Burden of Proof and Cause of Action
Stephen Wexler* and Jack Effron**
Burden of proof must be understood in terms
of burdens not proof. It is one in a series of
burdens the law assigns and every burden as-
signed to one side is a benefit conferred on
the other Burden of proof is substantive rather
than procedural. Who must prove and what
must be proven cannot be separated. The dis-
tinction between burden of proof and cause
of action makes us misunderstand the law.
II faut insister sur l’aspect ondreu du fardau
de la preuve plut~t que sur sa force probante.
Plusieurs fardeaux doivent 6tre r~partis lors
d’une poursuite judiciaire et l’imposition d’un
fardeau sur une des parties en litige confer
un avantage net A l’autre. Le concept du far-
deau de la preuve fait davantage partie des
r~gles de droit substantif que des rfgles de
procedure. La personne sur laquelle repose le
fardeau de la preuve et le fait A prouver sont
indissociables. La distinction que l’on trace
parfois entre le fardeau de ]a preuve et la
cause d’action n’est pas justifie6 et s6me la
confusion.
“* “* ”
Burden of proof is a very simple concept. It is a matter of common
sense. If you are driving your car and you come to an intersection at which
a traffic light is showing both red and green, you may not assume that the
green light is correct and proceed through. You must obey the red light and
come to a stop. When you are able to assure yourself that it would be safe
to proceed, you may then take advantage of the green light and go. The
green light has the burden of proof. The red light has the benefit of the
doubt. These two terms, burden of proof and benefit of the doubt, work
together and since burden of proof only arises when there is a contest of
some sort, the first and most important thing to notice is that burden is the
central notion, not proof. Every burden allocated to one side is a benefit
conferred on the other.
Legal theory has beclouded burden of proof by putting the stress on
proof Actually, burden of proof has very little to do with proof It is a
technique for making decisions in the absence of proof, a technique for
*Associate Professor, Faculty of Law, University of British Columbia.
**LL.B., University of British Columbia (1984).
McGill Law Journal 1984
Revue de droit de McGill
1984]
BURDEN OF PROOF AND CAUSE OF ACTION
469
drawing conclusions from inconclusive evidence. This is the reason why
the notion of a burden never arises in logic, mathematics, or science, three
areas in which proof is at least as important as it is in law. Mathematical
and logical proofs are abstract; they do not depend on evidence at all, and
though scientific proofs do depend on evidence, a scientist is never called
upon to come to a conclusion unless the evidence is conclusive.
Only in law do we have an articulated, well-developed notion of burden
of proof and that is because judges (and jurors with the help of judges) are
regularly called upon to come to conclusions in the absence of conclusive
proof. Consider the phrase: “conclusive proof”; anywhere else but in law,
it is a tautology. It is only in law that “inconclusive proof’ can count as
proof; indeed, inconclusive proof is characteristic of law. If either side to a
legal dispute could really prove its case, the case would never go to trial.
The burden of proof allows us to make a decision where we are not
sure which decision to make. It serves the same function that ffipping a
coin does, except for one thing: the burden of proof is always weighted on
one side or another.
The second thing to notice about the burden of proof is that it is only
one in a series of legal burdens, and while it is the best known of the legal
burdens, it is not the most important. Every legal procedure, and that in-
cludes every form and application, imposes a legal burden, a hurdle in the
way of achieving certain legal results.’ Far and away the most important
burden is that of initiating the proceedings in which proof will count. This
is the burden of suing, filing, applying, petitioning, charging or claiming and,
in general, all other burdens follow it. That is what we mean when we say
that the law is conservative: it places hurdles in the way of whoever asks
it to effect a change in the status quo.
The person who has the burden of initiating the proceedings wil usually
have the burden of pleading, that is, of stating a legally sound claim or cause
of action, and the person who has the burden of pleading, will usually have
the burden of proving as well. There are exceptions to both rules, especially
the latter. Criminal law provides the most obvious examples of legal pro-
ceedings in which the burden of proving does not follow the burden of
pleading; for instance, it is the defendant who must plead provocation, but
‘There is a sense in which legal procedures can be seen as facilitating or even permitting
certain legal rights to be vindicated. This is because the burden of being outside the law is
greater than any of the burdens inside it. “Before the law stands a doorkeeper. To this doorkeeper
there comes a man from the country and prays for admittance to the law.” Franz Kafka, The
Trial (New York: Alfred A. Knopf, 1982) 267.
McGILL LAW JOURNAL
[Vol. 29
the prosecution which must prove the lack of it.2 Tax audits and loyalty
checks are examples of proceedings in which the burden of pleading does
not follow the burden of initiating. From these examples it is easy to see
why this particular break in the usual chain of burdens occurs so infre-
quently. Finally, of course, there is the burden of enforcement; this burden
can only fall on the initiator since he is the one who is asking the law to
effect a change.
It is important to see burden of proof in this context as one in a series
of burdens and to note that, like the burden of proof, all legal burdens favour
one side over another. You always want your opponent to have to do the
applying or petitioning; then the law is on your side to start with.
The third thing to understand about the burden of proof, and in a way
it is the most difficult to grasp because it is the most opposed to traditional
theory, is that the burden of proof is a matter of substance, not procedure.
Legal burdens make certain changes in the status quo less likely and others
more likely. They do not absolutely determine what will happen. If they
did, criminals would rarely be convicted, since nearly all of the burdens in
a criminal case are on the prosecution. What the procedural structure of
criminal law does is make it less likely that a criminal will be convicted;
this is an expression of the substantive policy which is often summed up
in the phrase: “better set ten guilty men free, than punish an innocent one.”
Every legal burden, which means every legal procedure, manifests a
substantive policy decision, a choice to weight the coin on one side or
another. Stressing the proof part of burden of proof, with its connotations
of truth and logic, makes it possible to deny the policy implication of legal
burdens. The only coherent explanations that legal theory gives for burdens
of proof are that they are either self-evident truths (as the phrase res ipsa
loquitur implies) or logical necessities (as is implied by qui affirmat). But
what speaks for itself and who is put in the position of affirming are always
a matter of choice.
The real problem lies in the distinction between substance and pro-
cedure. That there is a close connection between the two has long been
understood. As A.W.B. Simpson has remarked:
[I]t is never quite clear whether the rules of law were sanctioned by an appro-
priate procedure, or whether the rules were developed to explain the existing
procedure; the truth no doubt in many cases was that law and procedure grew
together.3
2 Woohnington v. Director of Public Prosecutions [1935] A.C. 462, [1935] All E.R. Rep. 1,
and see generally P Jones & R. Card, eds, Cross and Jones’ Introduction to Criminal Law, 91h
ed. (1976) 48-52.
3A.W.B. Simpson, An Introduction to the History of Land Law (1961) 43.
1984]
BURDEN OF PROOF AND CAUSE OF ACTION
471
And the statement by Sir Henry Maine on the subject is quoted so often
that it has almost become trite:
[S]ubstantive law has at first the look of being gradually secreted in the in-
terstices of procedure….4
Now both of these quotations minimize the distinction between sub-
stance and procedure, but they both allow that the distinction exists, and
in one sense it does. Salmond described this sense best:
Substantive law is concerned with the ends which the administration ofjustice
seeks; procedural law deals with the means and instruments by which those
ends are to be attained. The latter regulates the conduct and relations of courts
and litigants in respect of the litigation itself; the former determines their
conduct and relations in respect of the matters litigated. Procedural law is
concerned with affairs inside the courts of justice; substantive law deals with
matters in the outside world.5
The distinction Salmond draws is a real one, but it has nothing to do
with legal theory. As Salmond uses the terms, the substance of the law is
that I own my shirt, that I must pay my bill in a restaurant or that I may
discipline my son. In this sense, the substance of the law is a social fact,
not a legal one. The distinction between substance and procedure makes
sense so long as you are talking only about the way in which people conduct
their ordinary affairs. As soon as lawyers get involved in the law, the dis-
tinction between substance and procedure vanishes. When lawyers draft
contracts, draw up wills, register deeds or give tax advice, they are acting,
to use Salmond’s phrase, “in the outside world”. They always have one eye,
however, on what will happen should their work wind up “in the courts of
justice”. For a lawyer, substantive legal questions all come down to this: “if
this matter becomes the subject of a dispute, will my clients have the benefits
or the burdens?” His job is to make sure that his clients have the benefits,
rather than the burdens, and in this sense he makes no distinction between
substance and procedure. From a legal as opposed to a social point of view,
the substance of law is equivalent to its procedure. The substance exists in
procedure and nowhere else.
This was better understood when the writs were still in use. Take, for
example, the creation of what we now call future interests. These were first
recognized by the statute De Donis Conditionalibus in 1285.6 To put it in
the simplest terms, this statute provided that when a “gift was given upon
certain conditions, what we now call a fee tail, the courts were obliged to
4Sir Henry Sumner Maine, Dissertations on Early Law and Custom (1883) 389.
5J. Salmond, Jurisprudence (1902) 577, as quoted in Risinger, “Substance” and “Procedure”
6The Statute of Westminster the Second, 13 Edward I, c. 1 as reproduced in T. Tomlins, ed,
The Statutes at Large (1811), vol. 1, 83 et seq.
Revisted (1982) 30 UCLA L.R. 189, 197.
REVUE DE DROIT DE McGILL
[Vol. 29
give effect to those conditions. And as the statute unashamedly admits, there
was only one way to give effect to legal rights in the Middle Ages –
a new
writ had to be created:
And forasmuch as in a new Case new Remedy must be provided, this Manner
of Writ shall be granted to the Party that will purchase it.7
In order to give effect to the rights of reversioners and remaindermen,
De Donis created the writ offormedon. In that writ one could allege:
(I) that the manor in question was given to a certain man and his wife and
their begotten heirs, or
(2) that it was given to a certain man and the heirs of his body begotten, to
descend after his death as it ought according to the form of gift, or
(3) that it was given to a man and his wife in “frankmarriage” and, at their
deaths, to their named son to descend thereafter according to the form of gift
(forman donationis).
The particularity of these allegations is important because, unlike the
modem “writ”, the medieval writ was more than just a summons. Each
writ was a pre-packaged case which included a very specific set of allegations.
At trial, the case that was presented had to follow the allegations in the writ
and could not differ from them in even the slightest detail. If the plaintiff
did not prove exactly what was alleged in his writ, the writ would “abate”
and the case would be dismissed. 9 As Bereford J. said in Bardolfv. The
Prioress of B.:
[A] man must count according to the facts of his case and suit his count to
his right.’0
Notice that “right” is used here to mean the same thing as “writ.” That
is what we mean when we say that, in medieval law, the writ “embodied”
the right. But notice, too, that the writ embodied the right by setting out
the burden of proof which had to be met in asserting it. This burden of
proof was contained in the pre-packaged allegations which defined the cause
of action.
7The Statute of Westminster the Second, De Donis Conditionalibus, (1285) 13 Edward I, c.
1, s. III., ibid., 85.
sThe Statute of Westminster the Second, De Donis Conditionalibus, (1285) 13 Edward 1, c.
1, s. III, ibid., as translated by the authors.
9See, for example, (1302) 30 Edward I Cornish Eyre, in A. Horwood, ed., Year Books of the
Reign ofEdwardI (1863) 182, where the plaintiffwas non-suited for bringing a writ offormedon
which did not match the actual form of the gift, even though he claimed that there was no
merit in the Registry to describe his conveyance.
‘0Year Book (1308-9) 2 Edward II, as translated by FW. Maitland in Yearbooks of Edward
H1, vol. 1, Seldon Society (1903) vol. 17, 118, pl. 57.
1984]
BURDEN OF PROOF AND CAUSE OF ACTION
473
Legal theory calls these pre-packaged allegations the “cause of action”
rather than the “burden of proof”. Who must prove is said to be procedural;
what must be proven is said to be substantive. The distinction is not helpful.
Who must prove and what must be proven are too closely linked to be
separated. The attempt to keep them separate serves no function except to
disguise the policy nature of choices about the allocation of burdens. It also
creates intellectual tangles. We will come to the tangles later, but before we
do, it is necessary to demonstrate two things. First, that one could say
everything one wished to say about a cause of action in terms of the burden
of proof. Second, that talking in terms of burden of proof would enable one
to say some additional things about a cause of action, some things that
would be missed if the two concepts were kept separate.
This demonstration could be made in terms of any cause of action; let
us take nuisance as an example. I I This action arose out of novel disseisin.
Under novel disseisin the question for the assise was whether the defendant
had iniuste et sine iudicio (“unjustly and without judgment”) disseised the
plaintiff of his freehold. The question in nuisance was whether, “unjustly
and without judgment”, the defendant had done anything to the “harm”
of the plaintiff’s freehold. 12 (The word for “harm” in the Latin writ was
nocumentum, with our word “nuisance” developing through the French
from nocumentum.13)
Since nuisance came from the same writ as novel disseisin, the burden
of proof in the two actions was the same. As in novel disseisin, the plaintiff
had to show first a freehold and then the nocementum, or harm. If he was
able to show both, he succeeded and, in theory, the strictness of pleading
and proof which applied to the writ of novel disseisin applied to the writ of
nuisance as well. The assise had to declare that the case was sicut breve dicit
(“just as the writ said”) or the plaintiff failed. Thus, one ancient plaintiff
lost his action because, although the assise agreed that the defendant had
blocked off the plaintiff’s right of way, he had not done it where the plaintiff
had said he had. 14
“This analysis of nuisance follows Coquillette, Mosses From an Old Manse: Another Look
at Some Historic Property Cases About the Environment (1979) 64 Cornell L.R. 761; Provine,
Balancing Pollution and Property Rights: A Comparison of the Development of English and
American Nuisance Law (1978) 7 Anglo-American L.R. 31; and Hecht, From Seisin to Sit-in:
Evolving Property Concepts (1964) 44 B.U.L.R. 435.
2 G. Woodbine & S. Thorne, Bracton on the Laws and Customs of England (1977), vol. 3,
192.
13J. Baker, Manual of French Law (1979) 148.
14D. Stenton, Rolls of the Justices in Eyre: Being the Rolls of Pleas and Assizes For Yorkshire
in 3 Henry Il (1218-19), Seldon Society (1937), vol. 56, 173, pl. 404.
McGILL LAW JOURNAL
[Vol. 29
But the writ of nuisance presented a problem which the assise of novel
disseisin did not, a problem that tended to erode the theoretical strictness
of the writ system. Just as there is a moral element in the question of what
constitutes “title”, which is absent in the question of what constitutes “pos-
session”, so there is a moral element in the question of what constitutes
“harm” that is absent in the question of what constitutes “title”. Any in-
terference with your neighbour’s title is theft and there can never be any
justification for it. The same cannot be said of the “de facto” theft of nuis-
ance; some harmful interference with the property of your neighbours is
inevitable and all harmful uses of land are not treated as if they disseised
the neighbouring owners. Unlike title, nuisance is a balancing of interests
rather than a determining of rights.
The long common law history of nuisance evidences an effort to draw
the line between a harmful use the law will stop and a harmful use it will
allow to continue. This history can be expressed, as the history of any
substantive legal question can be expressed, in terms of the burden of proof.
In 1611 Aldred’s Case’5 set out the basic burden. The defendant had built
what is delicately referred to as a “house for hogs” next to the plaintiff’s
“habitation of man”. The harm to the latter was so obvious that the court
found the action “well-maintainable” and let the trial judgment of damages
stand:
[F]or in a house, four things are desired habitatio hominis, delectation inha-
bitantis, necessitas luminis, et salubritas aeris, and for nusance done to three
of them an action lies…
16
It was for a nuisance done to the delectatio inhabitantis, the delight of the
inhabitants, that an action was said not to lie, because “the law does not
give an action of such things of delight”.17
Legal conclusions are always expressed in this way, as if they were
matters of fact (the “law does not give an action”). There is a choice being
made here however. The law could have given an action for the delight of
the inhabitants, but chose not to. The choice can be expressed as a burden
of proof If the plaintiff proved that his house had become unlivable, he
succeeded (“for that is the principal end of a house” 8); if all he managed
to prove was that his house had become less delightful, he failed.
1577 E.R. 816 (K-B.). Note that Aldred’s Case and the cases that follow arose under the writ
of trespass on the case, which replaced the assise of nuisance in the 1600’s.
16Ibid., 817.
7Ibid., 821.
I81bid., 817.
1984]
BURDEN OF PROOF AND CAUSE OF ACTION
475
But an important dictum also came out of Aldred’s Case, one which
influenced the burden of proof far beyond the specifics of harm outlined
above.
And this stands with the rule of law and reason, sc. Prohibetur ne quisfaciat
in suo quod nocere possit alieno: et sic utere tuo ut alienum non laedas.19
The Latin phrase means: “It is prohibited that anyone should do on his
own [land] what can harm a stranger; and thus use [your land] so that you
do not harm others.” It is the clause after the colon, the sic utere clause as
it came to be called, which, with some objection, has been quoted most
often as the burden of proof in nuisance law. In 1629, Jones v. Powel 20
applied the sic utere maxim in a case where a coal-burning brewery had
driven the plaintiff from his home with smoke and “unwholesome vapours”.
[A]lthough sea-cole be a necessary fuell to be used, and that brew-houses are
necessary, yet the rule in law is, sic utere…: and chimneys, dye houses, and
tan-fats are also necessary, but so to be used, that they be not prejudicial to
their neighbors. 2′
Jones v. Powell established the rule that the plaintiff had only to prove
the harm and the defendant’s act, for the defendant would not be heard to
say that the nuisance was “unavoidable” or necessary to the public good.
Baten’s Case,22 decided at almost the same time as Aldred’s Case, further
reduced the burden of proof for plaintiffs by quoting the maxim lex non
requirit verificare quod apparet curia. In other words:
the plaintiffs need not in this case assign any special nuisance, for here it appears
to the Court that it is to the plaintiffs’ nuisance;.. .for in this case the defendant
has built a new house, which overhangs part of the plaintiffs’ house… so that
of necessity the rain which falls from the new house must fall upon the plaintiffs’
house… . [A]nd that which appears to the Court need not be averred…. 23
In 1701, in the case of Jones v. Hammond,24 the defendant asked for
a non-suit because the plaintiff’s declaration had admitted that the defend-
ant was possessed of the close on which the nuisance had arisen and for
that reason “such a general declaration is only good against a wrong doer …25
The defendant was arguing that if he acted on his own land, the plaintiff
had to prove more than just the act and the damage; he had to show that
the defendant was in some way a “wrongdoer”, but Holt C.J. rejected this
argument.
191bid.
20123 E.R. 1155 (C. P.).
21Ibid., 1155-6.
2277 E.R. 810 (KIB.).
23Ibid., 811.
2492 E.R. 2 (KB.).
25Ibid.
REVUE DE DROIT DE McGILL
[Vol. 29
A defendant made a similar objection before Holt C.J. three years later
in Tenant v. Goldwin,26, and he also lost. At issue was the duty to maintain
a party wall, which was wholly on the defendant’s property. The plaintiff
sued in nuisance to have it repaired by court order, but the defendant said
that the plaintiff “ought to shew a title” if he claimed a right to order repair.
The last day of the term Holt Chief Justice delivered the opinion of the Court,
that the declaration was sufficient. … [T]he reason of this case is upon this
account, that every one must so use his own, as not to do damage to another27
We witness in this case the least onerous burden of proof for nuisance.
The plaintiff did not have to prove entry of the nuisance onto his land nor
even the harm itself (where it could be presumed for him by the court). As
Horwitz says:
In the eighteenth century, the right to property had been the right to absolute
dominion over land, and absolute dominion, it was assumed, conferred on an
owner the power to prevent any use of his neighbour’s land that conflicted
with his own quiet enjoyment.28
The Industrial Revolution brought a change in the burden of proof for
nuisance. Up until the middle of the 1700’s the burden had been getting
progressively less onerous. From then on through the 1800’s the burden got
more and more onerous, as the courts restricted the operation of sic utere
even as they cited it. Economic development was considered essential, so
the plaintiff had to bear a more onerous burden to stop it. One early example
of this can be found in the 1752 case of Fishmongers’ Co. v. East India
Co.2 9, where a wall, built by the defendant, had darkened the window lights
of the plaintiff. Lord Hardwicke C. stated:
[I]t is not sufficient to say that it will alter the plaintiffs’ lights, for then no
vacant piece of ground could be build on in the city… Therefore, take nothing
by the motion. 30
The high-water mark of this wave came in St. Helen’s Smelting Co. v.
Tipping (1865).31 The policy decision which lay behind the raising of the
burden is set out nicely in Lord Wensleydale’s short speech in the House
of Lords judgment The plaintiff, he said, “must not stand on extreme rights…
Business could not go on if that were so.”’32
2692 E.R. 222 (K.B.).
27Ibid., 224.
28Horwitz, The Transformation in the Conception of Property in American Law, 1780-1860
(1973) 40 U. Chi. L.R. 248.
2921 E.R. 232 (Chancery).
3OIbid.
3111 E.R. 1483 (H.L.).
32Ibid., 1487.
1984]
BURDEN OF PROOF AND CAUSE OF ACTION
477
This abbreviated history of the action in nuisance is meant to show
that collapsing the distinction between burden of proof and cause of action
would not hurt our understanding of the substantive law. On the contrary,
uniting “who must prove” with “what must be proven” brings out some
interesting and otherwise hidden facts about the substantive law. For in-
stance, it shows that two things remained constant through the whole long
common law development of nuisance. First, though the action came to be
seen as a private tort action, the plaintiff always had the burden of proving
that the defendant had harmed his property. It was never enough to show
that the factory smoke merely smelled terrible or made his children cough
at school. Nuisance grew out of novel disseisin and it never shed the burden
of proof associated with the real actions.
The second constant in the history of nuisance is even more basic. The
burden of proof gets more and less onerous, but it never shifted. The basic
burden of initiating legal action to stop a harmful use of private property
remained on the person who was hurt. This reflects a very deep substantive
point in our law. The privacy of property is always taken for granted; an
owner of private property is almost never called upon to bear the burden
of defending his claim to exclusive, private possession. A rich man is entitled
to throw away food, even if the people around him are starving.
This represents a choice, of course, just as the laws of zoning and plan-
ning reflect a choice in a different if not exactly an opposite direction. Zoning
and planning laws place the burden on the land owner, not on those he
might hurt. In modem times, the nuisance defendant has been turned into
a planning plaintiff. No longer can the owner just use his land as he wishes
and put the burden on those who wish to challenge him. Now, certain uses
of land are permitted, and the owner who wishes to use his land in other
ways has to apply for permission to do so.
The legal position of the planning plaintiff is instructive. Take a statute
such as the British Columbia Agricultural Land Commission Act)33 which
creates by section 15(2) a “green belt”:
No person shall use agricultural land for any purpose other than farm use,
except as permitted by.. .an order of the commission….
Suppose a land owner wishes to use his property in a new way, a way
which might be a “farm use” within the meaning of the statute, but which
is not certainly so. The Act provides that the commission may hear appli-
cations for other than farm use, so the owner may file an appliction. Or, on
33R.S.B.C. 1979, c. 9.
McGILL LAW JOURNAL
[Vol. 29
the other hand, he may choose not to file an application, in which case
section 34 of the Act governs:
Where the commission believes that a present or future activity or use of land
in an agricultural land reserve may contravene this Act, … the commission
(a) may order the owner or occupant to refrain from the activity or use for a
period not exceeding 60 days, and to make written or oral submissions to the
commission as it requires … ;
(b) may apply to the Supreme Court for an order restraining the owner or
occupant from commencing or continuing the activity or use of land….
Three different procedures are established under the Act: one in which
the owner applies for permission to use his land in the way he wishes; a
second in which he is required to make written or oral submissions after
having been temporarily prohibited from using his land as he wishes; and
a third in which the commission applies to the Supreme Court for an order
permanently enjoining the owner from using his land as he wishes. There
is a sense in which one might say that the same substantive question will
arise in each of the proceedings; namely, is the anticipated use a “farm use”
within the meaning of the statute. But is the burden of proof the same in
all the proceedings? Probably not. If the land owner applies, the burden
would probably be on him to show that the use he proposes is a “farm use”.
If the commission goes to court, the burden would probably be on it to
show that the use was not a “farm use”. Where the burden would lie in the
middle case is unclear.
Whether the land owner is trying to show that the use is a “farm use”
or the commission is trying to show that it is not a “farm use”, the same
statutory provision and the same precedents will be applicable. In this sense,
the two questions are substantively equivalent. But this sense is purely
formal. In reality, particularly in the overall reality of the long range of
cases, the outcome will depend on who has the burden of proof and the
outcome is the substance of the law.
Separating procedure from substance and putting burden of proof in
the former category hides the fact that the allocation of the burden of proof,
like the allocation of any legal burden, gives effect to a policy decision. This
alone would be enough reason to avoid the distinction between burden of
proof and cause of action, but there is another reason as well. You cannot
think sensibly about” who must prove” without thinking about “what must
be proven”. Attempts to do so have generated intellectual tangles. Foremost
among these is the problem of the shifting burden of proof.
No subject in the law is more confusing; indeed, the shifting burden is
one of the few things that everyone in law feels compelled to say he does
1984]
BURDEN OF PROOF AND CAUSE OF ACTION
479
not fully understand.3 4 The one thing everyone does know is the distinction
between the burden of producing evidence and the burden of persuasion.
The difference between the two was first pointed out at the end of the last
century by Professor Thayer of Harvard 35 and you cannot get through law
school without learning it.
There is no doubt in anyone’s mind that the burden of producing evi-
dence can and does shift. The question is whether the burden of persuasion
does. No one is sure, or rather, some people are sure it does and some people
are sure it does not.3 6 The debate is very lively, with both sides exhibiting
intense commitment to their position. But if you stop thinking about burden
of proof as covering solely who must prove, the whole problem vanishes.
It is criminal law which generates the strong ideological pressure for
the saying the burden of proof never shifts. The prosecution in a criminal
case is said to have the burden of proving every element of its case beyond
a reasonable doubt. No phrase is better known than that last one. It is the
“procedural” expression of the “substantive” notion “innocent until proven
guilty”. We take great and justifiable pride in that notion, and when we
criticize other societies as “totalitarian”, we mean that their governments
do not put that burden of proof on themselves.
But the burden of proving criminal guilt is different from the burden
of proving the presence or the absence of civil liability. Because we adhere
to the ideology which says the crinimal defendant can say nothing and still
be acquitted, he is not called upon to explain his behaviour even when the
prosecution has made out a good case against him. But once a civil plaintiff
proves what he must prove to establish his cause of action, there is no
reason in either good sense or conscience why the burden should not “shift”
to the defendant. If he has what the law calls an affirmative defence, he can
be required to raise it, to offer evidence on it or even to persuade the trier
of the facts.
Notice that this shift can only occur at one point in a trial – where
the defendant loses on his motion to dismiss. After that, it is his turn to
put in evidence. He can seek to disprove some part of the plaintiff’s case,
or to prove something outside the plaintiff’s case but whichever he chooses,
the burden of producing evidence has shifted to him. If he tries to prove
something outside the plaintiff’s case, an affirmative defence such as truth
34See, for example, Stark, Burden of Proof- New Light on an Old Problem (1930) 8 C.B.R.
608; Hanbury, The Burden of Proof(1949) 61 Jurid. Rev. 121.
35J. Thayer, Preliminary Treatise on Evidence at the Common Law (1898) 355.
36Dworkin, Easy Cases, Bad Law, and Burden of Proof(1972) 25 Van. L.R. 1151, 1153 is
sure it does not, and Denning, Presumptions and Burdens (1945) L.Q.R. 379, 380 is sure it
does.
REVUE DE DROIT DE McGILL
[Vol. 29
in libel, he also has the burden of persuasion; if he tries to disprove some
element of the plaintiff’s proof, the burden of persuasion will nearly always
remain with the plaintiff.
There is nothing particularly complicated about this. All the confusion
about the shifting burden arises out of the refusal to include what must be
proven in the burden of proof. That refusal hides the fact that the law can
“shift” any affirmative defence over into the cause of action or make any
element of the cause of action into an affirmative defence. This is what
Professor Stone means when he says: the burden of proof never shifts; the
issue to be proven does. 37 But why make this distinction? It does not lead
to intellectual clarity; on the contrary, the distinction between burden of
proof and cause of action makes us misunderstand the law.
37Burden of Proof and the Judicial Process (1944) 60 L.Q.R. 262, 263-4.
