Causal Attributions and Sentencing Linda Jane Coates
B.A., University of Victoria, 1989 M.A., University of Victoria, 1991
A Dissertation Submitted in Partial Fulfillment of the Requirements for the Degree of
DOCTOR OF PHILOSOPHY
in the Department of Psychology- We accept this thesis as conforming
to the required standard
Dr. J ^ B e a v i n Bavelas, Supervisor (Department of Psychology)
Dr
Psychology)
Departmental Member (Department of
Dr. R. Routledge, Departmental Member (Department of Psychology)
P r o f e s s ^ K. Jobson, Outside Member (Faculty of Law)
(£) LINDA JANE COATES, 1996 University of Victoria
All rights reserved. May not be reproduced in whole or in part, by photocopy or other means, without the permission of the author.
il h1 Clark, External
Dr. H. H. Clark, External Examiner (Department of Psychology, Stanford University)
(c) LINDA JANE COATES, 1996 University of Victoria
All rights reserved- May not be reproduced in whole or in part, by photocopy or other means, without the permission of the author.
Supervisor: Dr. Janet Beavin Bavelas ABSTRACT
In this dissertation, I examined the construction, function, and consequences of causal attributions in all British Columbia sexual assault trial judgements available through Quicklaw from 1986-1994. Most
judgements contained causal attributions, that is, the judge discursively constructed an explanation of the assault. These attributions were analysable as
internalizing versus externalizing (i.e., under the individual's control or not); situating or saturating (i.e., situated in a particular time and place versus abstracted and pervasive), and as violent or nonviolent (i.e., caused by a choice or tendency to be violent versus a choice or tendency to use alcohol, to be sexually inappropriate, etc.). The preponderance of attributions for persons guilty of sexual assault cast the cause as something other than violence. This
suggests that, despite laws which defined sexual assault as violent per se, judges are not treating these offenses as violent.
These attributions along with two other sets of factors (case parameters identified by Ruby, 1994, and
stranger versus nonstranger status) were then tested to determine which factors were related to sentencing. Causal attributions were significantly correlated with sentence. In particular, nonviolent attributions were significantly related to lower sentences. Only the relatively rare violent and internalizing attributions
(e.g., "he chose to be violent") were significantly correlated with higher sentences.
Most case parameters were either not associated with sentence or were actually correlated in the
opposite direction. For example, breach of trust was negatively associated with sentence: judges gave
offenders who assaulted children in their care
systematically lower sentences. Stranger status was significantly correlated with the offender receiving a higher sentence, even when the stranger had not
physically touched the victim.
Overall, these findings illustrate the usefulness of examining judges’ discourse as well as the facts of the case. When judges discursively constructed an assault as nonviolent, the sentence was lower. When they attributed the assault to violent causes and cast the offender as responsible, the sentence was higher.
Moreover, the results suggest that the judges are not dealing with sexual assault and other sexualized
offenses in the way that is consistent with the relevant legislative laws and sentencing principles. In particular, the legal system fails to protect children from those charged with their care and protection.
Examiners :
Dr. J* Beavin Bavelas, Supervisor (Department of Psychology)
e. Departmental Member (Department of
Dr. R. Routledge, %partmental Member (Department of Psychology)
^_________________________________
Professo^K. Jobson, Outside Member (Faculty of Law)otessc^K. Jobson, Outsxd Û. H. Clark, External E
Dr. H. H. Clark, External Examiner (Department of Psychology, Stanford University)
Table of Contents
Title Page i
Abstract ill
Table of Contents vi
List of Tables x
List of Figures xii
Chapter 1: Discourse Analysis and the Legal System 1
Previous Research 10
Affectionate Lamguage for Violent Acts 15
Psychological Explanations 19
Chapter 2 : Sexual Assault and the Legal System
in Canada 22
Statistics on Sexual Assault 30
Reports, Charges, and Attrition 33
Conviction Rates 36
Range of Sentence 36
Sentencing Principles 38
Chapter 3 : Attribution 43
Attribution as an Individual Perception 43
Attribution Theories 44
Attribution Research 45
Attribution as Social Construction 47 Application to Sexual Assault Trial Judgements 66
Chapter 4: Method 73
Database 73
Analyses 74
Developing the Systems of Analysis 74 Analysis of Causal Attributions 75
Reliability 80
Demographic Characteristics of the Sexual Assault
Trial Judgements 81 Courts 82 The Accused 82 The Complaineints 83 Gender 83 Age 83
Relationship of the Accused to the
Complainant 84
Breach of Trust 84
Possibility of Consent 85
Crimes 87
Charges 87
Nature of the Criminal Acts 88 Magnitude and Impact of the Crime 89
Criminal Code Maximum 89
Severity Scale 90
Victim Suffering Scale 91
Collateral Violence 92
Guilt and Sentencing 92
Sentencing 92
Sentencing or Incarceration 92 Length of Incarceration Per
Victim 94
Proportion of Sentence 94 Proportion of Determinate
Sentence 95
Disposition or How the
Sentence Was to be Served 95
Counselling 97
Chapter 5: Results 98
Causal Attributions 98
Attributions and Sentence 101
Case Parameters and Sentence 103 Previous Criminal Record 103 How the Offender's Guilt was
Established 104
Magnitude and Impact of the Crime 105
Violence 105
Alcohol and Drug Abuse 106
Stranger Versus Nonstranger 107 Comparison of Factors that
Predict Sentence 107
Chapter 6: Discussion 110
Causal Attributions 110
Legal Approaches to Sentencing 117 "Why" Do Judges Do This? 119
References 123
Appendix A: Scoring Instructions 133 Appendix B: Attributions and Case Parameters
Correlated with Sentencing and
Proportion of Sentencing 146
Tables 151
Figure Caption 171
Table 1: Alternative Vocabularies for Describing
Acts of Sexual Assault 151
Table 2: Combinations of Attribution Functions 153 Table 3: British Columbia Courts Which Heard the
Cases 154
Table 4: Ratings of the Possibility that the
Complainants Could have Consented 155 Table 5: Frequencies of Criminal Acts 156 Table 6: Victim Suffering Scale Scores for
Strangers versus Non-Strangers, and
Children versus Adults 157
Table 7: Mean Sentence in years for Each
Sexualized Crime in the Database 160 Table 8: Length of Incarceration Period
Per Victim 161
Table 9 : Meein Proportion of Sentence for
Each Crime 162
Table 10: Mean Proportion of Determinate
Sentence For Each Crime 163
Table 11: Disposition: Types of Sentences
(i.e.. How Offenders were Ordered to Serve
Table 12 : Frequency of Causal Attributions
for each Function 165
Table 13: The Correlation Between Proportion of Determinate Sentence and Attribution
Functions 166
Table 14: Case Parameters Correlated with
Proportion of Determinate Sentence 168 Table 15 : Correlations Among Case Parameters
and Attributions Related to Proportion
List of Figures
Figure 1: Venn Diagrams Showing Attributions
CHAPTER 1:
DISCOURSE ANALYSIS AND THE LEGAL SYSTEM Discourse is central to the law. Through
discourse, legal professionals accomplish tasks such as gathering statements, questioning witnesses, making arguments, end giving reasons for a judgement. The discourse used in legal judgements is particularly important because it not only expresses the state of the law, it also affects future law by forming part of the commonlaw that will guide decisions in the future.
This dissertation is based on the assumption that the words that are used to describe acts in a legal setting give meaning to those acts and create certain realities. Danet (1980) pointed out that the words used to describe a particular act cannot be separated from what happened. The words not only represent what happened, they also create versions of what happened. For example, describing an act as a "slap" implies that the actor had a negative motive whereas describing the same act as a "tap" implies a friendlier motive. Thus, words create a version of the events; in court, one of these versions may be that a crime was committed (e.g..
criminal event (e.g., sexual intercourse) occurred. This dissertation focuses on discourse about sexual assault, and the terms we use for these crimes illustrate Danet's point, above. The legal term for these acts is sexual assault, which is therefore the term I must use when describing the act as a legal event. However, as Boyle (1985) has pointed out "in a world in which sex was understood to be a consensual activity" we would not leibel an assault as sexual (p. 104). Thus, the term sexual assault is an oxymoron; sexualized assault would be more accurate, and I will use that term when not referring to legal charges.
Whenever possible, I prefer to use physical
descriptions that do not characterize the act or, where assault has been legally established, terms that
emphasize the assaultive rather than the assumed sexual nature of the act. We (Bavelas, Chovil, & Coates,
1993) illustrated the choices of terms for the same act; see Table 1.
There has been a plethora of research identifying pervasive cultural attitudes about sexual assault, often called "rape myths" (Burt, 1991; Herman, 1984;
Lipton, McDonel, & McFall, 1987; Shetland, 1988): that rape is not all that bad because it is sex euid
everybody likes sex; that women really like having sex with men but feel compelled to deny it and call it rape; that women commonly "cry rape" when they want to get back at a man; that women who "come on" to men deserve what they get; that men have such poor communication skills that they cannot understand a woman when she says "No"; or that rapists are men who simply were so sexually aroused that they could not stop themselves from raping. Yet there has been surprisingly little research on the language used to describe actual sexual assaults.
Some researchers have examined the language used in our court system. All of these researchers
concluded that although Ccinada's laws now define sexual assault as a violence issue rather them a sexual, moral issue, the courts have not necessarily adhered to this change (see Boyle, 1985; Gunn & Minch, 1988; Ruebsaat, 1985; and Smart, 1989). In her analysis for the
Department of Justice, Ruebsaat (1985) found that the poUrts frequently focus on morality. Similarly, other researchers have emphasized that the judiciary needed
to "use terms that are consistent with the new law" (Gunn & Minch, 1988, p. 108). Using inappropriate terms trivialises the violent nature of sexual assault
(Boyle, 1985; Gunn & Minch, 1988; Ruebsaat, 1985). For example, the term sexual assault has also been
controversial because it implies that a unilateral
violent act could somehow be sexual, and therefore also a consensual activity (Boyle, 1985; Coates, Bavelas, & Gibson, 1994). Consistent with this, Kelly and Radford
(1990/91) argued that
the law not only reflects but also constructs a very limited definition of sexual violence, and thereby plays a significant role in denying or trivialising women's experience of male sexual violence. (p. 39)
A significant factor in the social and legal construction of sexual violence is the traditional
notion that sexual assaults are committed by strangers. However, the evidence suggests that a far greater
proportion of sexual assaults are committed by someone the victim knows (Bohmer, 1991; Koss, 1992). A male friend whom a woman has known for some time is unlikely to jump out of a bush with a weapon to sexually assault
might use personal threats or verbal abuse to
accomplish the assault (Bownes, O'Gorman, & Sayers, 1991; Rennie, 1992). Meuiy people may not recognise this as an assault because it does not fit the
traditional stereotype of stranger rape. A woman is more likely to physically resist (i.e., fight or flee) a stranger than an acquaintance who is trying to
sexually assault her (Bownes, et al., 1991). Unfortunately, by not engaging in what legal
professionals deem to be "appropriate" ways for women to resist such attacks by a stranger, the victim may limit her chances of receiving any help from the
justice system (Estrich, 1987). Counsel and judges may question the woman's credibility or trivialise her
experience if her resistance does not fit a male combat perspective, that is, if her actions do not resemble a man resisting am attack by a man of similar size and strength (Coates, et al., 1994; Estrich, 1987; Kelly & Radford, 1990/91). As Muelenhard and Schrag (1991) noted:
Force by a man is usually defined in terms of the amount of resistance used by a woman. If the
woman did not physically resist, the courts see the maui as not having used force, even if the woman did not consent. (p. 125)
Changes introduced in 1992 should have eliminated defining consent in terms of the victim's resistance. However, whether the courts have made this change is cin
empirical question, to be answered by future research. Members of our research group began to study descriptions of sexual assault in 1991. At that time, we heard news reports of many infamous cases in Ccinada and the United States:
In British Columbia court decisions, a judge lowered an offender's sentence because he considered the victim, a three-year-old girl, to be sexually aggressive; another
judge acquitted a man because he decided that when a woman says "No", the meaning is
ambiguous.
A priest who was convicted of aggravated sexual assault talked about people's
unrealistic expectations that priests be celibate ("Everyone knows a priest needs love"). He also talked about the assaults in
mysteriously appear in his bed. He cast these victims as willing participants in sexual activity. In short, the priest used language that obfuscated his assaults.
A man who had raped and otherwise
tortured several boys to death talked about abducting the children. This man said that if the children said "No", he would let them go.
My first reaction to this latter description was to question the truth of this account, asking "What does he mean the boys didn't say 'No'? "I don't believe the boys didn't protest!". An interview with the father of one of the murdered boys provided me with the answers to my questions. He pointed out that the murderer knocked the wind out of the boys, so that they could not easily move, let alone talk. The father also pointed out that the boys had resisted the tortures
inflicted upon them by the offender. For example, the children had cried and screamed. When I thought of what the murderer had done to the boys, I was
Something about these interviews disturbed me, in addition to the facts of the cases, and after much thought I realized why: I had been drawn into the
offenders' accounts. For example, I had questioned the child rapist and murderer's account and doubted its validity, but while doing so, I had been unwittingly recruited into this man's version of events. I had asked all of my questions within the bounds or
framework of the murderer's description (e.g., "I don't believe that the boys didn't protest"). I should have asked questions from outside his description: "What purpose does this description serve?", "What was the victim's experience?". By asking questions within the murderer's framework, I collaborated in perpetuating the view that what he said had any credence at all. Unwittingly, I supported the reality that he was trying to create for us (as he had no doubt tried to do to the victims).
It became clear from these news reports cuid
interviews that offenders and others strategically use communication devices to disguise acts of assault. We decided to investigate those devices. For example, what devices did the judge use to focus on the
possibility that a young child was sexually aggressive or that the word "No" had smother meaning? As
communication researchers, we felt that we might be able to elucidate these strategies. We hoped that this knowledge could have a beneficial effect for the
victims and for the prosecution of perpetrators. We did not assume that the judgements and
interviews we found so disturbing were representative of the majority of cases, but even if they were a small minority of cases, they were still importsmt.
Procedures to safeguard human rights must be present at an individual level. Protecting every individual from violation will ultimately result in protecting all people. Having the general philosophy of protecting people will not necessarily protect emy individual's human rights. A general human rights philosophy is merely a promise of justice. One must take action at an individual level to fulfil that promise; otherwise the general principle is just a broken promise of justice. As the poet William Blake remarked
He who would do good to another must do it in minute particulars, "General Good" is the plea of
the scoundrel, the hypocrite, and the flatterer, (in Stevenson, 1971, // 55-58, pp. 741)
One must act upon every case, no matter how
unrepresentative it may be of the majority of cases. From a humeui rights perspective, every case— or more accurately, every person— is important.
Previous Research*
In order to get some idea of the court system's handling of sexual assault cases, we examined written sexual assault trial court judgements, because they were readily and inexpensively available. In an
initial study of a small random sample of British Columbia trial court judgements, we identified five anomalous themes (Coates, et al., 1994). First, we found that these judgements frequently described sexual assaults in erotic, romantic, or affectionate terms. Many of these descriptions seemed more appropriate to romance novels than court judgements. For example, they frequently used erotic terms such as "sexual intercourse," and "fellatio" to describe acts of assault. Also, affectionate terms such as "fondle"
'When no citation is given, the data are as-yet-unpublished results of continuing research.
appeared in the judgements. These descriptions both minimized and mutualized the violence, thereby changing a unilateral act of violence into a joint or
collaborative action.
Second, contrary to Canadian law which defines sexual assault as violent per se, the judgements
frequently described the sexual assaults as nonviolent. The descriptions minimized the violence inherent in any assault in several ways: by comparing it to other
atrocities that the offender could have committed but did not; by suggesting that "internal" violence (e.g., forced penetration) was not as serious as external violence; or by focusing on collateral violence (i.e., violence additional to the sexual assault itself) as the most serious aspect of the case. In other
judgements, the sexualized violence was simply dismissed, even when there were contusions on the victim's sexual organs.
Third, these judgements tended to minimize the victim's resistance to the assault. They seemed to define appropriate resistance narrowly, as persistent physical struggle. There was implicit criticism of victims who ceased to struggle physically (even when
the struggle was ineffective or dangerous), and the victims' verbal refusals did not appear to be taken as sufficient evidence of lack of consent.
Fourth, and much to our surprise, the judgements frequently described convicted offenders as being of "good character". In fact, in our small random sample of 13 cases, one offender, who had (among other violent acts) raped a woman twice, was described as having an
"impeccable character". Another case said that a man who had sexually assaulted his stepson for several years was of "exceptional character".
Fifth, the judges constructed sentences in such a way that the accused was not the grammatical agent of his actions. They did not use clear active voice
(subject-verb-object) to describe the offender's assault. Instead, they often omitted the actor from the sentence, leaving an incomplete description of the assault. For example, "Ms. X sustained some bruises and a temporary limp" rather than "The accused hit Ms. X hard enough to inflict some bruises and cause a temporary limp" (Coates et al., 1994, p. 197).
We stressed that these five themes were not just random errors; instead, they were part of an overall
pattern. We used Wetherell and Potter's (1988) concept of interpretative repertoire to explain this pattern. When describing an event, one chooses particular words that draw upon an interpretative repertoire to fill out a specific version of events (Wetherell & Potter,
1988). Interpretative repertoires
can be seen as the building blocks speakers use for constructing versions of actions... . Any particular repertoire is constituted out of a restricted range of terms used in a specific stylistic and grammatical fashion. (p. 172)
For example, the terms "development" and "progress" are drawn from a different interpretative repertoire them the terms "destruction" and "exploitation".
Unfortunately, in cases of sexualized assault legal professionals, as well as the public, seem to have two equally inappropriate interpretative
repertoires, those of stranger rape and consensual sex (Coates, et al., 1994; Bavelas, Chovil, Coates, &
Collins, 1995). Both of these repertoires are
inappropriate for describing the majority of sexualized assaults, which are unilateral, violent acts committed by someone the victim knows. The victim's immediate
response to these attacks may not resemble those of a victim who is attacked by a stranger, who immediately recognizes her or his damger, and who physically
resists the attack. Moreover, offenders who assault people they know usually do not resemble the stereotype of the deranged, itinerant stranger.
We suggested (Coates, et al., 1994) that when the interpretative repertoire of stranger rape does not fit the situation at hand, the courts use the consensual sex repertoire to describe the crime before them. Thus, in keeping with the fact that more people are assaulted by people they know than by strangers, the judges' language was more suited to the consensual sexual activity repertoire. That is, we found that
acts that had been legally established as sexual assaults were often described as
erotic, non-violent acts; they involved (but were not necessarily the responsibility of) persons of good character; and they had been insufficiently resisted by the victim,
Affectionate Language for Violent Acts
In a subsequent analysis of all the available written sexual assault judgements in British Columbia^ we found, once again, a predominance of erotic,
romantic, and affectionate language to describe the assaults. Less frequently, legal professionals used physical descriptions that concentrated on describing the act in question while neither casting the act as an assault nor something non-criminal (e.g., as sexual). Even less frequent were descriptions of the act as violent or language conveyed disapproval.
Moreover, erotic, romantic, and affectionate language was most common in familial assaults (e.g., father/son) and least common in stranger assaults. Finally, when judges did use language capturing the violence inherent in the assault, it was most likely to be in a case where the offender did not know the victim and least likely to be in a case where a family member assaulted a child (e.g., daughter).
Thus, it appears that judgements most frequently consist of erotic, romantic, or affectionate
descriptions of inherently violent acts (particularly when the victims are children). These findings support
the notion that the courts are fluctuating between the two interpretative repertoires of consensual sexual activity and stranger rape. Judges describe only
stranger rape as violent and non-sexual; they describe all other sexualized assaults, including crimes against children, as sexual and non-violent.
Thus, our data concurred with Ruebsaat (1985), who concluded that "the courts have had trouble shifting their focus from morality to violence" (p. 107). More specifically, the courts are treating an autonomous, individual action (sexualized assault) as if it were a joint or mutual action (sexual activity). It is little wonder, then, that a Law Society of British Columbia report on gender equality and the law concluded that the justice system "is failing battered women and victims of sexual assault" (1992, p. 3).
The difference between conceptualizing sexualized violence as sex (albeit unwauited or violent sex) and as assault is the difference between joint and autonomous actions (Clark, 1992). When two or more people
collaborate in performing some action, they accomplish a joint action (Clark, 1992). One individual alone cannot accomplish these actions; instead, the actions
of one person combine with the efforts of at least one other person to collaboratively accomplish the joint act. An individual can accomplish an autonomous action
(Clark, 1992). These acts do not require the
participation of another person, and so they can be accomplished in the presence or absence of other
people. Examples of autonomous actions include a solo, a pirouette, or shaking a stick (Clark, 1992). Notice that these are all unilateral actions.
In contrast, joint actions include such things as a duet, a waltz, or a handshake (Clark, 1992). Notice that the participants of these acts follow the same rules, and are both active participants.
The distinction between individual and joint actions can be applied to violent acts. Beating
someone up is different from boxing with someone. The word "beating" describes an autonomous act; that is, it only takes one person to accomplish a beating. A
person can beat a person in the same way he or she can beat an inanimate object such as a rug. The object or receiver of the beating does not participate in the act. Boxing, however, requires the consent and participation of two people. They follow the same
rules, they respond to what each other does, in
essence, they cooperate in order to accomplish the act of boxing.
Similarly, sexualized assaults (such as rape) are autonomous actions, whereas sexual activities (such as sexual intercourse) are joint activities- A sexualized assault is a unilateral activity that one person does without the participation of the victim. The victim is merely the object of the perpetrator's violent actions. The victim does not cooperate with the perpetrator to accomplish the act. Instead, the perpetrator subdues the victim by whatever means available, for example, by exercising authority, using threats, administering
drugs, or physical force. The perpetrator forces the victim into actions in much the same way as pressing a light switch engages the lights. The victim is an object, not a participant.
In contrast, sexual activities are joint actions. These mutual, consensual activities require the
participation of at least two people to accomplish the act. These reciprocal activities necessarily involve each participant responding to the other (e.g., in the pace, or what they actually do).
The English language has words to differentiate between autonomous and joint actions. For example,
"he", "she", "I", "you", and "me" are frequently used to describe autonomous actions: "He is doing X" or "She did Y " . In contrast, words such as "we", "us", or "they" depict joint actions: "We are doing X", "They did Y " . The word "with" also indicates a joint action: "He is doing X with her", "He had sexual intercourse with her". Thus, using words that cast an assault as a joint action necessarily creates the fiction that the victim was an agent in his or her own assault. We
proposed that using sexual terms for assaults, as found in out data, makes the acts implicitly or explicitly
mutual and therefore consensual. Psychological Explanations
A new direction in our research concerns judges' explanations of violence. In a trial, part of a
judge's, or jury's, task involves making a causal
attribution about the offender's responsibility for the act in question. For exaimple, not only may a judge decide if an accused is guilty or innocent, he or she will frequently also decide why an accused did
attributions, we (Coates & Wade, 1994) conducted a study in which we reliably identified causal
attributions that were psychological in nature. Our sample consisted of 1986 to 1991 British Columbia sexual assault trial judgements where the offender pleaded or was found guilty of sexual assault.
We found that legal professionals auid witnesses frequently attributed these violent acts to
psychological causes, namely.
Alcohol abuse: "It's something, however, that arose out of release of
inhibition obviously to gross consumption of alcohol".
Biological or sexual drives: "There is no doubt on the evidence those offences
occurred because he was unable to control his sexual impulses".
Psychopathology: The offender "is a victim of . . . his own inadequate personality".
Lack of control: "He didn't intend to hurt them; things just got out of hand" .
Character : "This incident was an
isolated one, entirely out of character...". Dysfunctional family upbringing: "It stems from his own background".
Stress or trauma: "He (. . .) has
sought to relieye his stress and anxiety by inflicting stress upon others".
Emotional state : The assault was "a single act of lack of control oyer his emotions".
We (Coates & Wade, 1994) called these explanations psychologizing. That is, these explanations placed the cause of the yiolent actions within the offender's
hypothetical psychology or psychological history. We noted that psychologizing also seemed to place the cause of the yiolence outside of the offender's control.
We proposed that these attributions functioned to minimize the yiolence inherent in sexualized assaults and to separate the offender from ciny responsibility for his criminal actions. This dissertation seeks to confirm these hypotheses.
CHAPTER 2:
SEXUAL ASSAULT AND THE LEGAL SYSTEM IN CANADA Prior to 1983, the law divided sexualized crimes into four main crimes: rape, attempted rape, indecent assault against a female, and indecent assault against a male. At that time, many Canadians were dissatisfied with our existing sexual assault legislation. As Clark
(1993) summarized many of these criticisms:
The old Criminal Code provisions codified common- law practices, enshrined in case law that had
developed around the offences of rape. The result of this accretion was a special set of rules
because they permitted (and in some cases
virtually required) the admission in trials for sexual offences of evidence that was not
admissible in trials for any other crime, (p. 3) One example of these special rules included the
requirement of corroboration for sexualized offences but not for others (Clark, 1993; Ruebsaat, 1985). The net effect of this special rule was that "adult women testifying under oath in relation to purported sexual offences were given less credibility than children so young they were unable to understand the nature of an
oath" (Clark, 1993, p. 4). A woman's testimony in matters concerning sexualized violence was considered to be untrustworthy.
By the second special rule, the courts required victims to report the attack to the police promptly, or else judges took the "delay" as evidence of fraud.
A third special rule permitted defence council to introduce evidence for the sole purpose of attacking the complainant's credibility, which included
introducing evidence regarding the complainant's previous sexual activity and reputation.
A fourth special rule allowed the courts to treat a victim's compliance with or submission to the
offender's demauids as if it were consent. "What this meant in practice was that the victim had to prove lack of consent despite the presence of coercion" (Clark, 1993, p. 5).
According to the fifth special rule, the judges could acquit the accused if he mistakenly believed that the victim had consented, no matter how unreasonab1e that belief was.
Finally, under the sixth special rule, "husbands were legally permitted to rape their wives" (Clark,
1993, p. 5). It is hard to know whether this rule came about from regarding women as the property of their spouses, from equating sexualized violence with sex, or both. What is clear, however, is that the law offered married women no protection from a likely source of sexualized violence.
In addition to the special rules outlined above, critics pointed to the obvious sexism in having more severe indecent assault penalties for cases where the offender assaulted a male (ten year maximum) than a female (five year maucimum). Moreover, defining
penetration of a vagina by a penis as the only serious crime against females ignored the extensive physical and emotional trauma to victims when assaulted in other ways, for example, oral or anal penetration, or vaginal penetration with bottles, branches, etc. (Clark, 1993, p. 12). Also, critics felt that existing laws which defined rape (the most serious offence) as penetration of a vagina by a penis did not sufficiently protect males (Department of Justice, 1990).
These special rules within the legal system functioned "to discourage women from reporting instances of rape" (Clark, 1993, p. iii). Not
surprisingly then, women's and other human rights groups criticised these laws as insufficient and unjust.
In response to the general discontent with the existing laws, and after more than a decade of lobbying by women's groups and other human rights groups,
parliament drafted new legislation. Bill C-127 passed on August 4, 1982 and became law in 1983. Among the lawmakers' goals for this legislation were protecting children, protecting the "integrity of the person" (Law Reform Commission, 1978), and increasing public
confidence in the legal system (Clark, 1993; Roberts, 1990). Legislators hoped to accomplish this by
ensuring that the general rules of evidence applicable to other violent offences against the person would now apply in sexual assault cases (Ruebsaat, 1985). The new rules also de-emphasised penetration as the main criterion of a serious assault (Ruebsaat, 1985) and changed the court's focus from a moral, "carnal
knowledge" issue to a violence issue (Boyle, 1985; Gunn & Minch, 1988; Ruebsaat, 1985; Smart, 1989; see Chapter 1). Lawmakers hoped that these cheuiges would result in a higher percentage of victims reporting sexualized
crimes to the police, a higher percentage of these reports resulting in trials, and greater public
confidence that the legal system was providing justice to these victims (Clark, 1993).
Bill C-127 replaced rape, attempted rape, and indecent assault with three levels of sexual assault. Sexual Assault I (s. 271) carries a Criminal Code maximum of ten years and can encompass all acts of sexualized violence (e.g., from forced oral contact to forced vaginal or anal penetration). Sexual Assault II
(s. 272) is defined as sexual assault causing bodily harm or with a weapon. This crime has a maximum
penalty of 14 years. The final level is Sexual Assault III (s. 273) which is aggravated sexual assault, or an assault where the offender "wounds, maims, disfigures or endangers the life of the complainant" (Martin's Annual Criminal Code, 1992, p. 434). Offenders convicted of aggravated sexual assault could be sentenced to life imprisonment.
Sexual Assault I, as it currently stands, is a hybrid offence in that Crown can attempt to get an indictable or a summary conviction. That is. Crown Counsel typically proceed with sexual assault as an
indictable offence, in which case the ten year maximum would apply. However, sexual assaults can also be tried as a summary offence. Upon summary conviction, judges could give offenders a maximum of 18 months incarceration or a maximum fine of two thousand
dollars. (Until very recently the maximum jail term was six months, see Roberts, 1994, or Ruebsaat, 1985.)
Among some of the other changes provided by Bill C-127 were that (a) men who rape or otherwise assault women they are married to are no longer immune from prosecution; (b) corroboration is no longer required;
(c) equal maximum penalties are set for assaulting either a male or female; cuid (d) recognition that all acts of sexualized violence, not just penetration, are serious. The de-emphasis on penetration also
"eliminated a major hurdle for the Crown auid a major source of further trauma to victims, by eliminating the necessity for medical evidence proving vaginal
penetration" (Clark, 1993, p. 11).
While most people welcomed these changes to sexual assault laws, "there was [still] much dissatisfaction with the resulting reform" (Clark, 1993, p. 8) because it left some problems unsolved. Defence council could
still introduce information regarding the complainant's past sexual activity and reputation. Moreover, Bill C- 127 still permitted a defence of mistaken-belief.
Clark (1993) suggested that this was
meant to cover geing or group rape, but only in relation to a defence of mistaken belief in consent. In such cases, evidence of this sort might allow the defence to lead evidence of a mistaken belief in consent or to plead that the accused had honestly believed in the complaineint ' s consent. The problem seems to come back to the old failure to distinguish genuine consent from passive acquiescence, despite the presence of
coercion. The presence or participation of two or more men in such situations should prove coercion tout court. (p. 8-9)
Thus, Canada still needed further changes to our sexual assault laws. Bill C-49 (often called the "No means no" law), which was introduced in 1992, attempted to solve some of these problems.
Bill C-49 restricts the use of the mistaken-belief defence. The accused's belief in consent can no longer be used as a defence when it comes from self-induced
intoxication, reckless or wilful blindness, or because the accused failed to take reasonable steps to
ascertain consent. Moreover, this bill explicitly defines consent as the voluntary agreement of the complainant to the sexual activities in question. Consent cannot be given by someone else, when the complainant is incapable of doing so, or when the
accused abused a position of trust, power, or authority to induce the complainant into the activities.
Complaineints can express lack of consent verbally or nonverbally at any time.
Bill C-49 also limits the circumstances where the defence can introduce information regarding the
complainant's previous sexual history (i.e., the judge must not only rule that this information is relevant to a particular issue but that it will help resolve that issue). In brief, the bill is a "rape-shield" law, which also explicitly defines consent as voluntary agreement by the complainant and requires that the accused take measures to obtain the complainant's consent.
Statistics On Sexual Assault
The majority of sexual assault victims (84 - 90%) are female (Roberts, 1994; Canadian Advisory Council on the Status of Women, 1985). Roberts (1994) found that 39% of Canadian women report that they were sexually assaulted when they were 16 years of age or older. The Canadian Advisory Council on the Status of Women (1985) estimated that in Canada, a man rapes a woman every 17 minutes and assaults a woman in other sexualized ways every six minutes (as cited in Porteous & Alexander, 1989). But one need only to think of conditions in prisons to realize that Ceuiadian men are also victims of these crimes.
Perpetrators frequently assault teenagers. In a recent survey, teenaged victims made up about 63% of the cases reported to police (Roberts, 1994). Of
reported assaults, complainants named family members as the perpetrators in 28% of the cases (Roberts, 1994). Moreover, in most cases reported to the police (79% for females and 84% for males) the victims knew the person who attacked them (Roberts, 1994). It should be noted that the offenders of these reported crimes tend to be uniformly male (98%; Roberts, 1994).
When looking at these statistics, it is important to consider that the vast majority of victims do not report the attack to the police. A study by the Solicitor General of Canada (1985) found that fewer than 40% of sexualized assault victims reported the crime to the police. Other researchers have found much lower reporting rates. For example, van Dijk, Mayhew, and Killias (1989) conducted an international survey
(which included Ceuiada) and found that only one out of eight victims of sexualized violence reported the crime to the police. Similarly, a more recent study
(Roberts, 1994), found that only six percent of victims reported the sexualized assault to police.
These low reporting rates may reflect the all-too- real horror of women, children, and men living in
situations where they dare not report the attacks. Undoubtedly, the low reporting rates also reflect dissatisfaction with or distrust of the courts'
treatment of sexualized assault. In fact, in one study 44% of victims who failed to report their
victimization were concerned about the reaction and attitudes of the police and courts towards
this type of incident. (Department of Justice, 1990, p. 2)
However, even these very low reporting rates are higher than they were before Bill C-127 took effect in 1983. Roberts (1990) examined patterns of sexualized assault charges (mainly in B. C.) between the years 1983 and 1988. He concluded that the 126% increase in charges of Sexual Assault I indicated that victims' "attitudes towards the criminal justice system have changed, making them more likely to come forward" (p. xiv). Similarly, the authors of a recent report
(Department of Justice, 1990) concluded that "there has been a dramatic increase in the number of sexual
assault reports made to police since 1983" (p. 27). Reporting rate trends suggest that the number of reports will continue to increase yearly. Indeed, Roberts (1994) found a continued increase in reporting rates, with the 1992 rates being 164% above the pre- 1983 rates. So (assuming that there has not been a large increase in the number of assaults during this time period) it appears that Bill C-127 has been
followed by an increased willingness by victims to turn to the legal system for justice. Whether victims
continue to report these crimes to the police will undoubtedly depend upon the legal system's handling of these cases.
Reports. Charcres. and Attrition.
Reporting a sexualized assault to the police does not necessarily mean that the Crown will charge the offender. In fact, the majority of reports to the police (71%) do not result in charges being laid (see University of Manitoba Research Report, as cited in Roberts, 1990). More recently, Roberts (1994) found that only 49% of Sexual Assault I reports resulted in charges being laid against the offender. A greater percentage of reports of Sexual Assault II (53%) and III (58%) result in charges being laid than for reports of Sexual Assault I (Roberts, 1994). This may reflect a bias by legal professionals to view the additional violence in these sexualized assaults as more visible and therefore more serious. Visible physical injuries to non-sexual areas of the victim's body may also be easier to prosecute.
There are various reasons why charges are not laid against an accused. Legal professionals dropped some of these cases (about 14% for Sexual Assault I, 9% for
Sexual Assault II, and 14% for Sexual Assault III) because the police had declared the report to be "unfounded" (Roberts, 1994). Workers dropped other reports because the charge had been "cleared
otherwise". Official and unofficial reasons for police or Crown declaring a report "unfounded" or "cleared otherwise" included: (a) the police deemed that the victim would not be a suitable witness, (b) the police decided that there was a lack of acceptable evidence,
(c) the police were generally unwilling to pursue the complaint, (d) the police decided that the report was false, (e) the accused had diplomatic immunity, (f) the accused died before charges were laid, or (e) the
victim wished the investigation to cease (see Roberts, 1994, or Stainley, 1985). Moreover, Clark cind Lewis (1976) concluded from their study that the Crown was not willing to proceed with prosecution unless a
conviction was fairly likely. The police aided this by screening out cases so that "only the 'best' of even the founded cases proceed" (Clark & Lewis, as cited in Stanley, 1985, p. 98). "Best" here seems to mean those cases which are easy to prosecute and likely to result in a conviction. Roberts (1994) concluded that the
number of sexual assault reports that the police
declared to be unfounded was systematically higher than those for simple assault. Thus, inflated "unfounded" rates for sexual assault have continued to exist even after the enactment of Bill C-127 and Bill C-49.
While Roberts (1990) suggested that this massive attrition rate for sexual assault cases is
approximately equivalent to other crimes, it is
apparent that when the unreported and the reported-but- dropped instances of sexual assaults are added up, the court system never acts upon between 91% and 97% of these assaults.
After case attrition is taken into account, about 96% of charges laid are for Sexual Assault I, 4% are for Sexual Assault II, and less than 1% are for Sexual Assault III. Roberts (1990) suggested that the fact that legal professionals predominantly classify sexual assaults at the lowest level may be due to more victims reporting less serious crimes. However, einother
worrisome possibility is that serious sexual assaults are being inappropriately forced into the lowest levels of sexualized assault. Thus, researchers need to
complainant in order to determine the appropriateness of classifying 96% of these assaults at the lowest level.
Conviction Rates
Given that so few victims of sexual assault report the crimes to the police, and so many of those they report are dropped, one might think that conviction would be likely for the select few remaining "best" cases. Yet Kinnon (1981) found that approximately 50% of rape cases resulted in the accused being convicted whereas the overall conviction rate for all crimes was 86%. Further, in his examination of provincial court data, Roberts (1994) found that between 25 and 30% of sexual assault offenders were convicted summarily.
Offenders convicted summarily receive about one quarter of the sentence given to those who are convicted of indictable offences.
Range of Sentence
Most offenders convicted of sexual assault were incarcerated (Roberts, 1990, 1994). Sixty percent of those convicted for Sexual Assault I received prison terms, 94% for Sexual Assault II, and 89% for Sexual Assault III. Twenty percent of Sexual Assault I, 25%
of Sexual Assault II, and 0% of Sexual Assault III offenders received suspended sentences (Roberts, 1994).
Sentences for sexual assault vary across Canada but are low relative to the serious nature of the crime. For example, in one study, the normal sentencing range for rape in the Quebec cuid
Newfoundland Courts of Appeal was three to seven years. Similarly, the Alberta Court of Appeal started at three years' imprisonment for serious sexual assault (Ruby, 1994). The actual sentence given would then go up or down depending upon mitigating or aggravating factors. Other courts have authorised lighter sentences. For example, the Newfoundland Court of Appeal sanctioned some sentences of under three years, and the Ontario Court of Appeal "affirmed that there may be cases of rape in which a suspended sentence would be
appropriate" (Ruby, 1994, p. 544).
The median sentences in Roberts' (1994) study of provincial courts in six jurisdictions were six months for Sexual Assault I, two years for Sexual Assault II, and five years for Sexual Assault III. These low
sentences can be partially accounted for by summary conviction, and the fact that more serious cases are
heard in the higher courts. Roberts (1994) did not provide any information on how the sentence related to the maximum that would have been possible if judges gave offenders consecutive sentences (usually judges impose concurrent sentences).
Sentencing Principles
Sentences vary with the specific details of each individual case. The specific details help the judge in deciding which of the sentencing principles to emphasise. The utilitarian principles of sentencing suggest that the judge should give a sentence that will deter others who would do similar crimes (general
deterrence) as well as deter the offender currently before the court (individual deterrence). The moral principles of sentencing highlight denouncement (which
is thought to clearly indicate right from wrong) and rehabilitation (see Ruby, 1994). It is thought that applying these general principles will result in the protection of society both from the offender currently before the courts and from those who would commit
similar crimes. Judges consider any mitigating or aggravating factors when reaching a sentencing
decision. Ruby (1994) described several of those including:
(1) Method of the crime, which includes such considerations as the offender's plamning auid deliberation. "Where the course of conduct is
extended, elaborate and calculated, rather than just a momentary loss of control, this may contribute to a longer sentence" (Ruby, 1994, p. 559). Judges would consider duration of the crime to be an aggravating factor when the offender committed the crime over a lengthy period of time. In general, judges assume that offenders who commit the crime over a long period of time have made "a conscious and deliberate choice to engage in criminality" (Ruby, 1994, p. 136).
(2) Magnitude and Impact of the Crime. In determining cui offender's sentence the judge will
consider the harm caused by the offender. "The number of victims may also be important" (Ruby, 1994, p. 138). That is, the more harm caused or the more people harmed by the offender will result in judges' viewing the case as having an aggravating factor.
(3) Breach of Trust. When a person abused the trust extended to them by others, judges usually
consider it to be an aggravating factor. In fact, "where a person in a position of trust breaches that trust, a term of imprisonment will normally be imposed" (Ruby, 1994, p. 148).
(4) Violence and the Use of Weapons. Ruby (1994) noted that "the courts differentiate between violence that is planned and that which is 'spontaneous'; but penalize both" (p. 160). He also noted that "the fact that an offender did not apply violence to a victim is not a factor in the mitigation of sentence, but simply the absence of an act that might aggravate the
sentence" (p. 161).
(5) Guilty Plea. "A guilty plea is evidence of remorse and, depending on the circumstance, will
therefore have some mitigating effect" (Ruby, 1994, p. 178) .
(6) Time spent in custody. In general, when an offender has spent time in jail awaiting trial, judges consider this time to be the "equivalent of a sentence of twice that length" (Ruby, 1994, p. 183). The judge
would then reduce the sentence given to the offender accordingly. (7) Alcohol and drugs. Ruby (1994) noted that
aggravating or a mitigating circumstance. In general, when the criminal act can be cast as "out of
character", judges will use intoxication as a
mitigating factor. An important exception to this pattern is for crimes of violence. Ruby (1994)
observed that "involvement in crimes of violence seems usually to render alcohol an aggravating factor, or neutral at best" (p. 192).
One of the functions of this dissertation is to examine sentences for sexual assault, including testing to see whether the factors mentioned above are
correlated with the length of the jail term given. If Ruby's (1994) observations hold true for sexual assault sentences, then I should find that: (a) as the period of time covered by the charge (duration) increases, the sentence should increase, (b) as the eunount of harm increases, the sentence should increase, (c) as the number of victims increase, the sentence should
increase, (d) breach of trust cases should receive higher sentences than similar cases not involving breach of trust, (e) cases where the offenders used alcohol or drugs should result in higher (not lower)
sentences, and (f) guilty pleas should result in lower sentences.
Summary
Canadian law-makers have been changing our sexual assault laws in attempts to offer justice to victims. With Bill C-49, the law now provides more protection to victims of sexual assault. The proper implementation of these laws is crucial to safeguard the rights of victims. Yet as already mentioned, earlier studies (Bavelas, et al., 1995; Coates et al., 1994; Gunn & Minch, 1988; Ruebsaat, 1985) have pointed to problems
in the judiciary implementing new laws. Also, Coates and Wade (1994) suggested that judges' psychological explanations of sexual assault may be working to reduce the offender's responsibility for the crime. If this is true, then the legal system will continue to fail the victims of sexual assault. This dissertation will, first, develop an analysis of causal attributions for assault and then test whether or how judges' causal attributions are connected to the actual sentence given.
CHAPTER 3:
ATTRIBUTIONS
"Why?" A common and important question is why someone did something. What we accept as the answer to this question will have implications for our response to another person's behaviour. For example, our
response to being told that another person broke a vase will likely be different if we are told that the person is "clumsy" rather than "careless". Psychologists call the process of connecting an event to causes
attribution. Generally, most researchers working in the area called social cognition assume that people make causal attributions in an attempt to understand an event, thereby making sense of the world.
Attribution As An Individual Perception Most researchers cast attributions as an intrapsychic process, that is, as an individual's
perceptions of causality. These studies treat language only "as part of method" (Edwards & Potter, 1993, p. 23). That is, they view language as a medium for examining a subject's causal thinking. Accordingly, because these researchers conceptualise language as a
neutral reflection of an individual's intrapsychic processes, they do not study language itself.
Moreover, traditional studies prevent subjects from engaging in "the common conversational practice of
choosing their own description of an event or disputing a given description" (Antaki, 1985, p. 214).
Research on attribution as individual perception has been a focus of study for social psychologists for decades. Consequently, I will not try to present a comprehensive review of this massive literature. What follows are a few highlights of traditional attribution theory and research.
Attribution Theories
Generally, many attribution theories assume that attribution is a three-stage process. In stage one, the perceiver identifies, categorises, and labels the behaviour (Schneider, 1991). Next, the perceiver
determines whether a trait (also called a disposition) has caused this behaviour. According to these
theories, perceivers use situational cues, behavioural cues, and prior expectations (based upon stereotypes or acquaintance with the actor of the behaviour) to decide upon a cause (Kelley & Michela, 1980; Ross & Fletcher,
