Briginshaw v Briginshaw
60 CLR 336(Judgment by: Latham CJ.)
Briginshaw v.
Briginshaw
Judges:
Latham CJ
Rich J
Starke J
Dixon J
McTiernan J
Judgment date: 18, 19 May, 30 June 1938
Judgment by:
Latham CJ.
This is an appeal from a judgment of Martin, J, whereby the appellant husband's petition for divorce on the ground of adultery was dismissed. The appeal is based upon the following grounds: (1) That the learned Judge wrongly decided that he could not hold that adultery was proved unless he was satisfied of the fact of adultery beyond reasonable doubt; that is, that it was wrongly held that the criminal standard of proof should be applied in divorce proceedings, at least in relation to a charge of adultery. (2) That the reasons for judgment given by the learned Judge showed that he was satisfied of the fact of adultery according to civil standards of proof; that is, upon a preponderance of probabilities, and that therefore the petition should have been granted. (3) Alternatively, that upon the evidence the learned Judge should have been so satisfied. (4) Alternatively, a new trial is sought.
The question of the standard of proof required in order to establish adultery in divorce proceedings has been expressly considered in three cases in the Supreme Court of New South Wales. The cases are: Godfrey v Godfrey , (1907) 24 WN NSW 57; Tuckerman v Tuckerman and Hogg , (1932) 32 SR NSW 220; and Doherty v Doherty , (1934) 34 SR NSW 290.
In the former two cases it was held that, in a suit for dissolution of marriage, a charge of adultery must be proved to the satisfaction of the judge or jury beyond reasonable doubt, and this principle was applied in the third case in relation to proceedings for variation of a maintenance order under the Deserted Wives and Children's Act 1901. It is argued for the respondent that a charge of adultery should be treated in the same way as a criminal charge, and that this proposition is established by the principles applied in the Ecclesiastical Courts in relation to such charges.
The Ecclesiastical Courts had no jurisdiction to pronounce a decree of divorce a vinculo , but questions of adultery arose in suits for divorce a mensa et thoro , and in other courts in proceedings involving legitimacy of issue. In Dillon v Dillon , (1842) 3 Curt. 86, which was a suit for divorce a mensa et thoro , Dr Lushington said that where a charge of adultery was made against a wife the proceeding was in effect a criminal proceeding, and that if there were any reasonable doubt she was entitled to the benefit of it. He dismissed the suit, because he was unable to say that the evidence was free from reasonable doubt. Dr Lushington, however, described the case at p 117 as "a case of great doubt." Therefore, in this case, the question did not really arise as to any difference between civil and criminal standards of proof, although the language used tends rather towards the adoption of the criminal standard. In more recent times, after the Matrimonial Causes Act 1857, there is but little authority on the subject, and what there is is not very satisfactory in character.
In Allen v Allen , [1894] P 248, the court of Appeal approved the words of Sir William Scott in Loveden v Loveden , (1810) 2 Hagg. Cons. R at p 2 -- "In every case almost the fact (of adultery) is inferred from circumstances which lead to it by fair inference as a necessary conclusion." The judgment of the court of Appeal proceeds --
To lay down any general rule, to attempt to define what circumstances would be sufficient and what insufficient upon which to infer the fact of adultery, is impossible. Each case must depend on its own particular circumstances. It would be impracticable to enumerate the infinite variety of circumstantial evidentiary facts, which of necessity are as various as the modifications and combinations of events in actual life. A jury in a case like the present ought to exercise their judgment with caution, applying their knowledge of the world and of human nature to all the circumstances relied on in proof of adultery, and then determine whether those circumstances are capable of any other reasonable solution than that of the guilt of the party sought to be implicated.
I am unable to regard either Loveden v Loveden or Allen v Allen as conclusive of the question which arises. In the first place, the phrase "circumstances which lead to it by fair inference as a necessary conclusion" is not very informative. The phrase combines in one sentence two quite different ideas. A "necessary conclusion" is one thing; a conclusion reached by what is generally described as "fair inference" is another thing. A "necessary conclusion" partakes of the character of a conclusion reached by mathematical demonstration. "Fair inference" is a phrase which is more properly descriptive of a process of thought leading to a conclusion which on the whole, is regarded as justifiable as a proper conclusion, but which cannot be said to be absolutely demonstrated. Further, the subsequent reference in Allen v Allen to "the infinite variety of circumstantial evidentiary facts" suggests reasonable inferences rather than "necessary conclusions," in such infinitely varying cases. The final advice that a jury should exercise its judgment "with caution, applying their knowledge of the world and of human nature to all the circumstances," is a statement which tends against the requirement that any conclusion should be a necessary conclusion in the ordinary logical sense. On the other hand, the question with which the quotation which I have made concludes, namely, whether the circumstances are capable of any other reasonable solution than that of guilt, is a statement which rather supports the applicability of the criminal standard of proof, which involves exclusion of any other reasonable hypothesis than that of guilt -- Wills' Circumstantial Evidence (5th ed ), 262. Thus I am unable to regard Allen v Allen as decisive of the questions raised.
In the case of Ross v Ross , [1930] AC 1, there was a difference of opinion in the House of Lords upon an appeal on facts on the subject of adultery. None of the learned Lords, however, suggested that the rule of proof beyond reasonable doubt was applicable in such a case. The matter was determined in exactly the same way as any appeal in a civil case upon a question of fact would have been determined. But the question of the proper standard of proof was not raised, and the case can hardly be regarded as a decision upon that point.
There is no mathematical scale according to which degrees of certainty of intellectual conviction can be computed or valued. But there are differences in degree of certainty which are real and which can be intelligibly stated, although it is impossible to draw precise lines, as upon a diagram, and to assign each case to a particular subdivision of certainty. No court should act upon mere suspicion, surmise or guesswork in any case. In a civil case, fair inference may justify a finding upon the basis of preponderance of probability. The standard of proof required by a cautious and responsible tribunal will naturally vary in accordance with the seriousness or importance of the issue. See Wills' Circumstantial Evidence (5th ed ), p 267 -- "Men will pronounce without hesitation that a person owes another a hundred pounds on evidence on which they certainly would not hang him, and yet all the rules of law applying to the one case apply to the other, and the processes are the same."
In criminal cases it has long been established that there must be a moral certainty of the guilt of the accused; the presumption of innocence must be definitely displaced, either by direct evidence of facts which constitute the offence charged, or by evidence from which the jury can draw an inference which satisfies the mind beyond reasonable doubt. The difference between the civil standard of proof and the criminal standard of proof has been examined and explained in this court in the case of Brown v The King , (1913) 17 CLR 570 , 20 ALR 570. See particularly at pp 584 et seq . and pp 595-6 (20 ALR 202, 206). Accordingly I am not prepared to adopt the view, which was suggested in argument, that the difference between the criminal and civil standards of proof is really only a matter of words.
What then is the rule to be applied to proof of adultery in proceedings for divorce? In the first place, I am of opinion that little attention should be paid to any decision of the Ecclesiastical Courts upon such a matter, and that they should not be accepted as binding. The jurisdiction in divorce, conferred in England by the Matrimonial Causes Act 1857, and in the various States of Australia by similar legislation, was a new jurisdiction. The Ecclesiastical Courts had never had power to pronounce a divorce a vinculo . Such a divorce could only be obtained by legislative procedure. The new legislation not only permitted divorce to be obtained by legal proceedings, but also gave persons a right to obtain a divorce if the conditions of the statute were satisfied. The legislation was strongly resented in many quarters. It was evidently feared by Parliament that the old rules of the Ecclesiastical Courts, belonging to an entirely different order of ideas, might be used so as to impede the exercise of the new jurisdiction, and to deprive the public of its benefits. Accordingly, s 22 of the Act of 1857, while providing that in other matters the new court established by s 6 should proceed and act and give relief on principles and rules as nearly as might be conformable to those in which the Ecclesiastical Courts had theretofore acted and given relief, expressly excepted from this provision "proceedings to dissolve any marriage." The section corresponding to s 22 of the English Act is to be found in the Victorian Marriage Act 1928, s 109, and also in the New South Wales Matrimonial Causes Act 1899, s 5. Therefore, primá facie , any special principle according to which the Ecclesiastical Courts acted in relation to proof of adultery in proceedings for divorce a mensa et thoro or other proceedings, is irrelevant and not applicable in proceedings for divorce a vinculo in the new jurisdiction.
Next, the House of Lords has stated in most explicit terms that the new jurisdiction is not a criminal jurisdiction, and that it is to be exercised according to the provisions of the applicable statute and not in accordance with any analogy derived from the administration of the criminal law. In Mordaunt v Moncreiffe , (1874) LR 2 H.L. (Sc.) 374, the House of Lords took the opinion of the Judges with respect to the question of the power of the court to grant a decree where the respondent was insane. It was held that, by the law of England, "adultery, though a grievous sin, is not a crime, and analogies and precedents of criminal law have no authority in the Divorce court, a civil tribunal" -- Headnote. Brett, J, regarded divorce proceedings as criminal in character; but Lord Chief Baron Kelly, Denman and Pollock, BB., and Keating, J, took the opposite view, being of opinion that divorce proceedings were civil in character. Lord Chelmsford, at p 384, said that it was unnecessary to consider whether proceedings for a divorce were of civil or quasi criminal nature, and that no aid to the consideration of the Act could be obtained from analogies applicable to cases of those different descriptions respectively. He said --
It is only necessary to bear in mind that the Act gives a right not previously existing to obtain the dissolution of a marriage for adultery, by the decree of a newly created court of Law, and from its proprovisions alone we must learn the conditions upon which the jurisdiction is to be exercised.
At p 393 Lord Hatherley said "the procedure in divorce is not a criminal procedure," and at pp 394-395, referring to the Divorce Acts, said, "every enactment indicates an analogy to civil and not criminal process."
Therefore, in order to determine the principles regulating the standard of proof in the Divorce court, it is necessary to go to the provisions of the statute, which in this case is the Marriage Act 1928. Section 80 of that Act is as follows: --
Upon any petition for dissolution of marriage, it shall be the duty of the court to satisfy itself, so far as it reasonably can, as to the facts alleged, and also to inquire into any counter-charge which may be made against the petitioner.
Section 86 is in the following terms: -- "Subject to the provisions of this Act the court if it is satisfied that the case of the petitioner is established shall pronounce a decree nisi for dissolution of marriage."
The phrase "it shall be the duty of the court to satisfy itself so far as it reasonably can" is also used in s 81 with reference to a petitioner being accessory to or conniving at or condoning adultery. In ss 82 and 83 the word "find" is used in relation to collusion and the other matters mentioned. In s 84 (1) it is provided that the court shall not be bound to pronounce a decree of dissolution of marriage if it "finds" that the petitioner has during the marriage been guilty of adultery. The same word is used in s 84 (2), but with reference to desertion.
The sections which are directly relevant to the present case are ss 80 and 86. Section 80 is a governing section applying to all the facts alleged as grounds for a petition for divorce -- adultery, desertion, etc. So far from the Legislature having used the phrase "satisfy itself beyond reasonable doubt," or any similar phrase, the Legislature has simply used the word "satisfy." It can be assumed that the Legislature was aware of the difference between the civil standard of proof and the criminal standard of proof. It would not be a reasonable interpretation of s 80 to hold that the words "satisfy itself" meant "satisfy itself beyond reasonable doubt." But the actual phrase is not merely "satisfy itself," but "satisfy itself so far as it reasonably can." The addition of the words "so far as it reasonably can" strongly supports the view that the Legislature did not intend the court to reach that degree of moral certainty which is required in the proof of a criminal charge. The words are apt and suitable for applying in the new jurisdiction the civil standard of proof, but they are not apt words of description for the criminal standard of proof.
In s 86 the words are simply -- "The court if it is satisfied that the case of the petitioner is established shall pronounce a decree nisi ." These words, like those in s 80, are applicable to all the grounds upon which a petition can be presented. If they require the criminal standard of proof in the case of adultery, they also require that standard of proof in the case of desertion -- a proposition which has no authority to support it. The result is that the ordinary standard of proof in civil matters must be applied to the proof of adultery in divorce proceedings, subject only to the rule of prudence that any tribunal should act with much care and caution before finding that a serious allegation such as that of adultery is established. This view is supported by the decision of this court in Dearman v Dearman , (1908) 7 CLR 549 , 15 ALR 287 -- an appeal on facts in a divorce suit where adultery was the ground of the petition. Barton, J, at p 557, stated the rule which he applied in the following words: -- "Before we infer adultery from circumstances we must have strong circumstances, such as would impel a reasonable mind to the conclusion that a petitioner had proved adultery" -- 15 ALR 289. Isaacs, J (pp 562-563), adopted language from Grant v Grant , (1839) 2 Curt. 16, as "an authoritative statement as to what is sufficient to establish the charge of adultery": --
The court must look at all the circumstances together, and form its own opinion whether they lead to a fair and natural conclusion that an act of adultery has taken place between the parties at some time or other
-- 15 ALR 292. Accordingly, I agree with the contention of the appellant that it is not the law that adultery in a divorce proceeding must be proved beyond reasonable doubt, that is, in my opinion, the criminal standard is not applicable in such a case.
It is next argued for the appellant that the learned Judge stated in his reasons for judgment that, according to the civil standard of proof, he was satisfied that adultery had been committed. In my opinion the words of the learned Judge will not bear this construction. The reasons for judgment show, in my opinion, that the learned Judge was left in a state of complete uncertainty on the issue of adultery. He was not prepared to accept or to act upon the evidence of any witness in the case. His Honour said -- "I have read the evidence several times, and the more I read it the more difficult the case seems." He then recited the evidence against the co-respondent. He said -- "In fact, all the witnesses gave their evidence well, and I could gather nothing adverse to them from their demeanour." Coming to the case against the respondent, he recited the relevant evidence, referred to discrepancies, and said -- "I am unable to draw any certain conclusions from the discrepancies." He added -- "Then there is a total denial by the wife on oath, and there was nothing in her demeanour in the box to suggest that she was lying." The nearest approach to a definite finding of fact is the statement of His Honour that the account of a particular conversation given by the corespondent was "the more feasible."
His Honour concluded his judgment by saying -- "I do not know what to believe. I have been very troubled." After a reference to a witness who was not called, the learned Judge said --
I have done my best to decide, but the petitioner must satisfy me that his story is true. I think I should say that if this were a civil case I might well consider that the probabilities were in favour of the petitioner, but I am certainly not satisfied beyond reasonable doubt that the evidence called by the petitioner should be accepted.
The appellant relied upon the statement -- "If this were a civil case I might well consider that the probabilities were in favour of the petitioner." But this statement in the whole of the context to which I have referred cannot be regarded as a finding that the witnesses for the petitioner or any of them were to be accepted as having spoken the truth. I am unable to discover in the reasons for judgment any finding of any fact. It therefore cannot, in my opinion, be said that the learned Judge has made findings upon which the petitioner is entitled to a decree.
It is then argued for the petitioner that, even if this be so, the learned Judge should have been satisfied by the evidence for the petitioner that adultery had been committed, and emphasis is placed upon His Honour's statement that he could gather nothing adverse to the witnesses from their demeanour. It is therefore urged that this court is in as good a position as the learned Judge to determine all questions of fact, and that it should accordingly do so. For myself, I should be most reluctant, save in a quite exceptional case, to find any person guilty of adultery upon conflicting evidence of conversations (as in this case) unless I saw the parties and other witnesses who gave evidence. If one regards only the evidence given, this is an ordinary case of a conflict of evidence, with probabilities and improbabilities on both sides. The learned Judge has been unable to make up his mind on the issue of adultery. The petitioner carries the onus of persuading a judge to make up his mind in his favour. If he does not succeed in so persuading a judge, he fails in his petition, and the matter is at an end.
There is, however, in my opinion, a special circumstance in this case which makes it proper that a new trial should be ordered. That special circumstance is to be found in the fact that the learned Judge (in my opinion wrongly) considered that he was bound to be satisfied of the fact of adultery beyond reasonable doubt, that is, according to the criminal standard of proof. He regarded the following statement at the end of his judgment as decisive of the case: -- "I am certainly not satisfied beyond reasonable doubt that the evidence called by the petitioner should be accepted."
Accordingly, the learned Judge did not actually consider the evidence according to the relevant and proper standard of proof. It is true that he says -- "I think I should say that if this were a civil case I might well consider that the probabilities were in favour of the petitioner." This statement is, however, discarded by the learned Judge as irrelevant, and there is no actual decision according to the probabilities of the case. There ought to have been such a decision with, as I have already stated, a realisation of the serious nature of the charge made against the wife. His Honour limits himself to saying, "I might well consider." He did not actually direct his mind to a consideration of the evidence upon a proper basis. The petitioner is entitled to have his case considered and decided upon such a basis.
I am therefore of opinion that there should be an order for a new trial.