R v P, NJ (No 4) - [2008] SASC 97

SUPREME COURT OF SOUTH AUSTRALIA
(Court of Criminal Appeal)
R v P, NJ (No 4)
[2008] SASC 97
Judgment of The Court of Criminal Appeal
(The Honourable Justice Duggan, The Honourable Justice Gray and The Honourable Justice White)
18 April 2008
CRIMINAL LAW - JURISDICTION, PRACTICE AND PROCEDURE - ADJOURNMENT, STAY OF PROCEEDINGS OR ORDER RESTRAINING PROCEEDINGS - STAY OF PROCEEDINGS - ABUSE OF PROCESS
CRIMINAL LAW - JURISDICTION, PRACTICE AND PROCEDURE - JUDGMENT AND PUNISHMENT - SENTENCE - FACTORS TO BE TAKEN INTO ACCOUNT - MISCELLANEOUS MATTERS - DOUBLE JEOPARDY
Appeal against a ruling antecedent trial – application for permanent stay of proceedings refused – appellant convicted of wounding with intent to cause grievous bodily harm and acquitted of attempted murder – appellant sentenced to 7 years’ imprisonment with 4 years’ non-parole period – victim since deceased – appellant now charged with murder – charge arose out of same act giving rise to appellant’s conviction of wounding with intent to cause grievous bodily harm and acquittal for attempted murder – application to stay prosecution on basis that amendments to Criminal Law (Sentencing) Act 1988 (SA) effected by Criminal Law (Sentencing) (Dangerous Offenders) Amendment Act 2007 (SA) exposed appellant to double punishment, such as to render prosecution for murder an abuse of process – legislative amendments introduced, inter alia, a mandatory minimum non-parole period of 20 years for murder – whether pursuant to 30(1) of the Sentencing Act, a sentence for murder can be backdated – whether, in the event of the appellant being convicted of murder, the backdating of the sentence would avoid double punishment – whether proceedings should be permanently stayed as an abuse of process.
Held: (Per Duggan, Gray and White JJ) Pursuant to s 30(1) of the Sentencing Act the trial judge would have power to backdate a sentence for murder to the commencement date of the sentence for wounding with intent to cause grievous bodily harm.
(Per Duggan and Gray JJ – dismissing the appeal) In the event of the appellant being convicted of murder, the backdating of the sentence would avoid any relevant unfairness – permanent stay of proceedings not justified in the circumstances.
(Per White J – allowing the appeal) In the event of the appellant being convicted of murder, the backdating of the sentence would not wholly avoid double punishment – imposition of double punishment oppressive to appellant – proceedings should be stayed permanently as an abuse of process.
Criminal Law Consolidation Act 1935 (SA) s 7A, s 11, s 21, s 270A, s 352(1)(c)(i); Criminal Law (Sentencing) Act 1988 (SA) s 30(1), s 30(2), s 30(3), 30(4), 30(5), 30(6), 31(1), s 32(5)(ab), s 32A(1), s 32A(3); Criminal Law (Sentencing) (Dangerous Offenders) Amendment Act 2007 (SA); Correctional Services Act 1982 (SA) s 21, s 67(1), s 67(2), s 67(3), s 67(6), s 67(7), s 70(1), s 70(2), s 73(1) and s 75(1); Prisons Act 1869 (SA) s 32; Prisons Act 1936 (SA) s 24; Acts Interpretation Act 1915 (SA) s 18 and s 22(1); Interpretation of Legislation Act 1984 (Vic) s 35(a); Road Traffic Act 1961 (SA) s 47B; Victims of Crimes Act 2001 (SA); Young Offenders Act 1993 (SA), referred to.
R v Garrett (1978) 18 SASR 308; R v Jamieson (1988) 50 SASR 130; R v Colson (1999) 73 SASR 407; R v Thomas (1986) 41 SASR 566; R v Newman & Simpson [2004] NSWCCA 102; R v P,NJ (No 3) [2008] SASC 63; Pearce v The Queen (1988) 194 CLR 610; Mill v The Queen (1988) 166 CLR 59; House v The King (1936) 55 CLR 499; R v Pearce (Unreported, Supreme Court of New South Wales, Court of Criminal Appeal, Sheller JA, Dowd and Hidden JJ, 18 December 1998); R v P,NJ [2006] SASC 309; R v P,NJ (No 2) (2007) 99 SASR 1; Governor of Brockhill Prison; Ex parte Evans [1997] QB 443; Arts v Briggs (1997) 93 A Crim R 56; Mills v Meeking (1990) 169 CLR 214; Burch v South Australia (1998) 71 SASR 12; R v Di Maria and Others (1996) 67 SASR 466; Owen v South Australia (1996) 66 SASR 251; Project Blue Sky v Australian Broadcasting Authority (1998) 194 CLR 355; Platz v Osborne (1943) 68 CLR 133; Police v Nowak (2000) 76 SASR 551; Bropho v Western Australia (1990) 171 CLR; Daniels Corp v ACCC (2002) 213 CLR 543; The Queen v Smith & Shoesmith (1983) 32 SASR 219; R v Hoar (1981) 148 CLR 32; R v Hammoud (2000) 118 A Crim R 66; R v Gorman (2002) 137 A Crim R 326; Johnson v The Queen (2004) 205 ALR 346; R v Sessions [1998] 2 VR 304; Jago v The District Court of New South Wales (1989) 168 CLR 23; Rogers v The Queen (1994) 181 CLR 251; Walton v Gardiner (1993) 177 CLR 378, considered.

R v P, NJ (No 4)
[2008] SASC 97
Court of Criminal Appeal: Duggan, Gray and White JJ

DUGGAN J. The background of this appeal is set out in the judgment of Gray J.
The application for a stay of the proceedings brought against the appellant was based on what is said to be an abuse of process. The appellant was found guilty of the offence of wounding with intent to cause grievous bodily harm. On 29 August 2003 he was sentenced for that offence to imprisonment for seven years. The court fixed a non-parole period of four years. The head sentence and the non-parole period commenced on 26 September 2002. The appellant is presently serving that sentence.
The victim died on 28 June 2004. The prosecution alleges that he died from the injuries inflicted during the incident upon which the offence of wounding with intent to cause grievous bodily harm was based. On 4 January 2006 the appellant was charged with the murder of the victim. The acts relied upon to establish the offence of murder are the same as those previously relied upon by the prosecution to establish the wounding charge.
In the event of a conviction for murder, a mandatory sentence of life imprisonment would have to be imposed. Furthermore, it is said by counsel for the appellant that the court would be required to fix a non-parole period of imprisonment for 20 years. It was submitted that the case did not come within any of the special reasons which would permit the court to exercise a discretion to reduce the non-parole period, at least in the absence of a guilty plea.
Furthermore, it was argued that the sentencing judge would have no power to backdate the head sentence and the non-parole period to the date on which the appellant’s present sentence commenced. According to the argument, this would result in double punishment.
Counsel for the Director of Public Prosecutions submitted that the court was empowered to backdate the head sentence and the non-parole period in the event that the appellant was convicted of murder. He argued that this would avoid any element of double punishment and that, accordingly, there was no basis for the court to grant a permanent stay of the proceedings against the appellant.
In order to determine whether the court has the power to backdate the sentence in these circumstances, it is helpful to have regard to some of the historical background to the court’s power to fix the commencement date for a sentence of imprisonment.
In the case of the early English criminal courts which sat continuously, sentences commenced on the date on which they were passed. However, the practice at assizes and quarter sessions varied. It would appear that a practice developed of directing that the sentence commence on the first day of the assizes or the quarter sessions. The practice of taking into account earlier periods of detention, when fixing a sentence in relation to the same offence did not arise until the criminal lists became longer and the waiting times in custody increased.
The English practice of backdating sentences to the first day of the Criminal Sessions was adopted in Australia. In South Australia s 32 of the Prison Act 1869 provided that sentences imposed at the Criminal Sittings of the Supreme Court were to commence on the first day of the sittings.
More comprehensive provisions were included in the Prisons Act 1936. Section 24 of that Act provided as follows:
24(1)All sentences of offenders, convicted at any criminal sittings of the Supreme Court or a circuit court shall date from the first day of holding the sittings unless the court orders that the sentence of imprisonment shall date from any other day, in which case the sentence shall date from that day.
(2)All other sentences of imprisonment shall date from the date of signing the warrant of commitment under which any offender is detained in custody, unless the offender was at large at the time of signing the warrant, in which case the sentence shall date from the time of the arrest of the offender.
(3)Whenever any court, judge, special magistrate, justice or other tribunal or person, in the exercise of any power whether statutory or otherwise, awards or orders, in passing any sentence of imprisonment on any person, that the sentence shall commence at the expiration of any imprisonment to which the person has been previously, or is at the same time, adjudged or sentenced, the sentence shall commence as so awarded or ordered.
In The Queen v Garrett[1] one of the questions which the trial judge reserved for the consideration of the Full Court was whether the Prisons Act 1936 empowered the court to ante-date a sentence of imprisonment beyond the first day of the Criminal Sittings at which a prisoner was tried and convicted.
[1] (1978) 18 SASR 308.
Hogarth ACJ and White AJ held that there was a power to ante-date such a sentence and that the power was to be inferred from the phrase “unless the court orders that the sentence of imprisonment should date from any other day” (s 24(1)).
The relevant provisions were later incorporated into the Correctional Services Act 1982. Section 21 provided as follows:
21(1)A sentence of imprisonment imposed upon a person convicted of an offence by the Supreme Court or a District Court shall be deemed to have commenced on the first day of the criminal sittings in which he was so convicted, unless the Court imposing the sentence directs that the sentence shall be deemed to have commenced, or shall commence, on a day that is earlier or later than that day.
(2)A sentence of imprisonment imposed upon a person, not being a sentence referred to in subsection (1) –
(a) shall, where the person is at large at the time that the warrant of commitment in respect of the sentence is signed, commence on the day upon which he is arrested for the purposes of committing him to a correctional institution pursuant to the warrant;
or
(b) shall, in any other case, commence on the day upon which the warrant of commitment in respect of the sentence is signed,
unless the court imposing the sentence directs that the sentence shall be deemed to have commenced, or shall commence, on a day that is earlier or later than the day referred to in paragraph (a) or (b).
(3)[Irrelevant].
Section 21 of the Correctional Services Act was amended in 1985 by the addition of s 21(3) which provided as follows:
(3)The court by which a person is sentenced to imprisonment shall, except where it suspends the sentence –
(a) specify the day on which or the time at which the sentence is to commence, or is deemed to have commenced;
and
(b) if a non-parole period has been fixed by the court, specify the day on which or the time at which that non-parole period is to commence, or is deemed to have commenced.
In R v Jamieson[2] the issue was whether it was appropriate to backdate the sentence imposed so as to include a period when the defendant was not in custody. It was held that, although the Correctional Services Act permitted that course, it would be rare that a non-custodial period would be treated as part of the sentence.
[2] (1988) 50 SASR 130.
Cox J held that the relevant power derived from s 21(1). King CJ said:[3]
Section 21 of the Correctional Services Act authorises the sentencing judge to direct that a sentence shall be deemed to have commenced or shall commence on a day that is earlier or later than the commencing date of the sittings at which the sentence is passed, and subs (3) of s 21 requires the judge to specify the date on which, or the time at which, the sentence is to commence or is deemed to have commenced.
In R v Garrett[4] it was held on the legislation as it then stood, and I think that there is no material distinction between that legislation and the present provision, that there was power to order the commencement of a sentence prior to the date of the commencement of the sittings at which the sentence was passed. Moreover Hogarth J and White J took the view that the sentence is broad enough to give a court a discretion to make an order ante-dating the sentence so as to cover a period during which the offender is not in custody.
It seems to me, however, that the circumstances in which it would be proper to make the latter order must be very rare. The purpose of the section, I think, is plainly to enable the court to make an order directing the commencement of the sentence on a date which will result in incorporation into the sentence periods spent in custody prior to the actual passing of the sentence.
[3] Ibid at 134.
[4] (1978) 18 SASR 308.
The provisions in the Correctional Services Act were replaced by s 30 of the Criminal Law (Sentencing) Act 1988 (“the Sentencing Act”). In so far as it is relevant to the present case s 30 provides as follows:
30    Commencement of sentences and non-parole periods
(1)     Where a court imposes a sentence of imprisonment and does not suspend the sentence, the court must specify the date on which, or the time at which, the sentence is to commence or is to be taken to have commenced.
(2)     If a defendant has spent time in custody in respect of an offence for which the defendant is subsequently sentenced to imprisonment, the court may, when sentencing the defendant, take into account the time already spent in custody and—
(a)     make an appropriate reduction in the term of the sentence; or
(b)     direct that the sentence will be taken to have commenced—
(i)on the day on which the defendant was taken into custody; or
(ii)on a date specified by the court that occurs after the day on which the defendant was taken into custody but before the day on which the defendant is sentenced.
(3)     Where a court imposes a sentence of imprisonment on a defendant who is not present in court, the court must direct that the sentence is to commence—
(a)     on the day on which the defendant is taken into custody pursuant to the warrant of commitment issued in respect of the sentence; or
(b)     if the defendant is subject to some other sentence of imprisonment—on the completion of that other sentence of imprisonment or at some earlier time fixed by the court.
(4)     Where a court fixes a non-parole period, the court must specify the date on which the non-parole period is to commence or is to be taken to have commenced.
(5)     Where a court directs that a sentence of imprisonment is to be taken to have commenced on the day on which the defendant was taken into custody, any non-parole period fixed by the court in respect of that sentence will be taken to have commenced on that day.
(6)     If, on imposing a sentence of imprisonment, the court fails to specify the date on which or the time at which the sentence is to commence or is to be taken to have commenced, the sentence—
(a)     will, in the case of a defendant not then in custody, commence on the day on which the defendant is subsequently taken into custody for the offence; or
(b)     will, in the case of a defendant already in custody for the offence, be taken to have commenced on the day on which the defendant was last so taken into custody; or
(c)     will, in the case of a defendant in custody for some other offence, commence on the day on which the sentence is imposed, unless the sentence is to be served cumulatively pursuant to this Act or any other Act.
Section 30 was considered by the Full Court in R v Colson[5]. The appellant in that case had been sentenced to a period of detention in a youth training centre. While there he escaped. At the time of the escape he was no longer a youth for the purposes of the Young Offenders Act 1993. A District Court judge sentenced him to a term of imprisonment for the escape to commence at the expiration of the period of detention.
[5] (1999) 73 SASR 407.
Pursuant to s 31(1) of the Sentencing Act a court may direct that a sentence be served cumulatively upon another sentence then being served. At the time Colson was decided that power did not extend to detention in a training centre. The issue in Colson was whether the order that the sentence commence at a future date could have been made under s 30.
The court held that s 30(1), which directs the court to specify the time at which a sentence is to commence, supported the order which the judge had made. The argument that s 31(1) provided a code for the circumstances in which a sentence could be ordered to commence in the future was rejected. It was held that s 30(1) provided a general power to backdate a sentence or to order that a sentence commence at a future date.
Doyle CJ (Prior and Mullighan JJ concurring) said:[6]
It is to be noted that an explicit reference is made to fixing a later commencement date, and that s 30(1) of the Sentencing Act also appears to contemplate the fixing of a commencement date that is later than the time at which the sentence is imposed.
In relation to s 30(1) of the Sentencing Act it is also pertinent to bear in mind that the Sentencing Act is intended to regulate generally the sentencing of offenders. It is reasonable to assume that it was intended to provide a sentencing court with flexible powers. It is also apparent that it was intended to impose upon a sentencing court an obligation to avoid uncertainty about the time of commencement of a sentence.
For present purposes the issue is whether s 30(1) supports the order that the judge made. Can his order be supported on the basis that it is an order specifying the time at which the sentence that he imposed is to commence?
On its face, and bearing in mind the contrast between s 24 of the Prisons Act and s 21 of the Correctional Services Act, s 30(1) appears to confer a power to specify that a sentence of imprisonment will commence at a time later than the date on which it is imposed.
The argument to the contrary is that it is pretty well unheard of for a court to impose a sentence of imprisonment and to specify that it is to commence at a later time, other than when the sentence is made cumulative upon another sentence in exercise of the power under s 31(1), or in the particular case dealt with by s 30(3). The argument is that it could not have been intended to confer such a power.
However, to my mind it is difficult to treat the subsequent subsections of s 30 as, in effect, exhaustively stating the scope of the apparently general power conferred by s 30(1). If that is what they do, then s 30(1) is to be read as doing no more than imposing an obligation to specify a date or time, the relevant date or time being arrived at as a result of the operation of later subsections, and not by the exercise of a power conferred by s 30(1) to fix a date or time. On the other hand, it might be said that if s 30(1) confers a general power, why did Parliament make the particular provisions found in the following subsections? The answer to that question is that Parliament thought it appropriate to deal specifically with the more commonly encountered situations, and, to some extent, to limit the scope of the discretion conferred by s 30(1) by providing what a court must do in certain circumstances. As well, s 30(2)(a) confers a power that is not conferred by s 30(1). On the other hand, s 30(2)(b) appears to be no more than a particular example of what may be done pursuant to the power conferred by s 30(1). Section 30(3) appears to limit the scope of the discretion conferred by s 30(1), by requiring the court to make a particular direction when it imposes a sentence of imprisonment on a person who is not present in court.
[6] Ibid at 410.

After referring to the cases of Garrett, Jamieson and R v Thomas[7] Doyle CJ continued:
The cases to which I have referred clearly enough treat s 30(1), or its predecessor, as conferring a general power to direct that a sentence is to commence at an earlier date or time than the time at which it is imposed. Thomas, at least in the context of an order akin to an order for accumulation, treats the predecessor provision as empowering the fixing of a commencement time after the sentence is imposed.
Taking into account the general terms in which s 30(1) is expressed, and its history, and not without some hesitation, I conclude that it is intended to confer a general power to specify a date or time for the commencement of a sentence, earlier or later than the time at which the sentence of imposed.
One of the reasons for my hesitation is the fact that, as I said earlier, it is pretty well unheard of for a court to order that a sentence of imprisonment is to commence at a later time than the time at which it is imposed, except in the circumstances dealt with by subsections (3) and (6). But the same comment applies to fixing an earlier date or time for commencement of a sentence, other than in the circumstances contemplated by subsections (2) and (6). As King CJ said in Jamieson, it would only be in very rare circumstances that it would be appropriate for a court to exercise the power to fix an earlier commencement date or time so as to include a period during which the offender has not actually been in custody: 50 SASR 130 at 134. Likewise, it would only be in exceptional circumstances that it would be appropriate for the court to fix a later commencement time, other than in the circumstances contemplated by s 30(3) and (6).
Nevertheless, for the reasons I have indicated, I conclude that s30(1) does authorise the judge to make the order that he made.
[7] (1986) 41 SASR 566 at 412.
Mr Wells QC, for the appellant, submitted that Colson was wrongly decided. He drew attention to the wording of s 30(1) and observed that it imposed a duty on the court to specify the starting date of a sentence. He distinguished a duty of this nature from the grant of a power and said that the distinction was reinforced by the word “specify” in s 30(1) and the reference to “directs” in other sub-sections of s 30. According to Mr Wells, the scheme of the section was for s 30(1) to oblige the court to specify a date for the purposes of ss 30(2), (3), (4) and (5). If the date was not specified, then s 30(6) provided a default situation.
Until the passing of the Sentencing Act the relevant sections of previous legislation had been interpreted as providing a general power to fix the commencement date of a sentence. The decisions explain that there are certain cases in which it would be appropriate to use the power, although they caution against inappropriate use.
It must be acknowledged that the words relied upon in Garrett (“unless the court orders”) and in Jamieson (“unless the court imposing the sentence directs”) do not have a counterpart in the Sentencing Act. However, it would seem unlikely that the legislature would intend to remove that power, thereby preventing its use in appropriate cases to ensure that justice is done.
Such an intention would also involve, at least in part, a retreat to the common law, the limiting effect of which has been almost obliterated by legislation and modern practice.
Although it is appropriate for the legislature to regulate the more common examples of pre-dating and forward-dating sentences, it has proved difficult for legislation to cater for all the circumstances in which such orders are required to do justice. It is important to bear in mind that the early rules in this area were developed from practice and that the practice of the courts remains an important feature. This is particularly so where backdating is called for in special cases in order to avoid injustice.
An example is to be found in the court’s approach to backdating in the case of alternative offences. It would seem that an alternative offence would not come within the scope of s 30(2) of the Sentencing Act which permits backdating of the sentence if the defendant has spent time in custody “in respect of an offence for which the defendant is subsequently sentenced to imprisonment”. So, if a defendant had been held in custody on a charge of murder and subsequently pleaded guilty to or was found guilty of manslaughter, it is doubtful that a sentence of imprisonment for manslaughter could be backdated pursuant to s 30(2). However, the backdating of sentences for manslaughter in these circumstances is common. It is true that the sentence for manslaughter could be reduced on account of time spent in custody, but that option is often unsatisfactory for the reasons explained in R v Newman & Simpson[8].
[8] [2004] NSWCCA 102 at [29]f.
I have referred to the fact that the power to backdate a sentence began as part of the practice of the court. Although it was based on a fiction, it effected a measure of justice for those prisoners who had been in custody since the commencement of the assizes or the quarter sessions.
When the legislature intervened in South Australia the statutory provisions were interpreted widely by the courts so as to ensure that justice was done in cases involving pre-sentence detention. The earliest legislation in this State re-affirmed the existing practice that sentences were to commence on the first day of the sittings. This was repeated in the Prisons Act 1936 with the qualification that it was subject to the order of the court. A similar provision was included in the Correctional Services Act 1982.
The Sentencing Act does not contain a provision relating to the commencement of the sittings of the court. Section 30 begins with a direction to the court that it is to specify the date on which the sentence is to commence or is deemed to have commenced. There follows a series of directions which the court may or must give in relation the commencement date of sentences in a variety of situations. There is nothing in the legislation which directly prohibits the court from following the previous practice of backdating sentences in circumstances not specifically provided for in the Sentencing Act. As I have said, it would be surprising if Parliament had decided to remove the power of the courts to do justice in those cases which had been identified by decisions of this court as appropriate situations for backdating.
This court should not overrule a previous decision of the Court of Criminal Appeal which is directly applicable unless it is of the view that it was arrived at per incuriam or was plainly wrong. In my view, this cannot be said of the decision of this court in Colson’s case, although I share the hesitation of Doyle CJ in arriving at a firm conclusion on the construction of s 30. It follows that, subject to the other arguments presented on behalf of the appellant, we should proceed on the basis that the trial judge would have the power to backdate the sentence for the crime of murder if the appellant were to be convicted of that offence. The power would enable the sentence for murder to be backdated to the commencement date of the sentence for wounding with intent to cause grievous bodily harm.
Mr Wells further submitted that the backdating of the non-parole period for the offence of murder would be contrary to recent amendments to the Sentencing Act. Section 32(5)(ab) of the Sentencing Act provides for a minimum non-parole period of 20 years for the offence of murder.
Section 32A(1) states that a minimum non-parole period prescribed by the Act represents the non-parole period for an offence at the lower end of the range of seriousness for offences to which the minimum period applies. The court cannot fix a non-parole period shorter than the prescribed period unless it is satisfied that special reasons exists for shortening the period. Section 32A(3) prescribes the only matters which might constitute special reasons. The court is to have regard to:
(a)[whether] the offence was committed in circumstances in which the victim’s conduct or condition substantially mitigated the offender’s conduct;
(b)if the offender pleaded guilty to the charge of the offence – that fact and the circumstances surrounding the plea;
(c)the degree to which the offender has co-operated in the investigation or prosecution of that or any other offence and the circumstances surrounding, and likely consequences of, any such co-operation.
Mr Wells submitted that the backdating of the non-parole period of 20 years would have the consequence that the minimum period of parole prescribed by s 32A would not be served. According to the argument, the backdating of the minimum non-parole period would shorten it in circumstances which were not authorised by the Act.
I am of the view that the judge at first instance was correct in her response to this submission. Her Honour said:[9]
In my view, the accused’s submission as to the interpretation and application of the legislation is flawed for a number of reasons. First, where an offence has a mandatory minimum non-parole period, s 32A(2)(b) gives the Court the power to fix a “shorter non-parole period”, but this power is limited by the “special exceptions” identified in s 32A(3). I note that the shortening of a non-parole period is different from backdating. The section has nothing to say about backdating, which if used to take into account time spent in custody, would not result in a shortening of the non-parole period. The prisoner at the end of the period would still have served a period of a minimum of 20 years non-parole. Backdating in this circumstance simply allows a person to have taken into account an actual period already served in custody as part of the overall sentence. The sentence would still indicate that the period of 20 years or more would be served, but was backdated to reflect time already served.
[9] R v P,NJ (No 3) [2008] SASC 63 at [24].
If a non-parole period of 20 years is to take account of pre-sentence custody, the period of pre-sentence custody is deemed to have been served as part of the 20 year period. If Mr Wells’ argument is correct, it would apply equally to the situation where a defendant who is on remand on a charge of murder is found guilty of that offence and the court backdates the non-parole period to the day on which that person was taken into custody. The pre-sentence period of custody is not served at the time as part of the non-parole period. However, the backdating of the non-parole period has that effect in law. The fact that, in the present case, the appellant was not in custody on the charge of murder throughout the entire period of pre-sentence custody does not detract in any way from the legal effect of the court’s order, namely, that the entire period of custody is to be considered as part of the non-parole period. The backdating of the sentence does not result in a reduction in the length of the sentence, but permits a period in custody to be counted as part of the sentence.
I would also reject the further argument put forward on behalf of the appellant that, if there is a power to backdate the sentence, the backdating cannot take effect from a date earlier than the commission of the alleged murder. This would be the date on which the victim died.
In my view, this argument is again one which ignores the nature of a deeming provision such as that in s 30. Reference has been made to the element of fiction involved in the backdating of a sentence. The practice previously referred to - of backdating sentences to the commencement of the Criminal Sittings - applied to persons who were not in custody as well as those who were in custody. Furthermore, it is obvious that a backdated sentence is not actually being served during the period of pre-sentence custody. The law deems it to have been served.
In exercising the power to backdate a sentence, care must be taken to ensure that there is a proper reason for doing so and that it is necessary in order to do justice in the case.
The circumstances of the present case are exceptional. The appellant is serving a sentence for an offence which is based on the same facts as those which form the basis of the prosecution for murder, though of course, with the added element of causing death. Furthermore, the same intention suffices for both offences.
The law regards the offence of murder as having been committed on the date on which the victim dies. However, the discretion which is recognised in Colson’s case would permit the court to backdate a sentence for murder so as to take account of the fact that the appellant had been serving a sentence based on the same circumstances.
Next, it was argued that, even if the sentences were made concurrent (as a result of the backdating of the sentence of life imprisonment and the non-parole period of 20 years), the sentence would offend against the principle discussed in Pearce v The Queen[10] in that it would result in a double penalty.
[10] (1998) 194 CLR 610.
In Pearce it was alleged that the appellant broke into the victim’s home and caused him grievous bodily harm. The appellant pleaded guilty to inflicting grievous bodily harm with intent to cause grievous bodily harm and to the further offence of breaking and entering a dwelling house and inflicting grievous bodily harm therein. Each of the offences was punishable by a maximum sentence of imprisonment for 25 years. It can be seen that a common element of the offences was the intention to do grievous bodily harm. The sentencing judge sentenced the appellant to imprisonment for 12 years on each count, to be served concurrently.
The High Court held that the principles involved in the concept of double jeopardy applied to the sentencing process. In their joint judgment McHugh, Hayne and Callinan JJ said[11]:
To the extent to which two offences of which an offender stands convicted contain common elements, it would be wrong to punish that offender twice for the commission of the elements that are common. No doubt that general principle must yield to any contrary legislative intention, but the punishment to be exacted should reflect what an offender has done; it should not be affected by the way in which the boundaries of particular offences are drawn. Often those boundaries will be drawn in a way that means that offences overlap. To punish an offender twice if conduct falls in that area of overlap would be to punish offenders according to the accidents of legislative history, rather than according to their just deserts.
[11] Ibid at [40].
Their Honours went on to conclude that the sentence on each count contained a portion calculated to punish the appellant for inflicting grievous bodily harm. Accordingly, it appeared that he was doubly punished for the one act.
After posing the question whether this made any difference bearing in mind that the sentences were ordered to be served concurrently, their Honours said:[12]
To an offender, the only relevant question may be "how long", and that may suggest that a sentencing judge or appellate court should have regard only to the total effective sentence that is to be or has been imposed on the offender. Such an approach is likely to mask error. A judge sentencing an offender for more than one offence must fix an appropriate sentence for each offence and then consider questions of cumulation or concurrence, as well, of course, as questions of totality[13].
Sentencing is not a process that leads to a single correct answer arrived at by some process admitting of mathematical precision[14]. It is, then, all the more important that proper principle be applied throughout the process.
Questions of cumulation and concurrence may well be affected by particular statutory rules[15]. If, in fixing the appropriate sentence for each offence, proper principle is not applied, orders made for cumulation or concurrence will be made on an imperfect foundation.
[12] Ibid at [45].
[13] Mill v The Queen (1988) 166 CLR 59.
[14] cf House v The King (1936) 55 CLR 499.
[15] See Crimes Act, s 444(2) and (3); Sentencing Act 1989 (NSW), s 9; see also Sentencing Act 1991 (Vict), s 16.
The appeal against sentence was allowed and the case was remitted to the New South Wales Court of Criminal Appeal which quashed the sentence on the charge of breaking and entering a dwelling house and inflicting grievous bodily harm therein and substituted a sentence of imprisonment for two years to be served concurrently with the other sentences.[16]
[16]R v Pearce (Unreported, Supreme Court of New South Wales, Court of Criminal Appeal, Sheller JA, Dowd and Hidden JJ, 18 December 1998).
The error of sentencing principle in Pearce’s case involved the failure to take into account the fact that there was a common element in each offence by imposing a sentence which did not punish twice for that element. Remedying this error required no more than the adjustment eventually made by the Court of Criminal Appeal.
However, in the present case there will be no error in sentencing principle if the appellant is found guilty of murder and sentenced to imprisonment for life with a non-parole period of 20 years backdated to the commencement of the sentence for wounding with intent to cause grievous bodily harm.
The important distinction between Pearce’s case and the present case arises from the fact that the head sentence for murder and the minimum non-parole period for that offence are prescribed by statute. The sentencing judge has no option but to impose the sentence of life imprisonment for murder. The maximum non-parole period cannot be shortened except in the circumstances outlined above. Even if the result did involve an element of double punishment, it would be the consequence of the application of the Sentencing Act.
In Pearce’s case there was a reason for the court to interfere with the sentence. The consequences of the sentence were unfair to the appellant and its structure disclosed an error of sentencing principle.
In my view, the same reasoning cannot be applied to the unique circumstances of the present case. It was held in previous proceedings in this case that it is not an abuse of the process of the court for the appellant to be presented on a charge of a murder despite the fact that he has been previously convicted and sentenced on the charge of wounding with intent to cause grievous bodily harm.
As the penalty of life imprisonment for murder is mandatory, it makes no practical difference whether it commences at the time of sentencing or is backdated to the time of the commencement of the sentence imposed on the appellant for the offence of wounding with intent to cause grievous bodily harm. In my view, any element of double punishment by reason of the imposition of the sentence of life imprisonment is more theoretical than practical and does not lead to the unfair consequences which were evident in Pearce’s case.
However, regard must be had to the non-parole period. Automatic release on parole is not available in the case of a sentence of life imprisonment. Section 67 of the Correctional Services Act 1982 provides that an application for release on parole by a person serving a sentence of life imprisonment cannot be made more than six months before the expiration of the prisoner’s non-parole period. After considering an application for release, the Parole Board may recommend to the Governor that the prisoner be released on parole (s 67(6)). The decision whether to release rests with the Governor.
In a sense, the non-parole period is part of the punishment. However, the primary sentence is life imprisonment. The decision to release on parole whilst remaining subject to the life sentence depends upon an assessment of the prisoner and the circumstances at the relevant time. Again, the backdating of this period does not result in unfair consequences such as those which were present in Pearce’s case. On the contrary, it would enable the appellant’s application for parole to be made at an earlier date.

In my view the appellant’s arguments do not provide a sufficient basis for the court to exercise the exceptional power of ordering a permanent stay of the hearing of the charge of murder.
I would dismiss the appeal.
GRAY J.
Introduction
This is an appeal against a ruling on an issue antecedent to trial. An application for a permanent stay had been refused. Permission to appeal pursuant to section 352(1)(c)(i) of the Criminal Law Consolidation Act 1935 (SA) was granted by the trial Judge.
On 24 September 2002, P, NJ, the defendant and appellant, on the prosecution case, stabbed H in the left temple with a knife. As a consequence, the appellant was arrested and taken into custody on 26 September 2002. The Information alleged attempted murder, and in the alternative, wounding with intent to cause grievous bodily harm. The appellant was tried before a judge without a jury. On 14 August 2003, he was found not guilty of attempted murder but guilty of wounding with intent to cause grievous bodily harm.[17]
[17] R v P, NJ [2003] SASC 308.
Following these verdicts H died on 28 June 2004, on the prosecution case, as a result of the stabbing. On 4 January 2006, an Information was laid against the appellant for the offence of murder, contrary to section 11 of the Criminal Law Consolidation Act 1935 (SA). The charge arose out of the same act giving rise to the appellant’s acquittal for attempted murder and the conviction of wounding with intent to cause grievous bodily harm. The particulars allege that on 24 September 2002 at Thevenard, the appellant murdered H.
On 5 September 2006, the appellant made application for an order that he not be prosecuted for murder. He advanced pleas of autrefois convict and autrefois acquit. In the alternative, the appellant sought an order that the proceedings be stayed as an abuse of process. On 29 September 2006, the application was heard by the trial Judge prior to trial. On 5 October 2006, the application was dismissed.[18] On 20 April 2007, this Court upheld the decision of the trial Judge.[19]
[18] R v P, NJ [2006] SASC 309.
[19] R v P, NJ (No 2) (2007) 99 SASR 1. Special leave to appeal to the High Court of Australia with respect to this decision was refused on 16 November 2007: P, NJ v The Queen [2007] HCATrans 691.
On 1 November 2007, the Criminal Law (Sentencing) Act 1988 (SA) was amended by the Criminal Law (Sentencing) (Dangerous Offenders) Amendment Act 2007 (SA). The 2007 Act relevantly amended section 30(2), section 32(5) and also inserted a new section 32A to the Sentencing Act. Section 11 of the 2007 Amendment Act provides that the 2007 amendments are retrospective.[20]
[20] Section 11 provides that: “An amendment made by Part 2 of this Act to the Criminal Law (Sentencing) Act 1988 applies whether the offence to which a sentence of imprisonment or non-parole period relates was committed before or after the commencement of that Part.”
Section 32 of the Sentencing Act, as amended, relevantly provides:
(1) Subject to this section, where a court, on convicting a person of an offence, sentences the person to imprisonment, the court must—
(a)if the person is not subject to an existing non-parole period—fix a non-parole period; or
…
(5) The above provisions are subject to the following qualifications:
…
(ab)if fixing a non-parole period in respect of a person sentenced to life imprisonment for an offence of murder, the mandatory minimum non-parole period prescribed in respect of the offence is 20 years;
…
The provisions of section 32A of the Sentencing Act should also be noted:
(1)     If a mandatory minimum non-parole period is prescribed in respect of an offence, the period prescribed represents the non-parole period for an offence at the lower end of the range of objective seriousness for offences to which the mandatory minimum non-parole period applies.
(2)     In fixing a non-parole period in respect of an offence for which a mandatory minimum non-parole period is prescribed, the court may –
…
(b)if satisfied that special reasons exist for fixing a non-parole period that is shorter than the prescribed period, fix such shorter non-parole period as it thinks fit.
(3)     In deciding whether special reasons exist for the purposes of subsection (2)(b), the court must have regard to the following matters and only those matters:
(a)the offence was committed in circumstances in which the victim's conduct or condition substantially mitigated the offender's conduct;
(b)if the offender pleaded guilty to the charge of the offence—that fact and the circumstances surrounding the plea;
(c)the degree to which the offender has co-operated in the investigation or prosecution of that or any other offence and the circumstances surrounding, and likely consequences of, any such co-operation.
…
Further subsections of the Sentencing Act are of relevance. Sections 30(1), (2) and (4)[21] provide:
[21] As earlier observed, section 30(2) was the only portion of section 30 amended by the Criminal Law (Sentencing) (Dangerous Offenders) Amendment Act 2007 (SA).
(1)     Where a court imposes a sentence of imprisonment and does not suspend the sentence, the court must specify the date on which, or the time at which, the sentence is to commence or is to be taken to have commenced.
(2)     If a defendant has spent time in custody in respect of an offence for which the defendant is subsequently sentenced to imprisonment, the court may, when sentencing the defendant, take into account the time already spent in custody and—
(a)    make an appropriate reduction in the term of the sentence; or
(b)    direct that the sentence will be taken to have commenced—
(i)     on the day on which the defendant was taken into custody; or
(ii)on a date specified by the court that occurs after the day on which the defendant was taken into custody but before the day on which the defendant is sentenced.
…
(4)     Where a court fixes a non-parole period, the court must specify the date on which the non-parole period is to commence or is to be taken to have commenced.
On 5 February 2008, the appellant applied for an order that the Information be permanently stayed on the ground that, as a result of the 2007 Amendment Act, the appellant was exposed to being punished twice for the same conduct.
The trial Judge, when addressing the second application for a stay, accepted the appellant’s contention that section 30(2) of the Sentencing Act would not allow the period already spent in custody by the appellant to be taken into account for any sentence that would be imposed if he were to be convicted of murder. This is because the appellant, if convicted, would not have spent time in custody for the offence of murder.
The Judge rejected the appellant’s contention that, in the event of the appellant being convicted of murder, section 30(1) of the Sentencing Act would not allow the sentence to be imposed to be backdated. The Judge construed section 30(1) in accordance with the earlier decision of this Court in Colson.[22] In her Honour’s view, in accordance with Colson, section 30(1) conferred a general power to specify a date or time for the commencement of a sentence, which may be earlier or later than the time at which the conviction was recorded.
[22] R v Colson (1999) 73 SASR 407.
The Judge considered that there was no reason why a backdated sentence for murder could not run concurrently with an earlier sentence that had been imposed for another offence. In the result, the Judge concluded that there would be power in the present case, if the appellant is convicted of murder, to backdate the sentence to be imposed to the date on which the appellant was first taken into custody. This would, in the Judge’s view, result in there being no double punishment. Having regard to the above reasons, the Judge dismissed the application.[23]
[23] R v P, NJ (No 3) [2008] SASC 63.
The Appeal
The appellant contended that the 2007 Amendment Act exposed the appellant, in the event of conviction for murder, to double punishment. It was said that this exposure rendered the prosecution of the Information an abuse of process and that any further prosecution should be stayed.
The appellant accepted that the Court in Colson[24] had concluded that section 30(1) of the Sentencing Act provided a general power to backdate a sentence. However, it was submitted that Colson was wrongly decided. It was further argued that the effect of the 2007 Amendment Act was that the sentencing court had no power to backdate the commencement of the mandatory non-parole period for murder.
[24] R v Colson (1999) 73 SASR 407.
The appellant submitted that even if section 30(1) had application, and there was power to backdate, in the circumstances of the present case it was inevitable that the defendant, if convicted of murder, would still be exposed to impermissible double punishment. It was contended that the power under section 30(1) would not permit backdating of the non-parole period to a commencement date earlier than the date on which the offence of murder was alleged to have been committed. That date was 28 June 2004, the date that H died. The appellant has been in custody from 26 September 2002. Finally, it was argued that even if Colson was correctly decided and there was power to backdate any sentence to be imposed for a conviction of murder to the date that the appellant was taken into custody, there would still be an exposure to double punishment sufficient to justify a stay.
The Director of Public Prosecutions submitted that the provisions of the 2007 Amendment Act should be interpreted as having application together with section 30(1) of the Sentencing Act, which was not amended by the 2007 Amendment Act. It was argued that section 30(1) continued to apply and that in the event of the appellant being convicted of murder, the Court was empowered to backdate a sentence. It was said that the decision in Colson remained good law. It was further submitted that there was power to backdate to the date on which the appellant was first taken into custody. The Director contended that in this circumstance there would be no double punishment.
The Common Law
In Pearce,[25] the High Court considered the common law principles with respect to double jeopardy. It was pointed out that the expression has been used in relation to a number of aspects of the criminal justice process, including punishment. McHugh, Hayne and Callinan JJ noted that the expression has been used to encompass a wider principle “that no-one should be punished again for the same matter”.[26] Their Honours then addressed double punishment in the context of two offences with overlapping elements and observed:[27]
To the extent to which two offences of which an offender stands convicted contain common elements, it would be wrong to punish that offender twice for the commission of the elements that are common. No doubt that general principle must yield to any contrary legislative intention, but the punishment to be exacted should reflect what an offender has done; it should not be affected by the way in which the boundaries of particular offences are drawn. Often those boundaries will be drawn in a way that means that offences overlap. To punish an offender twice if conduct falls in that area of overlap would be to punish offenders according to the accidents of legislative history, rather than according to their just deserts.
[25] Pearce v The Queen (1998) 194 CLR 610.
[26] Pearce v The Queen (1998) 194 CLR 610 at [9].
[27] Pearce v The Queen (1998) 194 CLR 610 at [40].
It is relevant to observe that at common law all periods spent in custody before sentence, other than custody wholly unrelated to the offence for which sentence is passed, count against the period of sentence to be served. An application of this principle is to be found in the judgment of Bingham LJ in Governor of Brockhill Prison; Ex parte Evans.[28] Recent Victorian authority has referred to this principle and to its implementation in circumstances where it is not possible pursuant to relevant legislation to make a reduction in the sentence otherwise to be imposed. In Arts v Briggs,[29] Callaway JA observed:
There are, of course, many cases where a person cannot be given credit for pre-sentence detention. He or she may be on remand for several months and then acquitted. The time spent on remand cannot be regarded as a bank balance on which to draw in relation to offences unconnected with the reason for custody, but that is not the case here. As Lord Bingham of Cornhill CJ said on behalf of the Divisional Court in Governor of Brockhill Prison; Ex parte Evans …:
“It has in our experience been the practice to assume that all periods of custody before sentence, other than custody wholly unrelated to the offences for which sentence is passed, will count against the period of the sentence to be served” (emphasis added).
Applying Heaney and Renzella, I would therefore deduct from the head sentence of six months to be imposed on Briggs a period that takes account of the time during which his detention was doubly warranted. As explained in the authorities, that is done as part of the exercise of the sentencing discretion and not pursuant to s 18 [of the Sentencing Act 1991 (Vic)].
Harper AJA commented:[30]
Justice requires that Mr Briggs, when resentenced, is given appropriate credit for the period during which he has been in custody since being sentenced on the present charge on 21 November last year. There may be no warrant in the Sentencing Act for reckoning that period as a period of imprisonment already served under the resentence, but in my opinion the court has such inherent jurisdiction as would enable it to make a declaration about post-sentence detention in the same way as s 18 requires where the offender was held in pre-sentence custody in relation to the relevant proceedings and for no other reason.
[28] Governor of Brockhill Prison; Ex parte Evans [1997] QB 443 at 462.
[29] Arts v Briggs (1997) 93 A Crim R 56 at 59.
[30] Arts v Briggs (1997) 93 A Crim R 56 at 68.
These authorities confirm the continued operation of the common law to protect offenders against double punishment when legislation does not provide the relevant protection. The courts have demonstrated considerable flexibility in the application of the common law principle. On occasions the problem has been addressed by the use of concurrent sentencing. On other occasions a reduction in one or other sentence may be made. The declaration power has also been utilised.
The Statutory Scheme – Principles of Interpretation
The provisions of the Sentencing Act deal with the imposing of terms of imprisonment, fixing of non-parole periods, the backdating of sentences and the taking into account of time spent in custody. The relevant provisions together form a scheme. The purpose of this scheme should be identified and given effect to. The relevant provisions should be viewed in their entirety including the 2007 amendments. There should be a consistent interpretation of the provisions.
Generally speaking the grammatical or literal interpretation of legislation should achieve the drafter’s desired purpose. Accordingly, words used in a section should be given their ordinary, everyday meaning, and the interpretation of legislation should be guided by common sense, experience and local knowledge. The particular provision under scrutiny should not be read in isolation from the rest of the enactment of which it forms part.
A purposive approach to statutory construction is statutorily prescribed by section 22(1) of the Acts Interpretation Act 1915 (SA):
[W]here a provision of an Act is reasonably open to more than one construction, a construction that would promote the purpose or object of the Act (whether or not that purpose or object is expressly stated in the Act) must be preferred to a construction that would not promote that purpose or object.
Dawson J, in Mills v Meeking,[31] when addressing section 35(a) of the Interpretation of Legislation Act 1984 (Vic), which is in similar terms to section 22(1) of the Acts Interpretation Act, observed:
[T]he literal rule of construction, whatever the qualifications with which it is expressed, must give way to a statutory injunction to prefer a construction which would promote the purpose of an Act to one which would not, especially where that purpose is set out in the Act. … The requirement that a court look to the purpose or object of the Act is thus more than an instruction to adopt the traditional mischief or purpose rule in preference to the literal rule of construction. The mischief or purpose rule required an ambiguity or inconsistency before a court could have regard to purpose. … The approach required by s 35 needs no ambiguity or inconsistency; it allows a court to consider the purposes of an Act in determining whether there is more than one possible construction. Reference to the purposes may reveal that the draftsman has inadvertently overlooked something which he would have dealt with had his attention been drawn to it and if it is possible as a matter of construction to repair the defect, then this must be done. However, if the literal meaning of a provision is to be modified by reference to the purposes of the Act, the modification must be precisely identifiable as that which is necessary to effectuate those purposes and it must be consistent with the wording otherwise adopted by the draftsman. Section 35 requires a court to construe an Act, not to rewrite it, in the light of its purposes.
[31] Mills v Meeking (1990) 169 CLR 214 at 235 (footnotes omitted).
In Burch v South Australia,[32] Cox J considered that section 22 of the Acts Interpretation Act (SA) should be similarly construed. Cox J considered that it was not necessary to first identify that the section under consideration was open to more than one construction before the purpose of the enactment could be referred to:[33]
It would be strange if the mischief rule could not be used where there is no apparent ambiguity but a literal interpretation would lead to inconsistency or injustice.
[32] Burch v South Australia (1998) 71 SASR 12.
[33] Burch v South Australia (1998) 71 SASR 12 at 18.
It should be noted that the interpretation of section 22 of the Acts Interpretation Act (SA) is still open to doubt. As Doyle CJ observed in Di Maria:[34]
[I]n more recent times the so-called purposive approach to interpretation has predominated. Under this approach, the purpose of the legislation, as understood by the court, may influence the process of construction. But it could do so only if the literal approach resulted in an ambiguity or inconsistency. These approaches, and the use of the purposive approach, are discussed in D C Pearce & R S Geddes, Statutory Interpretation in Australia (4th ed, 1996), Ch 2. The use of the purposive approach has, in my opinion, been assisted by rather than been dependent upon statutory provisions requiring courts to adopt a purposive approach. In my opinion, the purposive approach should be adopted by this Court and has, generally, been adopted.
That view is reinforced by s 22 of the Acts Interpretation Act 1915 (SA) …
However, it must be noted that in South Australia there must first be more than one construction “reasonably open”. Under such a provision it is open to argument whether the purpose of the legislation may be taken into account in deciding whether more than one construction is open: cf Mills v Meeking (1990) 169 CLR 214 at 234-235, per Dawson J. Having heard no submissions on this point, or indeed on the relevant principles of interpretation at all, I put that issue to one side, it not being appropriate to decide it now.
In discovering the purpose of a statutory amendment it is permissible for a Court to have regard to the content of the Minister’s Second Reading Speech in addition to reports of law reformers.[35]
[34] R v Di Maria And Others (1996) 67 SASR 466 at 472.
[35] Owen v South Australia (1996) 66 SASR 251; Burch v South Australia (1998) 71 SASR 12.

In Project Blue Sky v Australian Broadcasting Authority,[36] McHugh, Gummow, Kirby and Hayne JJ reviewed the principles of statutory interpretation and observed:
The primary object of statutory construction is to construe the relevant provision so that it is consistent with the language and purpose of all the provisions of the statute. The meaning of the provision must be determined “by reference to the language of the instrument viewed as a whole”. In Commissioner for Railways (NSW) v Agalianos, Dixon CJ pointed out that “the context, the general purpose and policy of a provision and its consistency and fairness are surer guides to its meaning than the logic with which it is constructed”. Thus, the process of construction must always begin by examining the context of the provision that is being construed.
A legislative instrument must be construed on the prima facie basis that its provisions are intended to give effect to harmonious goals. Where conflict appears to arise from the language of particular provisions, the conflict must be alleviated, so far as possible, by adjusting the meaning of the competing provisions to achieve that result which will best give effect to the purpose and language of those provisions while maintaining the unity of all the statutory provisions. Reconciling conflicting provisions will often require the court “to determine which is the leading provision and which the subordinate provision, and which must give way to the other”. Only by determining the hierarchy of the provisions will it be possible in many cases to give each provision the meaning which best gives effect to its purpose and language while maintaining the unity of the statutory scheme.
Furthermore, a court construing a statutory provision must strive to give meaning to every word of the provision. In The Commonwealth v Baume Griffith CJ cited R v Berchet to support the proposition that it was “a known rule in the interpretation of Statutes that such a sense is to be made upon the whole as that no clause, sentence, or word shall prove superfluous, void, or insignificant, if by any other construction they may all be made useful and pertinent”.
[36] Project Blue Sky v Australian Broadcasting Authority (1998) 194 CLR 355 at [69]–[71]. (footnotes omitted).
On occasions later amendments are made to legislation without affecting parts of the amended legislation that have been the subject of judicial interpretation. The view has been taken that in this circumstance there is an indication that the legislature was satisfied that the judicial interpretation was correct and should continue to apply.[37] In Police v Nowak,[38] the appellant challenged the settled interpretation of section 47B of the Road Traffic Act 1961 (SA). In declining to overrule previous decisions dealing with this section, Doyle CJ observed:[39]
S18 of the Acts Interpretation Act 1915 does away with any presumption that the enactment or re-enactment of a provision that has been construed in a particular manner creates a presumption that Parliament has approved that construction. But the fact that Parliament has amended many of the provisions relating to drink driving offences, but has not sought to amend s47B(4), at least suggests that the settled interpretation has not been seen as giving rise to unsatisfactory results.
The same can be said in the present case. Despite the operation of section 18 of the Acts Interpretation Act, the fact that section 30(1) was not amended by the 2007 Amendment Act indicates that Parliament has not viewed the previous interpretations of this section as leading to unsatisfactory results.
[37] Platz v Osborne (1943) 68 CLR 133; Police v Nowak (2000) 76 SASR 551.
[38] Police v Nowak (2000) 76 SASR 551.
[39] Police v Nowak (2000) 76 SASR 551 at [16].
In Bropho v Western Australia,[40] the High Court referred to rules relating to the construction of statutory provisions that would abolish or modify fundamental common law principles or rights. The Court noted:[41]
[T]he rationale of the presumption against the modification or abolition of fundamental rights or principles is to be found in the assumption that it is “in the last degree improbable that the legislature would overthrow fundamental principles, infringe rights, or depart from the general system of law, without expressing its intention with irresistible clearness; and to give any such effect to general words, simply because they have that meaning in their widest, or usual, or natural sense, would be to give them a meaning in which they were not really used” [Potter v. Minahan (1908) 7 CLR 277 at 304, and see, also, Ex parte Walsh and Johnson; In re Yates (1925) 37 CLR 36 at 93.]
More recently in Daniels Corp v ACCC,[42] McHugh J observed:
Courts do not construe legislation as abolishing, suspending or adversely affecting rights, freedoms and immunities that the courts have recognised as fundamental unless the legislation does so in unambiguous terms. In construing legislation, the courts begin with the presumption that the legislature does not interfere with these fundamental rights, freedoms and immunities unless it makes its intention to do so unmistakably clear. The courts will hold that the presumption has not been overcome unless the relevant legislation expressly abolishes, suspends or adversely affects the right, freedom or immunity or does so by necessary implication. They will hold that the legislature has done so by necessary implication whenever the legislative provision would be rendered inoperative or its object largely frustrated in its practical application, if the right, freedom or immunity were to prevail over the legislation. A power conferred in general terms, however, is unlikely to contain the necessary implication because “general words will almost always be able to be given some operation, even if that operation is limited in scope”.
These principles of statutory interpretation guide the construction of the relevant provisions.
[40] Bropho v Western Australia (1990) 171 CLR 1.
[41] Bropho v Western Australia (1990) 171 CLR 1 at 18.
[42] Daniels Corp v ACCC (2002) 213 CLR 543 at [43] (footnotes omitted).
The Statutory Scheme – the Sentencing Act
Part 3 of the Sentencing Act addresses “imprisonment”. The Part has two divisions – the first addresses sentences of imprisonment, and the second non-parole periods. More particularly, section 30 addresses the commencement of sentences and non-parole periods; section 31 deals with cumulative sentences; and sections 32 and 32A, the final sections in Part 3, address the duty of the Court to fix or extend non-parole periods. As earlier observed, the 2007 Amendment Act was enacted with retrospective effect.[43]
[43] See section 11 of the Criminal Law (Sentencing) (Dangerous Offenders) Amendment Act 2007 (SA).
The overriding object of Part 3 of the Sentencing Act is to provide a sentencing court with wide powers to impose sentences of imprisonment and fix non-parole periods that are otherwise determined to be appropriate for the offending. The legislation acknowledges that occasions will arise for the need for reductions to be made on account of time spent in custody, and allows backdating of a sentence to occur. In providing the requisite powers, the legislation respects the need to avoid double punishment.
The Sentencing Act, and in particular Part 3, does not seek to erode the common law principle providing protection against double punishment. Had the legislature intended to erode this principle, in accordance with the earlier referred to authorities, a clear intention to do so would have appeared in the legislation. There is no such intention evident. To the contrary, the legislation discloses an intention to respect and support the common law. In Colson,[44] Doyle CJ reviewed the progenitor provisions to section 30(1). Those provisions have been the subject of consistent interpretation. In making the 2007 amendments, the legislature has seen fit not to disturb this interpretation. The purpose of the provisions of the Sentencing Act is to allow the Court to deliver just and appropriate punishments. It is against this background that the amending legislation is to be understood.
[44] R v Colson (1999) 73 SASR 407.
The legislature, in enacting the 2007 Amendment Act, has determined that subject to limited exceptions, a mandatory non-parole period should be fixed when sentencing for murder. The legislature intends that those convicted of murder serve a minimum of 20 years in custody. However, there is nothing in the 2007 Amendment Act that would suggest, in appropriate circumstances, that backdating of a sentence including a non-parole period should not take place, providing the minimum term is served in custody.
Colson[45] addressed the power of the Court to make an order to set the commencement of a sentence at a later time than the date at which the sentence was imposed. The Court was concerned with the interpretation of section 30(1) of the Sentencing Act. As earlier observed, although section 30(2) has been amended by the 2007 Amendment Act, section 30(1) was not amended.
[45] R v Colson (1999) 73 SASR 407.
Doyle CJ traced the history and purpose of section 30(1) and summarised the history of the progenitor provisions:[46]
[46] R v Colson (1999) 73 SASR 407 at [16], [17], [22], [23].
Section 30(1) was enacted to overcome a common law rule that a sentence of imprisonment was treated as commencing on the first day of the sitting of a Court sitting under a Commission of Oyer and Terminer or Gaol Delivery, and to overcome a further common law rule that there was no power to make an order that a sentence commence at an earlier time: R v Garrett (1978) 18 SASR 308 at 314. The common law rule was displaced in 1934, when the Prisons Act 1936 (SA) was amended so that s 24(1) read as follows:
“All sentences of offenders, convicted at any criminal sittings of the Supreme Court or a Circuit Court shall date from the first day of holding the sittings unless the Court orders that the sentence of imprisonment shall date from any other day, in which case the sentence shall date from that day.”
That provision was carried forward into s 21(1) of the Correctional Services Act which provided as follows:
“A sentence of imprisonment imposed upon a person convicted of an offence by the Supreme Court or a District Court shall be deemed to have commenced on the first day of the criminal sittings in which he was so convicted, unless the Court imposing the sentence directs that the sentence shall be deemed to have commenced, or shall commence, on a day that is earlier or later than that day.”
It is to be noted that an explicit reference is made to fixing a later commencement date, and that s 30(1) of the Sentencing Act also appears to contemplate the fixing of a commencement date that is later than the time at which the sentence is imposed.
In relation to s 30(1) of the Sentencing Act it is also pertinent to bear in mind that the Sentencing Act is intended to regulate generally the sentencing of offenders. It is reasonable to assume that it was intended to provide a sentencing court with flexible powers. It is also apparent that it was intended to impose upon a sentencing court an obligation to avoid uncertainty about the time of commencement of a sentence.
...
The predecessor provisions of s 30(1) have been considered by this Court. However, the earlier provisions did not contain the detailed subsequent provisions to be found in s 30. In R v Garrett the Court emphasised that s 24 of the Prisons Act conferred a power to order that the sentence shall date from any other day, but in context it may be that the Court was directing its mind only to a date before the first day of the sittings at which the sentence was imposed. In R v Thomas (1986) 41 SASR 566 a judge sentenced an offender to imprisonment for an offence against Commonwealth law and for an offence against State law. Commonwealth law then required the Court to impose a minimum sentence in respect of the offence against Commonwealth law. State law then required the Court to impose a single non-parole period in respect of all sentences imposed under State law. The question arose as to whether the Court had power, in effect, to make the sentence under State law, and the non-parole period, cumulative upon the sentence imposed under Commonwealth law. In particular, could the judge order that the sentence under State law, and the non-parole period under State law, would commence at the expiration of the service of the minimum term of imprisonment fixed for the Commonwealth offence? This Court held that s 21(1) of the Correctional Services Act was sufficiently flexible to allow for an order in those terms. It is to be noted that this meant that the Court was directing that the sentence would commence at a time later than the time at which it was imposed. Granted, this was in the context of, in effect, making one sentence of imprisonment cumulative upon another, although that is not exactly what the Court did. It is to be noted that it was not suggested that the relevant result was to be arrived at in exercise of the power then conferred by s 310 of the Criminal Law Consolidation Act to make a sentence of imprisonment cumulative upon another sentence of imprisonment. In R v Jamieson (1988) 50 SASR 130, without making any reference to the decision in Thomas, the Court referred to s 21 of the Correctional Services Act as conferring a power, quite generally, to order the commencement of a sentence prior to the date of the commencement of the sittings at which the sentence was imposed. The Court did not have to consider the use of that power to specify a commencement time later than the time at which the sentence was imposed.
The cases to which I have referred clearly enough treat s 30(1), or its predecessor, as conferring a general power to direct that a sentence is to commence at an earlier date or time than the time at which it is imposed. Thomas, at least in the context of an order akin to an order for accumulation, treats the predecessor provision as empowering the fixing of a commencement time after the sentence is imposed.
Doyle CJ addressed the construction of section 30(1):[47]
On its face, and bearing in mind the contrast between s 24 of the Prisons Act and s 21 of the Correctional Services Act, s 30(1) appears to confer a power to specify that a sentence of imprisonment will commence at a time later than the date on which it is imposed.
The argument to the contrary is that it is pretty well unheard of for a court to impose a sentence of imprisonment and to specify that it is to commence at a later time, other than when the sentence is made cumulative upon another sentence in exercise of the power under s 31(1), or in the particular case dealt with by s 30(3). The argument is that it could not have been intended to confer such a power.
However, to my mind it is difficult to treat the subsequent subsections of s 30 as, in effect, exhaustively stating the scope of the apparently general power conferred by s 30(1). If that is what they do, then s 30(1) is to be read as doing no more than imposing an obligation to specify a date or time, the relevant date or time being arrived at as a result of the operation of later subsections, and not by the exercise of a power conferred by s 30(1) to fix a date or time. On the other hand, it might be said that if s 30(1) confers a general power, why did Parliament make the particular provisions found in the following subsections? The answer to that question is that Parliament thought it appropriate to deal specifically with the more commonly encountered situations, and, to some extent, to limit the scope of the discretion conferred by s 30(1) by providing what a court must do in certain circumstances. As well, s 30(2)(a) confers a power that is not conferred by s 30(1). On the other hand, s 30(2)(b) appears to be no more than a particular example of what may be done pursuant to the power conferred by s 30(1). Section 30(3) appears to limit the scope of the discretion conferred by s 30(1), by requiring the Court to make a particular direction when it imposes a sentence of imprisonment on a person who is not present in court.
...
Taking into account the general terms in which s 30(1) is expressed, and its history, and not without some hesitation, I conclude that it is intended to confer a general power to specify a date or time for the commencement of a sentence, earlier or later than the time at which the sentence was imposed.
One of the reasons for my hesitation is the fact that, as I said earlier, it is pretty well unheard of for a court to order that a sentence of imprisonment is to commence at a later time than the time at which it is imposed, except in the circumstances dealt with by subss (3) and (6). But the same comment applies to fixing an earlier date or time for commencement of a sentence, other than in the circumstances contemplated by subss (2) and (6). As King CJ said in Jamiesons, it would only be in very rare circumstances that it would be appropriate for a court to exercise the power to fix an earlier commencement date or time so as to include a period during which the offender has not actually been in custody: 50 SASR 130 at 134. Likewise, it would only be in exceptional circumstances that it would be appropriate for the Court to fix a later commencement time, other than in the circumstances contemplated by s 30(3) and (6).
Nevertheless, for the reasons I have indicated, I conclude that s 30(1) does authorise the judge to make the order that he made.
It is this reasoning which was said to be incorrect.
[47] R v Colson (1999) 73 SASR 407 at [19]-[21], [24]-[26].
As earlier observed, the appellant submitted that there was no power to backdate the mandatory non-parole period. It was contended that any such backdating would in substance effect a reduction of the non-parole period below the statutory mandatory minimum, and as such was forbidden by the 2007 Amendment Act. This submission necessarily challenged the correctness of this Court’s decision in Colson.
The appellant submitted that section 30(1) was not an empowering provision, but was rather a provision which only required the Court to specify a date for the commencement of a sentence of imprisonment. It was contended that the empowering provisions were to be found in subsequent subsections of section 30. It was then suggested that the only power relevant to the present case was contained in subsections 30(2), (4) and (5).
The Director submitted that the decision in Colson was correct. Section 30(1) conferred a general power on a sentencing court to specify a date and time for the commencement of a sentence. The Court had power to post-date or ante-date the time of commencement. It was said that this construction was supported by authorities dealing with progenitor provisions – Jamieson,[48] Smith & Shoesmith[49] and Garrett.[50] It was pointed out that section 30(2) did not allow a sentence to commence at a date later than the date of the imposition of the sentence. In certain exceptional circumstances it may be appropriate in the interests of justice for a sentence to pre-date the date on which a person was taken into custody. It was also pointed out that section 30(1) would allow the Court to have regard to time spent in custody with respect to alleged offending, not the subject of the particular sentence. This might arise when a defendant was ultimately charged with a different offence from that for which he had been initially remanded and the earlier charges abandoned. It was pointed out that in these circumstances, section 30(1) allowed for a flexible approach to be taken to ensure that the appropriate sentence was served. These submissions should be accepted.
[48] R v Jamieson (1988) 50 SASR 130 at 134.
[49] The Queen v Smith & Shoesmith (1983) 32 SASR 219.
[50] R v Garrett (1978) 18 SASR 308.

In my view Colson was correctly decided. There is no basis to doubt its continued applicability having regard to the 2007 Amendment Act. The interpretation does not strain the language of section 30 of the Sentencing Act. It provides a commonsense interpretation that reflects and supports the avoidance of double punishment. It also accords with the fundamental precept of sentencing law – that punishment is to be appropriate to the criminal conduct of the offender. There is no suggestion in the 2007 Amendment Act that a sentencing court should not be able to backdate a mandatory non-parole period so as to avoid a defendant being doubly punished. The legislative scheme is not expressed to be a code, and accordingly the common law powers to act to avoid double punishment also remain.
As pointed out in Pearce,[51] where crimes are committed with overlapping elements, double punishment may be able to be avoided by the ordering of concurrency of sentences. The same consequence would follow from the coincidence of two sentences in the circumstance where the sentences are not imposed at the same time but the later sentence is backdated so as to coincide with the commencement of the earlier sentence.
[51] Pearce v The Queen (1998) 194 CLR 610.
The Avoidance of Double Punishment
Counsel for the appellant accepted that if the Court had power to backdate a sentence imposed for the offence of murder, and also to backdate the mandatory non-parole period, in the ordinary case, the possibility of double punishment may be avoided. However, it was said that in the present case, in the event of there being a conviction for murder, and if the power to backdate existed, the earliest possible date for backdating would be to the date of death of H. Having regard to the delay between the actus reus and the death of H, it was contended that substantial double punishment would follow.
The Court, in the event that the appellant is convicted of the offence of murder, has power to backdate the sentence to be imposed including the mandatory non-parole period. That power is provided by section 30(1) of the Sentencing Act.
The power to backdate in section 30(1) is not limited by reference to the date on which the elements of the offence occurred. The act, the subject of the charge of murder, was committed on 24 September 2002. On 26 September 2002 the appellant was taken into custody. There is a clear nexus between the offence of murder and 24 September 2002. This nexus would provide a principled basis on which the section 30(1) power may be invoked.
Even with a sentence for the offence of murder backdated to the time when the appellant was taken into custody, there still remains the possibility of some perceived double punishment. Although the two sentences would then be served at the same time, the appellant would have been twice convicted of crimes with overlapping elements. The appellant would have been the subject of separate sentences being served at the same time. However, this is a necessary consequence of the mandated life imprisonment sentence and the non-parole period of 20 years for the crime of murder.
It is to be observed that the appellant was properly convicted of the offence of wounding with intent to cause grievous bodily harm. There has been no appeal against that conviction. The appellant was properly sentenced with respect to that offence. There has been no appeal against the sentence imposed. The appellant now stands properly charged with the offence of murder. This Court has ruled that there is no abuse of process in his being so charged.
In the event that the appellant is convicted of the offence of murder, he would be sentenced to mandatory life imprisonment and a mandatory non-parole period of 20 years would be fixed. If convicted of murder, this is the penalty that the law would require to be imposed. It is open, as earlier observed, for the appellant’s sentence to be backdated to commence from the time when he was first taken into custody. His sentence and non-parole period would commence from that time. In this event, the backdated sentence for murder would coincide with, and be served at the same time as, the earlier sentence. The appellant would only serve the one period of imprisonment and, in that respect, would not be doubly punished. In a very real sense, the earlier sentence would be absorbed into, and form part of, the mandatory sentence to be imposed in the event that the appellant is convicted of, and sentenced for, the offence of murder. It is true that the appellant would then be convicted of two offences in circumstances where there were overlapping elements. However, this is a consequence of the abolition of the year and a day rule, and of the bringing of the lesser charge prior to the death of H.
If the appellant’s contentions were to be accepted, it would follow that in other similar circumstances there would be a stay of the subsequent murder charge. The suggested double punishment would arise. In my view, such a result would reflect adversely on the administration of justice. Victims’ families, and the general community, would feel justifiably aggrieved. The event of a delayed death would result in the offender escaping the legal consequences of his or her actions. Justice would not be seen to be done.
As pointed out in Pearce,[52] the sentencing process is a practical process undertaken in accordance with settled principle. There needs to be adequate transparency. A defendant and the public must be able to see that a just and appropriate punishment has been imposed with respect to the relevant criminal conduct. In my view, in the event of the appellant being convicted of the crime of murder, the backdating of the sentence to be imposed to the time when the appellant was first taken into custody, would, from a practical point of view, avoid any relevant double punishment. In this circumstance, there would be transparency of the sentencing process. The appellant and the public would observe that there had been accorded a just and appropriate punishment.
[52] Pearce v The Queen (1998) 194 CLR 610 at [39], [121].
Insofar as it may be suggested that there remained some perception of double punishment, that circumstance would not be sufficient to justify a stay of proceedings.
Conclusion
This appeal should be dismissed.
WHITE J: The appellant is charged with the murder of DH. On the prosecution case, DH died on 28 June 2004 from the effects of a stab wound inflicted on 24 September 2002 by the appellant.
The appellant was previously charged with two offences arising from his stabbing of DH: attempted murder,[53] and wounding with the intention to inflict grievous bodily harm.[54] On 14 August 2003 he was acquitted by a judge sitting without a jury of the first of those offences but convicted of the latter. The judge acquitted the appellant of the offence of attempted murder because he was not satisfied beyond reasonable doubt that the appellant had, at the time of the stabbing, an actual intention to kill DH. He was, however, satisfied beyond reasonable doubt that the appellant had stabbed DH causing him a very serious wound and that his action in stabbing had been a conscious, voluntary and deliberate act carried out with the intention of inflicting grievous bodily harm.
[53] See Criminal Law Consolidation Act 1935 (SA) (CLCA) ss 11 and 270A.
[54] CLCA s 21 as in force at 24 September 2002.
On 29 August 2003 the appellant was sentenced to imprisonment for seven years with a non-parole period of four years. Both the sentence of imprisonment and the non-parole period were specified to commence on 26 September 2002, being the date upon which the appellant had been taken into custody. This meant that the appellant became eligible for parole on 26 September 2006. However, because of the charge of murder he has not been released. The appellant has therefore already served in custody some five and a half years of the sentence of seven years imposed for the offence of wounding with intent to do grievous bodily harm.
This is the second appeal to this Court against a decision of the judge refusing to stay the prosecution as an abuse of process.[55] The abuse of process which the appellant relied upon in the previous appeal was said principally to lie in the re-litigation of issues involved in the decision of 14 August 2003 and in the oppression and vexation to the appellant in having to face a further trial when he has already served a substantial portion of the sentence imposed on 29 August 2003.
[55] R v P, NJ (No 2) [2007] SASC 135; (2007) 99 SASR 1.
The present application and appeal focuses on the prospect of double punishment. They have been prompted by the amendments to the Criminal Law (Sentencing) Act 1988 (SA) (CLSA) effected by the Criminal Law (Sentencing) (Dangerous Offenders) Amendment Act 2007 (the 2007 Amendments) which came into operation on 1 November 2007. The appellant’s principal submission is that the 2007 Amendments expose him unavoidably to double punishment, such as to render the prosecution for murder an abuse of process.
The judge at first instance held that in the event of a conviction for the offence of murder, double punishment could be avoided by a direction that the head sentence and non-parole period be taken to have commenced on 26 September 2002, being the date upon which the sentence for the offence of wounding with intent to inflict grievous bodily harm was taken to have commenced.
The appellant contended that ss 30 and 32 of the CLSA would not permit any backdating at all of the sentence which would have to be imposed in the event of a conviction for murder. Alternatively, he submitted that any such backdating as is permissible would not allow the commencement of a sentence for murder to be before 28 June 2004 (the date of death of DH). Finally, the appellant contended that even if the sentence for murder could be backdated to 26 September 2002, it would still involve double punishment.
The respondent’s submissions accepted that unavoidable double punishment may indicate that the prosecution for murder involves an abuse of process. The Director submitted, however, that double punishment could be avoided by the sentencing judge directing that the sentence imposed for murder be made concurrent with the sentence imposed by the judge on 14 August 2003 and, for that purpose, its commencement backdated to 26 September 2002.
The provisions of the CLSA relevant to the appeal are set out in the reasons of Duggan J and of Gray J.
The appeal raises the following issues:
1.Does s 30(1) of the CLSA authorise a sentencing judge to direct (other than in the circumstances to which s 30(2) refers) that a sentence, including a non-parole period, be taken to have commenced on a date antecedent to the date of sentencing? The appellant’s submissions on this issue challenged the correctness of this Court’s decision in R v Colson.[56]
2.Even if s 30(1) would otherwise permit the commencement of a sentence to be backdated, is that operation now precluded by the 2007 Amendments to s 32 of the CLSA in relation to sentences for murder?
3.If the commencement of the sentence for murder could be backdated, can that backdating be to a date which antecedes the date of the death of the victim?
4.If the sentence for murder can be backdated to 26 September 2002 so as to be concurrent with the sentence imposed for the offence of wounding with the intention of inflicting grievous bodily harm, would it nevertheless involve double punishment in the circumstances of this case?
5.If so, is that double punishment sufficient to warrant an order staying the prosecution?
[56] [1999] SASC 184; (1999) 73 SASR 407.
As to the first three of those issues, I agree with the reasons of Duggan J. That means that I can turn directly to issue No 4.
Would a Backdating of the Sentence for Murder Avoid Double Punishment?
The Effect of Backdating a Sentence of Life Imprisonment
If the appellant is convicted of murder, the mandatory sentence of life imprisonment will have to be imposed.[57]
[57] CLCA s 11.
The effect of a sentence of life imprisonment has been modified by the provisions in the CLSA concerning non‑parole periods and by certain provisions of the Correctional Services Act 1982 (SA) (CSA). Without those provisions the imposition of the mandatory sentence of life imprisonment would plainly involve a duplication of punishment. That is because the sentence would amount to the imposition of an additional penalty at a time when the appellant has completed the greater part of the sentence imposed on 29 August 2003. In a practical sense it would be the imposition of a sentence of indefinite duration for an offence with elements common to the offence for which a sentence of finite duration was imposed. This would occur at a time when the latter has been substantially served. The backdating of the commencement of the sentence of life imprisonment would not avoid that duplication. The commencement date of such a sentence is immaterial because (without the statutory modifications) the sentence would operate for the balance of the offender’s life. The taking account of time already served in a head sentence of life imprisonment is of no practical consequence because it is impossible to deduct a fixed number of days (being the period already served) from the sentence.[58]
[58] Cf Fox and Freiberg “Sentencing: State and Federal Law in Victoria” 2^nd Edn OUP at [9.804].
However, the combined operation of the provisions concerning non-parole periods in the CLSA and the provisions in the CSA concerning release on parole indicate that a sentence of life imprisonment will not necessarily result in a liability to imprisonment for life.
By s 32(1) of the CLSA, a court sentencing a person to imprisonment must fix a non-parole period. That obligation is subject to certain qualifications. The court may, in some circumstances, decline to fix a non-parole period (s 32(5)) but for practical purposes, that possibility can presently be ignored. In the case of a sentence of life imprisonment for murder, the non-parole period must be at least 20 years (s 32(5)(ab)). That minimum period can be reduced only for “special reasons” (s 32A(2)(b)), the nature of which are narrowly circumscribed (s 32A(3)).
The CSA provides for the effect of a non-parole period. It is necessary to refer only to the provisions concerning a non-parole period imposed in respect of a sentence of life imprisonment. An application for release on parole may be made by a prisoner, the Chief Executive Officer of the Department of Correctional Services, or by an employee of the Department authorised by the Chief Executive Officer (s 67(1) and (2)). Such an application may not be made more than six months before the expiration of the non-parole period (s 67(3)). If satisfied that release on parole is appropriate, the Parole Board may recommend to the Governor that the prisoner be released on parole (s 67(6)) and, amongst other things, must then recommend a period of between three and 10 years during which the prisoner is to remain on parole. The Governor may, on receiving the Parole Board’s recommendation, order that the prisoner be released on parole for a period specified in the order, being a period of between three and 10 years (s 67(7)). It is not necessary for the Governor to act in accordance with the recommendation of the Parole Board, and there have been instances in the past when the Governor has not done so, including by refusing the release on parole. Persons released on parole will, unless the release is later cancelled or suspended, remain on parole for the period specified in the Governor’s order (CSA s 70(1)). On the expiry of that period, the sentence of life imprisonment is taken to have been wholly satisfied (s 70(2)).
Parole is cancelled automatically on the imposition of an unsuspended sentence of imprisonment for an offence committed while the person is on parole (s 75(1)) or for the breach of a condition which the Parole Board designates will have that effect (s 73(1)).
In summary, a sentence of life imprisonment does not necessarily result in offenders being subject to the sentence for the balance of their lives. In those cases in which the offender is granted parole, the offender will be wholly discharged from the sentence upon completion of the period (between three and 10 years) specified in the order for release on parole. The offender will, however, remain subject to the sentence of life imprisonment if the Parole Board declines the application for release on parole, if the Governor rejects the Parole Board’s recommendation for release on parole, or if the offender, once released, breaches a designated condition of the parole or commits a further offence or offences for which an unsuspended sentence of imprisonment is imposed. Put more shortly, offenders may remain subject to the sentence for the balance of their lives, depending upon events in the future, some of which will be within their own control, and some which will not.
The backdating of the commencement of the sentence of life imprisonment would be of advantage to the appellant in this case. If the commencement of the sentence of life imprisonment was directed to be the date upon which he was taken into custody (26 September 2002) the non-parole period fixed in his case would be taken to have commenced on that same day.[59] In that way, the appellant would become eligible for parole sooner than would otherwise be the case. Further, if granted parole, he may commence serving the period of between three and 10 years before being discharged wholly from that sentence sooner than would otherwise be the case. However, it is not inevitable that the appellant will be granted parole either at the time when he first becomes eligible, or at some later time. Whether or not that will occur cannot be known at the present time. If it should turn out that the appellant is not granted parole, then the backdating of the commencement of the head sentence and of the non-parole period will, for the reasons earlier given, have been of no effect in avoiding double punishment.
[59] CLSA s 30(5).
I conclude therefore that a backdating of the commencement of the sentence imposed for murder would involve the possibility, but not the inevitability, of double punishment in the time served in custody. It cannot be said however that backdating would avoid altogether the possibility of double punishment.
Would Backdating Avoid Double Punishment for the Common Elements of the Two Offences?
The appellant’s submissions indicated another way in which a sentence for murder would involve double punishment. Even with a backdating of the sentence to make it concurrent with the sentence imposed on 29 August 2006, he would, it was submitted, be punished twice for the common elements of the two offences.
When the appellant was sentenced for the offence of wounding with intent to do grievous bodily harm on 29 August 2003, the judge had regard, amongst other things, to the act of stabbing, the circumstances in which that act occurred, the appellant’s intention, and the very serious and permanent effects of the stabbing upon the victim. The appellant contends that if convicted of the murder of DH, the mandatory sentence of life imprisonment, together with the mandatory minimum non-parole period required by s 32(5)(ab) of the CLSA, will include punishment for elements for which he has already been punished, including for the physical actions constituting the offence and the accompanying intention to inflict grievous bodily harm.
The appellant relied on the decision in Pearce v The Queen.[60] In that case the defendant was charged with two offences arising out of a single incident: wounding with intent to inflict grievous bodily harm; and breaking and entering a dwelling and, while present, inflicting grievous bodily harm on an occupant. Both offences involved the one assault causing grievous bodily harm. An application for a stay of the prosecution of one or other of the two counts on the basis of double jeopardy was refused. The defendant was sentenced to imprisonment for 12 years for each of these two offences but those sentences were ordered to be served concurrently. The High Court dismissed the appeal against the refusal of the stay of prosecution, but allowed the defendant’s appeal against sentence. In relation to the latter, McHugh, Hayne and Callinan JJ held:
[60] [1998] HCA 57; (1998) 194 CLR 610.
To the extent to which two offences of which an offender stands convicted contain common elements, it would be wrong to punish that offender twice for the commission of the elements that are common. No doubt that general principle must yield to any contrary legislative intention, but the punishment to be exacted should reflect what an offender has done; it should not be affected by the way in which the boundaries of particular offences are drawn. Often those boundaries will be drawn in a way that means that offences overlap. To punish an offender twice if conduct falls in that area of overlap would be to punish offenders according to the accidents of legislative history, rather than according to their just deserts.
In the present case we need not decide whether this result is properly to be characterised as good sentencing practice or as a positive rule of law. There is nothing in ss 33 or 110 of the Crimes Act more generally which suggests that Parliament intended that an offender such as the appellant should be twice punished for his inflicting grievous bodily harm on his victim. …
It is clear in this case that a single act (the appellant's inflicting grievous bodily harm on his victim) was an element of each of the offences under ss 33 and 110. The identification of a single act as common to two offences may not always be as straightforward. It should, however, be emphasised that the inquiry is not to be attended by "excessive subtleties and refinements". It should be approached as a matter of common sense, not as a matter of semantics.
The trial judge sentenced the appellant to identical terms of imprisonment on counts 9 and 10 and made those sentences wholly concurrent. We can only conclude that the sentence on each of those counts contained a portion which was to punish the appellant for his inflicting grievous bodily harm on his victim. Prima facie, then, he was doubly punished for the one act.
Does that matter if, as was the case here, an order was made that the sentences be served concurrently?
To an offender, the only relevant question may be "how long", and that may suggest that a sentencing judge or appellate court should have regard only to the total effective sentence that is to be or has been imposed on the offender. Such an approach is likely to mask error. A judge sentencing an offender for more than one offence must fix an appropriate sentence for each offence and then consider questions of cumulation or concurrence, as well, of course, as questions of totality.
Sentencing is not a process that leads to a single correct answer arrived at by some process admitting of mathematical precision. It is, then, all the more important that proper principle be applied throughout the process.
…
Looked at overall, it may well be said that the effect of the sentences imposed on this appellant was not disproportionate to the criminality of his conduct. Nevertheless, we consider that the individual sentences imposed on counts 9 and 10 were flawed because they doubly punished the appellant for a single act, namely, the infliction of grievous bodily harm. Further, to make the sentences imposed on those two counts wholly concurrent may also be said to reveal error in that to do so failed to take account of the differences in the conduct which were the subject of punishment on each count. …[61] (Citations omitted) (Emphasis added)
[61] Ibid at [40]-[49]; 623-24.

The appellant’s submissions emphasised the italicised passages and, in particular, the proposition that an offender should not be punished twice for the elements of two offences which are common, the statement that the issue of double punishment should not be determined simply by an enquiry about the effective total period of imprisonment to be served, and the statement that, despite the order for concurrency, the sentences were flawed because of the double punishment for a single act which they involved.
In his separate judgment in Pearce, Kirby J also cautioned against regarding an order for concurrency as avoiding the imposition of double punishment:
It is tempting to regard the imposition of common concurrent sentences as a practical way of avoiding the risk of double punishment. There may be cases where it does so. But there are distinct risks in proceeding in that way. The duplication of sentences, although to be served concurrently, may yet amount to double punishment. The differential features of the successive offences (which alone justify double prosecution and punishment) may not be taken into account, adequately or at all. In short, the judicial discretion exercised in the consideration of punishment may not readily provide the means of curing the defects of unjustifiable vexation or the risks of double punishment. In the imposition of a sentence in such circumstances, great care must be taken to avoid double punishment for the same conduct. That care should be manifest in the reasons of the sentencing judge.[62]
[62] Ibid at [121]; 650.
Relying on these passages, the appellant submitted that he would still be punished twice for the same conduct even if the sentence for murder was backdated. The sentence imposed on 29 August 2003 was a “full” sentence for the offence of wounding with the intention of inflicting grievous bodily harm and, because of the mandatory sentence for murder, he would, in the event of conviction, also receive a “full” sentence for that offence. The sentence for murder could not be reduced so as to avoid punishing him again for the elements common to both offences.
The circumstances of this case are unusual. The unusualness arises from the combination of the circumstance that the appellant has already substantially completed service of the sentence imposed on 29 August 2003, the mandatory nature of the sentence for murder (including, since the 2007 Amendments came into effect, the mandatory minimum non-parole period), and the fact that there are substantial elements common to both offences.
Unlike Pearce, in the event of a conviction for murder in this case, there would not be a simultaneous sentencing for two offences arising out of the one episode. The flexibility which the Court would otherwise have if sentencing simultaneously for the two offences will not be available in this case. The sentence for wounding with intent to inflict grievous bodily harm has already been imposed and substantially served. It cannot be modified by the sentencing judge. As already noted, the mandatory nature of the sentence for murder also removes the Court’s ability to adapt a sentence to meet the particular circumstances of this case.
The present circumstances are analogous to those cases in which a person who has been convicted and sentenced for an offence of conspiracy is later charged with an offence constituted by the overt acts of the conspiracy. In such cases, the person is not to be doubly punished for those overt acts. So much is apparent from R v Hoar.[63]
We are told that, notwithstanding the conviction for conspiracy, there are pending against Hoar charges for summary and substantive offences under the Fisheries Act based on the same transactions as those involved in the offence of conspiracy. The Solicitor-General for the Northern Territory informed this Court that the Crown had not decided whether to proceed with these charges. Twelve of the offences are alleged to be constituted by acts which were the overt acts of the conspiracy, and were taken into account by the Federal Court in determining the sentence to be imposed for that offence. Further prosecutions would therefore seek further punishment for the same acts.
That suggests that the Crown’s advisers have overlooked a practice, if not a rule of law, that a person should not be twice punished for what is substantially the same act … It has long been established that prosecutions for conspiracy and for a substantive offence ought not to result in a duplication of penalty. In 1848 Lord Denman CJ in Reg v Button said: “If, however, a prosecution for a larceny should occur after a conviction for a conspiracy, it would be the duty of the Court to apportion the sentence for the felony with reference to such former conviction”.
In exceptional cases the element of concert may justify a more severe penalty for conspiracy than for the substantive offence which the conspirators commit … , but where a court, imposing a penalty for conspiracy, takes into account the overt acts of the conspiracy, it would be wrong to impose a further penalty in respect of those acts.[64] (Citations omitted)
[63] (1981) 148 CLR 32.
[64] Ibid at 37-8.
The respondent submitted that the difficulty raised by the appellant would be avoided by what was described as the orthodox sentencing process outlined in Pearce, namely, the sentencing judge imposing the “appropriate” sentence for the offence of murder, then considering the question of accumulation or concurrence, and then the question of totality.[65] As I understood the submission, both the sentence imposed on 29 August 2003 and the mandatory sentence which would have to be imposed on a conviction for murder should, for this purpose, be regarded as “an appropriate sentence”. The respondent submitted that an order that the sentence imposed for murder be served concurrently was both available and inevitable, and would “necessarily” avoid double punishment. The submission may derive some support from the judgments of Simpson J in R v Hammoud[66] and of Beazley JA in R v Gorman.[67] Each judge expressed the view that the paragraph in Pearce to which the respondent referred required the sentence for an individual offence to reflect the criminality involved in that offence untainted by reference to the other offences for which the offender is to be sentenced.[68]
[65] Pearce v The Queen [1998] HCA 57 at [45]; (1998) 194 CLR 610 at 624.
[66] [2000] NSWCCA 540; (2000) 118 A Crim R 66.
[67] [2002] NSWCCA 516; (2002) 137 A Crim R 326.
[68] Hammoud at [8]; 67; Gorman at [6]; 328.
In my opinion, the course of action proposed by the respondent, which is well adapted to the circumstance of simultaneous sentencing for multiple offending, does not adequately address the issue of duplication of punishment which arises in the unusual circumstances of this case. In the first place, the submission overlooks the limitation on the sentencing judge’s discretion which would exist in this case. The appropriate sentence on a conviction for murder will be the mandatory orders required by statute, including the mandatory minimum non-parole period. That sentence cannot be reduced on account of a previous sentence. Secondly, fixing one sentence without reference to the sentence imposed for the other will mean that no account is taken in the second sentence of the fact that elements common to both offences have already been punished in the first sentence. Thirdly, the submission assumes that regard need only be had to the total effective sentence to be served by the appellant. Both Pearce and the later decision of Johnson v The Queen[69] make it plain that this is not appropriate. It is not just a question of “how long” the appellant will serve in custody. An order that the separate sentences be served concurrently does not avoid altogether the double punishment involved for the elements which are common to each offence nor the appellant being convicted of both offences.[70]
[69] [2004] HCA 15 at [26]; (2004) 205 ALR 346 at 356.
[70] A conviction is part of the sentence: R v Sessions [1998] 2 VR 304 at 313 per Hayne JA.
Summary of the Effect of Backdating
In summary, my opinion is that a backdating of the commencement of the sentence for murder in this case to 26 September 2002 will not remove altogether the double punishment of the appellant. This is because the appellant may in the future have to serve time in custody which takes no account of the period already served. More importantly, in the unusual circumstances of this case, the Court will not be able to avoid imposing a sentence which involves further punishment for the elements which are common to both offences. The concurrency of the operation of the two separate sentences will not avoid that effect.
Should There be a Stay of Proceedings?
The Court has an inherent power to stay criminal proceedings as an abuse of the Court’s process when that is necessary to prevent injustice to an accused.[71] An order staying a criminal prosecution is unusual and is to be made only in exceptional circumstances.[72] The question of whether a prosecution should be permanently stayed as an abuse of process is to be determined by a weighing process involving the balancing of a variety of considerations. These include the requirements of fairness to accused persons, the legitimate public interest in the disposition of charges of serious offences and the conviction of those guilty of crimes, and the need to maintain public confidence in the administration of justice.[73] To these considerations may be added the interests of victims of crime and of their relatives in having those responsible dealt with by the courts.[74]
[71] Jago v The District Court of New South Wales (1989) 168 CLR 23; Rogers v The Queen [1994] HCA 42; (1994) 181 CLR 251.
[72] Jago v The District Court of New South Wales (1989) 168 CLR 23 at 31 per Mason CJ.
[73] Jago Ibid at 30-34 per Mason CJ; Walton v Gardiner [1992] HCA 12 at [49]; (1993) 177 CLR 378 at 395-6 per Mason CJ and Deane and Dawson JJ; Rogers v The Queen [1994] HCA 42 at [5]-[6]; (1994) 181 CLR 251 at 256 per Mason CJ.
[74] See the Victims of Crime Act 2001 (SA) and CLSA s 7A. See also R v P, NJ (No 2) [2007] SASC 135 at [114]; (2007) 99 SASR 1 at 37.
In the present case, in the view I take of the matter, the appellant would be doubly punished if the matter proceeded to trial and, after a conviction, a sentence containing the mandatory elements required by statute is imposed. To some extent, that double punishment may be avoided by an order that the mandatory sentence be taken to have commenced on 26 September 2002. But, as Pearce indicates, the double punishment cannot be wholly avoided in that way. The imposition of a double punishment would be oppressive of the appellant. That circumstance, in my opinion, enlivens the Court’s power to stay the prosecution as an abuse of process.
It is a very serious matter for the Court to countenance double punishment. I cannot detect anything in the CLSA which indicates a legislative intention that an accused may be doubly punished. On the contrary, many of the powers vested by the CLSA seem intended to provide a sentencing court with the flexibility to avoid that consequence. There is the further consideration that this is not a case in which the appellant will, if a stay is granted, have avoided altogether the imposition of a serious criminal sentence. The sentence which he is currently serving is a significant penalty.
In my opinion, this is a case in which the Court’s exceptional power to stay a prosecution as an abuse of process should be exercised. I would allow the appeal, set aside the order of the judge dismissing the application for a stay, and substitute an order that proceedings on the information charging the appellant with the murder of DH be stayed permanently.