragged off by robbers, or if something else of that kind happens, so long as no fault is found on the borrower’s part. In the same way if a blaze or the collapse of a building or some other fate- destined loss is the cause of some disaster, then the borrower will not be liable.

The next pair is interesting because the first gives an extract from one jurist, Julian, and the second shows a later writer, Ulpian, citing and qualifying the earlier utterance. D.13.6.19 (Julian, Digest):

People who hire the job of keeping something or borrow something for use (qui servandum aliquid conducunt aut utendum occipiunt) do not, and this is far from being in doubt, have to bear the consequences arising from damnum iniuria datum (the delict of �loss wrongfully caused’) committed by another: for by what care or attention can we make sure that nobody commits wrongful loss against us?

Then D.19.2.41 (Ulpian, 5 On the Edict):

But Julian holds that you cannot sue him in respect of loss inflicted by another: for by what safe-keeping could you ensure that nobody did you wrongful loss? But Marcellus says it sometimes can work out like that. As where he could have arranged a guard or the guard he arranged himself did the wrongful loss. And this opinion of Marcellus ought to be followed.

These texts and others not dissimilar may have been tampered with. But if we use them to support the proposition that the borrower’s liability went beyond dolus and extended to culpa, we will be safe. It is only when we ask whether they show that the borrower could not be liable without at least the proof of culpa, i.e. show also that there was no strict liability, that we have to treat what they say with exactissima diligentia (the most exacting attentiveness).

One more thing can be said with confidence. If the borrower went outside the scope of the contract, as by putting a slave to work up a scaffolding when he was lent for gardening or by taking a horse into battle when lent for an ordinary journey, he would have to answer for every disaster, omne periculum, irrespective of culpa. Furthermore, though this is a separate matter, even if no disaster supervened, he would face a delictual liability for theft if he had dishonestly extended the use he had been allowed.

This brings us to security-liability (custodiam praestare). First, what was it? Second, did it at any time apply to the borrower by commodatum?

Security-liability was a species of strict liability; that is, liability independent of fault or, more accurately, of the proof of fault. But it was not an absolute liability; that is, it was not imposed totally without regard for the circumstances. A person subject to security-liability would be liable without proof of fault but not if the disaster occurred through an event beyond the power of any man to prevent, as death by old age or disease, earthquake, enemy action, incursion of armed bands. These forces are summed up in the phrase vis maior (too great force, �act of God'). Another way of looking at it is to say that one who, either expressly or by implication from the nature of the contract, warrants the security of the thing is understood to guarantee its safety not absolutely but so far as human means can achieve it.

You may say that this amounts to the same thing as fault liability. In nine out of ten cases it will produce the same result. If the disaster was an act of God outside the scope of the warranty, the warrantor will also be free of fault; if the disaster was not from act of God and therefore was caught by the security-warranty, the warrantor will have been at fault. In nine out of ten cases. In the tenth the conclusion will be different. I do everything that a bonus paterfamilias would do. I set two men to guard your plough. Nobody would do more. By an ingenious plan thieves nevertheless get the plough. Let it be that they lure the guards away with false messages apparently from myself. Anybody would have been taken in. Non-violent theft is not vis maior. So if I am under security-liability I must pay up. But I was not at fault. If my liability had been for culpa I would have escaped. It is an interesting question whether it is possible to construct an example the other way about. One, that is, in which I would be liable on a culpa basis but escape when the liability is for custodia. I think it is not. Because security-liability is something super-added to culpa liability. But this is historically, and analytically, more difficult than it seems.

This account is consciously coy about the kind of disasters within security liability. There is room to take one of two positions. The larger is that the warranty for security extended to all harm to the thing (damage and destruction) as well as loss, i.e. disappearance, typically by theft. The narrower view is that only loss, in the sense of disappearance, is in question, while harm is left to the culpa liability. I think the larger view is probably right. This doubt is an instability within the scope of security-liability. That is, whenever we say that security-liability did or did not attach to such and such a bailee (temporary transferee) we are not completely sure whether we mean a qualified strict liability for inability to produce the res or for both that and for damage apparent in the res as produced.

Did the custodia-liability attach to commodatum at any time? There is a good, though complex, argument to the effect that the compilers meant the liability to be limited to culpa except where more was expressly undertaken by the borrower. One way of undertaking omne periculum was to agree to take the res at a valuation, a sum to be restored at all events. But special arrangements between the parties aside, the Justinianic law made liability turn on culpa, fault. In other words, no security-warranty was implied into commodatum.

On the other hand it is almost certain that in classical law the borrower did have this security-liability unless he expressly excluded it. In this connexion the exact words of the formula in factum would be decisive. Did it say just �If it appears that the... was not returned...’? Or did it include some reference to a reason or qualification, as �If it appears that the... through the defendant was not returned...'? Sadly we do not know, though Lenel gives the unqualified version. If Gaius had not chosen the formulae of depositum to be his full example of doubling up pleadings in ius and in factum we would have had the commodatum formulae spelled out in full. And the commodatum case is much more interesting. The liability in deposit is uncomplicated.

However, Gaius himself does provide the best evidence for the custodia-liability of the classical borrower. It is in his discussion of theft. His question is, Who may sue for theft? The general answer is, whoever has an interest in the safety of the thing. But this means, turned about, that sometimes an owner cannot sue because the safety of the thing is of no interest to him. When can that be? When he has the benefit of a contract under which there is a security-guarantee. For then whatever happens he cannot lose the value of the thing by reason of theft. G.3.205-7:

205. Also, suppose a cleaner receives clothes to be laundered or treated, or a tailor get clothes for mending and they do this for a fixed reward. Then they lose the clothes by theft. They, not the owner, have the actio furti. Because the owner has no interest in the clothes not being lost. For in the trial on hire he can obtain his whole interest from the cleaner or tailor, so long at least as the cleaner or tailor has the money to answer for the goods. In fact, if the worker is insolvent then the owner himself has the actio furti. Because, unable to get the value of his interest by that route, he has in this case an interest in the safety of his goods. 206. Everything we have said about the cleaner and tailor applies across to the person to whom we have lent something for use. Those two have to warrant the security of the thing because they earn a reward (merces) from it. In the same way this borrower has to bear the same warranty, because he too takes the profit of its use. 207. What of the depositee? He does not warrant the security of the thing. And he is only liable in respect of it if he himself does something dolo malo (with malicious intent). It follows from this that if the thing is snatched away from him he cannot bring the actio furti. Because, having no liability in the action on deposit in respect of the restoration of the res, he lacks an interest in its safety. So the dominus has the actio furti.

The emphasis on theft comes from the context of the discussion. The inference however is quite clear: the liability of the borrower is such as to make the lender safe without inquiring into fault. That is the whole point.

There remains the actio contraria. What can the lender owe the borrower when the loan is gratuitous and in the borrower’s interest? Paul points out, at D.13.6.17.3,^[36] that as in negotiorumgestio what starts as a kindness becomes a duty once begun. If the lender gives something for a purpose and then leaves the borrower high and dry by recalling the res before the purpose is complete, he must pay his loss. As for example if you lend me a scribe to take down a contract or some scaffolding for building repair, and then, once I am relying on the loan to meet my need, you take the res back. If the res needs special attention, the lender should pay, as where a slave becomes ill and needs medical care. The borrower must meet daily and ordinary expenses, as for instance food. But not exceptional impositions. Again defective res can be more trouble than none at all. Vessels which leak or contaminate the contents cause loss, not commodum. And the lender must answer for it. This liability is referable to the good faith on which the action in ius is built. It appears not to be confined to dolus, despite the lender’s lack of material interest.