No. 4]
“Autonomous Acts of Things” in Quebec Law –
Legal Adventurism versus Legal Conservatism
M. A. Miner *
In every system of law there is tension between the fault prin-
ciple of liability for personal injuries and the risk principle. The
question is whether liability should follow strictly upon proof that
the defendant caused the harm (variously termed objective, causal,
or strict liability), or whether it should depend further on evidence
that the defendant was culpable in some way (sometimes termed
subjective liability).
It can safely be said that the general tendency since the first
decades of this century has been to strengthen the risk principle
at the expense of the fault principle. But the means used to achieve
this result have not been uniform from one country to another. In
many countries, certain areas of activity in which physical danger
has been a prominent feature have occasioned so much social
agitation as to bring about legislative changes to improve the
prospects of compensation.
Notably in the field of industrial injuries, statutes have been
introduced to deal with the problem of workmen’s compensation
on an administrative basis rather than through litigation in the
ordinary courts. Discarding the fault principle, statutory schemes
of compensation either displace the civil remedy entirely, as in
Quebec, or, as in the United Kingdom, supplement it. The major
defects of this system have been the inadequacy of the scales of
compensation and, in so far as the common law remedy survives,
the continuing hazards of proof.
Similar pressures are at work in regard to highway accidents,
although the legislation in this case has for the most part been
neither as comprehensive nor as widespread as industrial injury
legislation. Moreover –
the legislation in this regard has tended to leave the fault principle
and this is of far greater importance –
* B.A., LL.B., B.C.L., Ll.D., Professor, Faculty of Law, University College,
London.
McGILL LAW JOURNAL
[Vol. 17
formally intact,1 contenting itself with ensuring, through some form
of insurance, that the victim is not defeated by the anonymity or
the indigence of the wrongdoer. Detailed studies have, however.
revealed the inadequacies of this type of legislation as a solution
to the enormous problem of reparation for the victims of road
accidents.
The remedial deficiencies of the law governing reparation for
injuries in industry and on the roads have led to proposals for the
total displacement of the normal civil remedy by insurance schemes,
administered either by the state as a part of a national scheme of
social security, or under the aegis of private insurance companies.
In these schemes, reparation is to be wholly independent of proof
of fault. In effect, they recognise that industry and transport in
modern conditions expose so large a part of the population to the
risk of injury that it becomes a matter of elementary social justice
that all persons injured in these ways should receive compensation
as of right.1a
There is, beyond doubt, a powerful case to be made out upon
these lines for radical legislative reform. But these projects are for
the future. At present in Quebec, the workmen’s compensation
legislation 2 is neither comprehensive nor excessively generous. And
even if it were, that would not justify disregarding the plight of
those persons who do not come within its terms. And Quebec
legislation regarding liability for accidents on the roads,3 whilst it
greatly assists the accident victim by raising a prima facie case
against the owner and the driver of the vehicle involved in the
accident, nevertheless clings to the principle of fault, allowing the
defendant to escape liability, inter alia, where he is able to prove
that the damage is not imputable to the fault of himself, the driver,
or a person on board.3a It is especially essential to take account of
the cases where the circumstances of the injury take it out of the
1 In the practice of the courts, however, the fact that the true defendant
in really an insurance company undoubtedly influences judge and jury alike,
creating a secret sphere of influence for the risk principle, screened by the
verbal formulae of fault liability.
ia The most far reaching proposals of this kind are those contained in the
report of the Royal Commission on Compensation for Personal Injury, New
Zealand, 1967, recommending that all physical injuries, wherever and however
sustained by any citizen of the country, should confer upon such citizen, as
of right, a claim to reparation by the State, the normal remedies of Common
Law being totally superseded by a national scheme of social benefits.
2 Workmen’s Compensation Act, R.S.Q. 1964, c. 159.
3 Highway Victims Indemnity Act, R.S.Q. 1964, c. 232.
3al bid., s. 3(a3.
No. 4]
AUTONOMOUS ACTS OF THINGS
present industrial and road accident legislation. In all such matters,
the victim must look to his civil law remedies. It is desirable, there-
fore, to bring under review, from time to time, the manner in which
the provisions of the Civil Code are interpreted and applied in those
cases in which the plaintiff is dependent upon them.
The Jand’heur Jurisprudence of French Law
In the interpretation of articles of the Quebec Civil Code which
are inspired by the Code Napoleon, the developments in French
jurisprudence an ddoctrine are always relevant. This is especially
true of those provisions for delictual responsability which, in both
Codes, are stated in very general terms. Most particularly, Art. 1054
of the Quebec Code must run the gauntlet of French legal thinking
regarding the corresponding provision, i.e. art. 1384, of the Code
Napoleon, which has engendered in France the remarkable judicial
activity known as the Jand’heur jurisprudence.
Article 1053 of the Civil Code of Quebec 4 (corresponding to
articles 1382 and 1383 of the Code Napoleon) embodies the basic
principle of fault liability which seems to have been the common
law of the Province prior to the codification. As the Quebec legis-
lation referred to above indicates, art. 1053 was not felt to measure
up to the social needs of the twentieth century in regard to industrial
and highway accidents. In the fullness of time, more adequate laws
may find their way onto the statute book. Meanwhile the question
is to what extent other provisions of the Quebec Code, in particular
the provisions of article 1054, provide at least a palliative until such
time as the legislative millennium arrives.
Art. 1054 reads as follows:
He is responsible not only for damage caused by his own fault, but also
for that caused by the fault of persons under his control and by things
he has under his care;
The father, or, after his decease, the mother, is responsible for the
damage caused by their minor children;
Tutors are responsible in like manner for their pupils;
Curators or others having legal custody of insane persons, for the
damage done by the latter.
Schoolmasters and artisans, for the damage caused by their pupils
or apprentices while under their care.
4 Art. 1053: “Every person capable of discerning’ right from wrong is
responsible for the damage caused by his fault to another, whether by
positive acts imprudence, neglect or want of skill”.
McGILL LAW JOURNAL
[Vol. 17
The responsibility attaches in the above cases only when the person
subject to it fails to establish that he was unable to prevent the act which
has caused the damage.
Masters and employers are responsible for the damage caused by their
servants and workmen in the performance of the work for which they are
employed.
Now the first paragraph of this article contains the possibility
of a refuge for those who are defeated by the fault requirement
of art. 1053 – or at least it must have seemed a possible refuge ever
since the time, nearly a century ago, when the French law first
sought to discover in its art. 1384 some relief from the fault principle
embodied in arts. 1382 and 1383 of the Code Napoleon.5 Traditionally
art. 1384, al. 1, C.N. was treated as merely adumbrating
the
specific presumptions of fault, relating to damage caused by an
animal or by the disrepair of a building, which were set out in the
two succeeding articles, 1385 and 1386 (corresponding to Art. 1055,
Quebec C.C.). The subsequent liberation of art. 1384, al. 1, C.N.,
from this limitation is so well known that we may content ourselves
with a summary account here.
It is commonly accepted that the strongest impulse towards the
re-examination of art. 1384, C.N., came from the industrial revolution,
with its increasing toll of injury and death amongst French indus-
trial workers, and the resultant agitation of the victims or their
dependants for a more effective remedy than existed up to then.
Some courts began to see in the words of art. 1384, C.N., which rest
responsability, inter alia, on “le fait… des choses que l’on a sous
sa garde”, a basis for giving judgment in favour of injured workmen
without fault having been proven against their employers. This
development was interrupted by the enactment in France of the
Law on Industrial Accidents,
(1898), which assured to injured
workmen some compensation, however meagre. But the judicial
activity regarding damage caused “by the act of… things that he
has under his care”, while losing some of its impetus, was not
halted. It gathered renewed force with the advent of the motor car
and its grim harvest of road injuries and fatalities, and culminated
5<,Art. 1382. Tout fait quelconque de l'homme, qui cause h autrui un dom-
mage, oblige celui par la faute duquel il est arriv6, h le r6parer.
Art. 1383. Chacun est responsable du dommage qu'il a caus6 non-seulement
par son fait, mais encore par sa ndgligence ou par son imprudence.
Art. 1384. On est responsable non-seulement du dommage que l'on cause
par son propre fait, mais encore de celui qui est caus6 par le fait des
personnes dont on doit r6pondre, ou des choses que 'on a sous sa garde... ).
No. 4]
AUTONOMOUS ACTS OF THINGS
in the great legal battle in the Jand'heur case,6 in which the Cour
de Cassation finally established the new jurisprudence. Art. 1384 C.N.
was there interpreted to mean that the liability of the gardien of a
thing to the person damaged by that thing was defeasible only on
very limited grounds, namely, positive proof of the decisive inter-
vention of a cause external to the defendant and not imputable to
him (cas fortuit, force majeure, or the act of a stranger or of the
plaintiff himself). Moreover, the cause must be external to the thing
itself, so that a defect in the thing which makes it harmful, whether
known or knowable, or not, does not rank as force majeure or cas
fortuit, so as to exonerate the custodian. In short, art. 1384 C.N. was
deemed to sanction not simply a presumption of fault, displaceable
by proof of absence of fault, but a presumption of responsibility,
that is to say strict liability, defeasible in no way at all other than
by disproof of the causal nexus in the requisite manner.7
The risk principle thus enthroned reached its pinnacle in the
years which followed -
"the age of adventures", as Rodi6re called
it.' And indeed the triumph of the causal principle did threaten to
dismantle much of the jurisprudence based upon the fault require-
ment of art. 1382 C.N., and to reduce that article to a very subsidiary
role in the law of civil responsibility. The high water mark in this
development might well be taken to be the first decision of the Cour
de Cassation in Cannot v. Franck.9 It was there held that the gardien
continues in law to be the gardien of the thing even after it had been
stolen, so that he remains responsible to a person injured by his
car whilst it is being driven by the thief. Since then, there have
been some retreats. Thus the reasoning in Connot v. Franck no longer
prevailed in the Cour de cassation when the same case came before
the Chambres r6unies some five years later.10 The court now held
that the gardien in such cases had no longer any power of surveil-
lance over the thing, and, being deprived of the use, direction and
control of the car, could not rightly be regarded as the gardien
any longer.
In this new attitude there is seemingly some revival of fault
theory: an implicit argument that, having regard for the theft of
OJand'heur v. Les Galeries Belfortaises, Cour de cassation, Chambres
rdunies, 13 Feb., D. 1930.1.57.
7 Contrast Quebec law: ,
Insurance Co. Ltd. [1970] C.A. 734, per Brossard, J., at 735. See further below.
s Rodi~re, in Ch. Beudant, Cours de droit civil frangais, (2e 6d.), Tome IX bis,
No. 1497.
9 D. 1936.1.81 (note Capitant).
‘0 Connot v. Franck, DA. 19411.369.
McGILL LAW JOURNAL
[Vol. 17
the car, no blame could attach to the erstwhile gardien. Other
pointers in the same direction are to be found in decisions denying
responsibility in the case of mental illness, and denying responsi-
bility in the case where the thing is said to play a purely passive
role.1’ On the surface, the line of reasoning in this last class of case
is causal –
the assertion that a purely passive thing is not a “gener-
ating cause” of the damage. This does not mean, however, that an
inert thing cannot be a cause within the sense of Art. 1384 C.N.; for
a distinction is made between that which is inert and that which is
purely passive, the latter alone affording the defendant an answer
to the plaintiff’s claim. But the significant point is that the test of
“pure passivity” is sometimes said to turn upon the normality of the
thing, its position and behaviour. For it is easy enough to associate
normality with proper control by the gardien, abnormality marking
some deviation from the norms of the community –
and instantly
one is poised once again upon the brink of fault liability.
Despite these straws in the wind, the Jand’heur jurisprudence
remains substantially intact in France. Perhaps, as has been often
said, it maintains its position because of the continued lack of
adequate special laws to ensure the compensation of victims of
highway accidents. Carbonnier, for instance, suggests that it may
one day be said of the celebrated jurisprudence on art. 1384, al. 1,
C.N., that it was “an immense waste of intelligence and time” –
one good law on responsibility for car accidents, like the German
(3 May 1909) or Swiss (15 March 1932), would do the job at least
without having to embrace within the same formulae physical and
social realities of the most diverse kind. 2 However that may be the
Jand’heur jurisprudence does still dominate the scene, and affords
a truly striking illustration of the living process of interpretation
of a code in a manner responding to the felt need of the times. It
represents a remarkable adventure in judicial creativity, yielding
an interpretation of a provision of the Code Napoleon which, as
everyone agrees, the legislators of 1804 could never have imag-
inated.13
by F.M. at pp. 49-51; as well, see the note by Flow, D.C. 1941.1.85-90.
“1 See cass.-civ. January 21 and February 19, S. 1941.1.49., and the note
‘ 2 Carbonnier, Droit Civil, Vol. 2, no. 192.
13See Mazeaud et Tunc, La Responsabilitg Civile (5th ed.), vol. II, no. 1144.
Cf. the discussion before the legislature by Tarrible, Tribunat, in 15 Fenet,
Recueil complet des travaux prdparatoire du code civil (1827), p. 478: “In no
case is liability imposed if it be proved that the act causing the damage
could not have been prevented” –
the very provision which is, in substance,
embodied in the Quebec Code, art. 1054, para. 6.
No. 4]
AUTONOMOUS ACTS OF THINGS
Quebec Jurisprudence: the “autonomous acts” of things
It is plain that the Jand’heur jurisprudence depended upon the
firm rejection of the argument that a “thing” in terms of art. 1384,
al. 1, C.N., meant only a thing which was in some way self-activated,
and that this provision therefore excluded all cases (including the
paramount case of a motor car in motion) where the act of a thing
is in some way associated with human conduct. This “autonomous”
behaviour of things, if it can be understood at all, (for even the fact
that the thing is wherever it happens to be at the time of the accident
is normally attributable to human behaviour), must surely be con-
fined to a very limited number of cases indeed. In the ordinary way,
a thing, being inanimate, does not “act”. If it is in motion at the
time when it causes damage, it moves under human propulsion,
guidance or control. If it is inert when it causes damages, it is
nonetheless under human control or surveillance. “Self-activity” in
any meaningful sense would seem to be confined to internal chemical
or physical changes, like those giving rise to spontaneous com-
bustion or explosion,’14 and even then, only if such changes could
not be traced to the culpabale act or omission of a human being.
Thus even if one could make an intelligible category of self-activated
things, the effect would be to narrow the scope of a provision such
as art. 1384, al. 1, Code Napoleon, to the point where it could only
be applicable in very rare cases. Yet this is precisely what has been
done by Canadian courts in the interpretation of art. 1054 of the
Quebec Civil Code. Professor Cr6peau has remarked: ‘”
Quebec courts unlike the French courts, have always tried to maintain
a sharp distinction between, on the one hand, the ‘act’ (le fait) of a
thing, in which case the plaintiff has the benefit of the legal presumption
of liability attached to paragraph 1 of article 1054, and on the other hand,
the ‘act’ (le fait) of a person through the mere instrumentality of the thing,
in which case the plaintiff can only sue under article 1053 and must,
therefore bear the onus of proof.
It is ironic that, as Nadeau had observed, 6 art. 1384 of the Code
Napoleon, which speaks of le fait des choses, lends itself much more
readily to the Quebec theory of un fait autonome than does art.
1054 of the Quebec Civil Code, which drops the phrase fait des
choses (the act of things) and speaks, instead, simply of “dom-
mage…caus6 par les choses” (“damage caused by things”). Be that
14 Gas, electricity, steam, and ,,tous les 616ments d6ltres ou insalubres,>
are instanced by Bissonnette, J., in Federal Store Ltd. v. Tomy, [1954] B.R.
232, at p. 245.
15 “Liability for Damage Caused by Things”, (1962) 40 C.B.R. 222, at p. 234.
16 Traitg de droit civil du Quibec (1949), Vol. 8, p. 382.
McGILL LAW JOURNAL
[VoL. 17
as it may, the distinction is made and reiterated in a number of cases
by Anglin C. J. in the nineteen-twenties. Thus in Curley v. Latreille,”,
his Lordship said that responsability under art. 1054, para 1 “arises
only when the occurrence is due to the thing itself, not when
it is ascribale
to the conduct of the person by whom it is
put in motion, controlled or directed…” The learned Judge bases
himself upon the earlier French decisions, which were destined to
be submerged, in the decade that followed, by the Jand’heur juris-
prudence. This theme is repeated by Anglin, C. J., in two cases in
1928. In the first of these, Lacombe v. Power,18 having made the
distinction between an automobile starting” … of itself, i.e., without
the intervention of human agency, and owing to something inherent
in the machine… “, and one whose “.. . movement was due to an
act of the deceased… “, the Court holds that it is to the former
case alone that art. 1054 applies. This line of thinking appears again
in Pgrusse v. Sta ford,’9 which reaffirms that art. 1054, 1, has no
.where.., the real cause of the accident is the
application, “…
intervention of some human agency.. .”. Hence Taschereau J.,
declares in the Supreme Court that “la jurisprudence reconnue est
aujourd’hui [i.e. 1951] h l’effet que pour que cette derni~re partie
du premier paragraphe de 1054 C.C. s’applique, il faut ndcessaire-
ment que le dommage ait dt6 caus6 par le fait de la chose elle-mgme
sans aucune intervention humaine”.2 (Emphasis his own.)
This would mean that even a natural force which activates a
thing (e.g. the force of gravity) will not bring the matter within the
terms of art. 1054 (1), if some human agency has previously set
the stage by bringing about the situation which made that natural
force potentially operative. 21 On the other hand, a natural force
which is not thus “triggered” by preceding human conduct would
probably rank as force majeure or cas fortuit, so that the plaintiff
17 (1920), 55 D.L.R. 461 at 467. Cf. Montreal Tramways v. LaPointe, (1921), 31
B.R. 374, at 375, where Lamothe C.T said: ,
obligation a de moyens > expresses the fault principle. See, generally, Mazeaud
et Tunc, La responsabilitg Civile, (5th ed.), vol. I, 103-2, 66.
43 Supra, fni. 40.
No. 4]
AUTONOMOUS ACTS OF THINGS
to prevent the act which has caused the damage”. Having regard
for the plain words of the article, the qualification arbitrarily
lightens the obligation “de rdsultat”. In the light of this rider, it
is not entirely clear what is left of Lord Sumner’s statement in
Vandry’s case that there is a clear difference in law between a
rebuttable presumption of faute and a liability defeasible by proof
of inability to prevent the damage.44 At least, however, one may
say that the defendant’s rebuttal may not take the form of any
kind of evidence going to show absence of fault; 45 it may only
take the form of proof of inability by reasonable means to prevent
the damage complained of. But it would not be sufficient, seemingly,
for the defendant to show that he had used reasonable care in the
management of the thing or that he had no reasonable grounds
for anticipating that the thing would behave in the way that it
did, so long as evidence of that kind leaves open the question
whether he took reasonable steps to prevent the damage.46
Having regard to this exoneratory provision of the Quebec Code,
which has no parallel in the Code Napoleon, it is plain that even
had the Quebec courts chosen to follow the Jand’heur jurisprudence,
the result would have been quite different and far less dramatic
than in France. There, as we saw, no answer avails the defendant
other than proof of an external cause not imputable to the defend-
ant. But is this not, perhaps, the very reason why the Quebec courts
might have tackled para. 1 of art. 1054 less timorously than they
did, secure in the knowledge that the exoneratory clause continues,
in a modified way, to tie this provision to the fault principle? For
44 Thus Nadeau, supra, n. 16, at p. 406, says of the case of Watt and Scott
Ltd. that:
a… le Conseil priv6 ramenait le probl~me h ses v6ritables proportions et
tout en gardant dans les termes une pr6somption de responsabilit6 ordinaire,
il n’y a aucun doute qu’il assignait h la faute pr~sum~e du gardien sa place
vritable comme fondement de cette responsabilit6 lgale.,,
Thus he considers that it is not necessary for the defendant to prove the
direct cause and true nature of the accident, but enough to show positively a
complete absence of fault on his part (p. 441). Sed quaere: see the next foot-
note.
45Cf. Taschereau, J., in Cloaks Ltd. v. Cooperberg and Davis [1959], S.C.R.
785, at 788: [le gardien juridique] , peut s’exon~rer en d6montrant l’inter-
vention d’une force majeure, d’un cas fortuit, de l’acte d’un tiers, ou qu’il
n’a pu par des moyens raisonnables emp~cher le fait qui a caus6 le dommage.
Cette responsabilitd existe mdme en i’absence de faute attribuable au gardien
de la chose). (Italics added). Taschereau, I., dissented in the result, but
his statement of principle was expressly endorsed by Abbott and Judson,
JI., at p. 791.
46But see Nadeau, supra, fn. 44. It is submitted that this goes too far.
McGILL LAW JOURNAL
[Vol. 17
the exoneration permitted by art. 1054, para. 6, as understood by
the Privy Council, depends upon exculpation in the prescribed way.
In short, Art 1054, para. 1, read together with para. 6, establishes
a kind of rebuttable presumption of fault. It amounts to treating
all instances of damage caused by things in defendant’s care as
raising against the defendant a prima facie case, which he may
answer in any way which shows that he could not by reasonable
means prevent the damage. This surely does relatively small violence
to the fault principle so cherished in Quebec. Indeed it differs very
little from other legal presumptions of fault in which the judiciary
acquieces without misgiving! 7 Perhaps, after all, it was not really
necessary to insist so rigorously on the doctrine of the autonomous
act of things; a less destructive interpretation of art. 1054, para. 1,
was possible without total surrender to the risk principle.
“Garde de Structure” and “garde de comportement”
It is curious that the devotion of the courts to the fault doctrine
in Quebec, which has given rise to the restrictive interpretation of
art. 1054, para. 1, has not deterred the courts from flirting with
the risk principle in regard to the meaning of “garde” in the very
same paragraph. In this area, the courts have borrowed from
France a theory of the division of garde which, in France itself,
is still highly controversial. This theory maintains that garde can
be divisible, one person having the garde of the structure of the
thing whilst, concurrently, another person has the garde of it’s
comportement, or behaviour 8 The result is to make it possible
for an injured person to sue either the original owner or manu-
facturer of the thing, or the person in whose garde it was at the
time of the accident, according to whether the accident is attrib-
utable, causally speaking, to a structural defect in the thing, or to
it’s comportement.
This theory clearly leans towards the risk principle in fixing
the owner or manufacturer with a continuing responsibility for
damage caused by structural defects, even although he no longer
retains control of the thing in question. His responsibility here
turns, not upon garde of the thing (except in an extended sense),
but rather upon his ownership, present or past. This is a movement
towards “products liability”, similar in function, though different
47 And compare the reversal of the onus where there is a presumption
of fact – C6tg v. Duchesne [1963] B.R. 748.
48 See Mazeaud et Tunc, La Responsabilitd Civile (5th ed.), vol. II, no. 1160.3.
No. 4]
AUTONOMOUS ACTS OF THINGS
in character and legal analysis, from the developments in this field
which have taken place in the common law in the United States.4 9
This notion of divisible garde has been severely attacked as
being inconsistent with the text of the Civil Code, in conflict with
the principle of fault which animates the jurisprudence of Quebec,
and as adding nothing to the scope of the remedies which the
injured party enjoys without benefit of any such theory.50 One
may, however, take leave to question the last mentioned criticism,
since it is impossible to doubt that the transfer of the onus to
the defendant, in the sense of requiring him to exculpate himself
in the terms of art. 1054, para. 6., is a great and often decisive
advantage to the plaintiff.
However, without pursuing this particular controversy, it is
appropriate to note here the apparent lack of a coherent theory
which has resulted in a timid flight from the risk principle in deal-
ing with Art. 1054, para. 1, where it refers to “things”, whilst em-
bracing the risk principle by adopting this radical interpretation of
the word “garde” in the very same paragraph. It might be that,
at any rate in the field of physical injuries to the person, con-
temporary social attitudes are better served by the latter approach
than by the former, and in so far as the provisions of art. 1054
of the Civil Code are capable of yielding a meaning which will afford
greater protection to the injured party, a departure from the dogma
of fault liability would be no bad thing. It is indeed true that, to
some extent, the safeguarding of the interest of persons suffering
physical injuries is independently catered for by separate statutes,
such as those which obtain in the field of industrial injuries and
highway accidents. But there would seem to be no good reason why
the Civil Code, without offending accepted canons of construction,
should not come to the aid of special laws affecting particular fields
of personal injury, so as to provide a jus in subsidio which would
generally broaden and strengthen the remedial position of the
injured party.
-9 See the American Restatement of Torts (2nd ed.) 402 A, Kessler,
“Products Liability” (1967), 76 Yale L.J 887.
50See Kamil Antaki –
(1966), 12 McGill L.J., 41.
((Garde de structure et garde de comportement>,
