United Kingdom Journals

The transmission of disease during consensual sexual activity and the concept of associative autonomy

Janet Dine LLB, PhD, Barrister, *

Professor of Law, University of Essex
<dinej@essex.ac.uk>

and

Bob Watt, BA, BCL, *

Senior Lecturer in Law, University of Essex
<watt@essex.ac.uk>

* Earlier versions of this paper were discussed at the National AIDS Trust conference and the Essex University Centre for Theoretical Studies. Simon Davies and John Holmes provided assistance with research on the case law. Steve Anderman, Peter Luther and Peter Stone read earlier drafts. The authors would like to thank the above. The usual disclaimers apply.

Copyright © 1998 Janet Dine and Bob Watt.
First Published in Web Journal of Current Legal Issues in association with Blackstone Press Ltd.

Summary

The authors consider the Home Office Consultative Document on reforming the Offences Against the Person Act 1861 and the associated draft Bill. In their article which responds, in part, to an earlier article by Ormerod and Gunn, they argue that the establishment of a consensual sexual relationship radically modifies the familiar atomic concept of autonomy. The authors develop a number of established theories of criminal liability to advance a concept of associative autonomy. They apply this theory to the transmission of disease during consensual sexual intercourse and conclude that where disease is transmitted during a relationship criminal culpability should only apply in very narrow circumstances. Furthermore, they reject one of the mentes reae proposed in the draft Bill especially in the context of transmission of the HIV. On the other hand, in circumstances of widespread transmission of disease, they argue that the established law of public nuisance should be applied.

Contents

Introduction
The ambit of the Government's proposals
The basis of criminalisation
Autonomy and Harm

The concept of associative autonomy.

The Role of Consent in the Formulation of Offences

Stepping Outside Relational Autonomy

(1) Deception
(2) Outside the Limits of Consent

Existing jurisprudence concerning individual transmission.

Widespread Transmission
Public Health Law
Actual Transmission.
Risk

Conclusion

Bibliography

Introduction

The Government has issued a consultation document and draft Bill proposing reforms to the Offences Against the Person Act 1861. (Violence: Reforming the Offences Against the Person Act 1861, the Consultation document, henceforth CD), and the Offences Against the Person Bill, (the draft Bill, references to Clauses are to those in the draft Bill). The proposals are numerous and wide-ranging. This article is focused upon only one of the recommendations, which would impose criminal liability for the transmission of disease. It addresses disease caused by the human immunodeficiency virus (HIV), the aetiological agent of acquired immunodeficiency syndrome (AIDS). The transmission of this disease poses special problems in the criminalisation debate because it is most commonly passed during sexual intercourse⁽¹⁾ (see, e.g., Howe and Jensen 1997), which is usually a consensual activity, and because there are certain aspects of risk assessment which are unique to this disease. The matter of risk assessment is considered in detail below.⁽²⁾² Both of these factors must, it is argued, shape the debate about criminalisation.

The Government takes the view that (CD paras. 3.13 to 3.20 and Clause 15) the transmission of disease is a special case of assault because it is concerned not to discriminate against those who `are HIV positive, have AIDS or viral hepatitis or who carry any kind of disease' particularly if the effect of legislation is to `discourage people from coming forward for diagnostic tests and treatment in the interests of their own health and that of others' (CD para. 3.16). Accordingly it proposes that `the criminal law should only apply to those whom it can be proved beyond reasonable doubt have deliberately transmitted a disease intending to cause a serious illness' (CD para. 3.18). Two questions arise; (1) whether the proposals as they stand achieve this aim, and; (2) if they do, has the right balance been struck between the competing interests? In relation to the first question, it will be argued that the proposals are too wide ranging. In relation to the second question, it will be argued that the proposals avoid the central issue at stake, that of consent. It is striking that the documents contain no discussion or proposals relating to the place of consent in the criminal law. Indeed, it is proposed that the law should remain in its current confused state (CD para. 3.37 and Clause 18). Whilst the aim of simplifying the technical rules relating to violence must be welcomed, it is argued that the failure to tackle the issue of consent especially, but not exclusively, in the present context is a fundamental flaw. Whilst it is plain that the Government has not addressed the issue of consent since it anticipates that the Law Commission will make further proposals on this matter and accordingly our criticism may be viewed as somewhat harsh, we argue that the issue is of such importance that the Government should have delayed publishing the draft Bill until the paper on consent is published.

Furthermore, in the context of the transmission of HIV it is argued that a wholly separate justification for transmission may be established. This should be seen as a `public health' justification. Where a significant section of the public is affected by the transmission of disease, society has a right to defend its interests. The restrictions and sanctions imposed by the existing common law relating to public nuisance and the Public Health (Control of Disease) Act 1984 as amended by the Public Health (Infectious Diseases) Regulations 1988 (SI 1988/1546) will be considered in this context. It is argued that, if the Government wishes to provide a sound and comprehensive legislative basis for the criminalisation of the transmission of disease, the passage of some codifying measures consistent with those set out here will be necessary.

Top | Contents | Bibliography

The ambit of the Government's proposals

Transmission of disease is only relevant for the purposes of the offence defined in Clause 1(1) of the draft Bill which provides that a person is guilty of an offence if he intentionally causes serious injury to another. Transmission of disease is excluded from the definition of injury in relation to all offences save for that defined in Clause 1(1). In relation to Clause 1(1), Clause 15(4) provides that the definition of physical and mental injury includes `those things caused by disease'.

The definition of intent is two-fold and is set out in Clause 14 of the draft Bill (see, generally, Smith 1998):

`A person acts intentionally with respect to a result if - (a) it is his purpose to cause it, or (b) although it is not his purpose to cause it, he knows that it would occur in the ordinary course of events if he were to succeed in his purpose of causing some other result.'

It is argued herein that the mens rea represented by definition (a) should be the exclusive foundation of criminalisation under the Government's present proposals. In respect of the extended definition of intent set out in (b) it is argued that the definition is overbroad and should be rejected. It is acknowledged that our proposal will lead to different definitions of intent being used for different crimes, but this is already inherent in the government's proposals (see Smith 1998, in his discussion of Woollin, [1998] 3 WLR 382). ⁽³⁾ In the context of consensual sexual activity the partners together and mutually take the risk of some physical and psychological harms. It is argued that where one partner acts outside the ambit of the parties' mutual consent the transmission of disease should be criminalised. Accordingly, the issue of consent and the definition thereof is central to the criminalisation debate. It is accepted that the argument here set out would logically lead to the legalisation of euthanasia in a number of circumstances. The present authors welcome this conclusion. Furthermore, at least as far as HIV is concerned, there are special factors relating to the risk of transmission which render the definition as set out in the draft Bill unworkable.

Top | Contents | Bibliography

The basis of criminalisation

Criminalisation involves a serious interference with the liberties of the citizen and should not be undertaken without providing a proper theoretical justification. Public respect for the law and acceptance of its strictures depends upon an understanding of the reasons which underlie it (Williams 1995). The Law Commission does not accept that a coherent philosophical base was required to justify criminalisation in situations where consent is an issue. They proposed the adoption of rules on a `pragmatic' basis which `may lead us into what our critics may believe to be attitudes on related issues that are mutually inconsistent in a philosophical sense' (Law Commission CP No139, para 2.17)⁽⁴⁾. The current case law is certainly internally inconsistent and difficulties have arisen on both sides of the Atlantic. The cases reflect an attempt to adopt simultaneously an approach recognising the value of individual autonomy, permitting consent as a defence in some circumstances whilst imposing paternalistic limits to the ambit of consent (Law Com. CP No 139 para. 1.11). The most famous (or infamous) British decision is that in Brown [1994] AC 212 HL (see also Laskey, Jaggard and Brown v UK 109/1995/615/703-705) in which the majority of the House of Lords held that people were legally incapable of giving consent to activities which gave rise to serious physical harm, but both the House of Lords and the Law Commission accepted a number of exceptions to this rule, exemplified in the cases of Billinghurst [1978] Crim LR 553, Jones [1986] 83 Cr App R 375 and, more controversially, Wilson [1997] QB 47 and Slingsby [1995] Crim LR 570. Similarly in the USA the doctrine set out in the leading case of US v Outhier 45 MJ 326, USCAAF 1996 in which `one cannot consent to an act which is likely to produce grievous bodily harm or death' led to the decision in US v Bygrave 46 MJ 491, USCAAF 1997 where Bygrave, who was HIV+, was convicted of aggravated assault when he had unprotected intercourse with a woman who knew of his seropositivity when she consented to unprotected intercourse. The parties subsequently married each other.

It is here argued that it is possible to formulate a coherent basis for criminalisation based upon a conception of human beings as autonomous beings that can and do form relationships. When they form relationships, their individual autonomy is subsumed into a wider autonomy, which may conveniently be termed `associative'.

Top | Contents | Bibliography

Autonomy and Harm

Liberals, at least since Mill (Mill 1989, p.16), have to a substantial extent accepted the proposition that

"[t]he only purpose for which power can be rightfully exercised over any member of a civilised community, against his will, is to prevent harm to others. His own good, either physical or moral, is not a sufficient warrant. He cannot rightfully be compelled to do or forbear, because it will be better for him to do so, because it will make him happier, because, in the opinions of others, to do so would be wise, or even right."

By `flipping-over' this proposition one can quickly and easily conclude that the state can rightfully restrain a person from causing harm to others. The question then becomes `what is it that constitutes harm to others?'

Joel Feinberg (Feinberg 1984) defines harm to another in terms of a setback to a `welfare interest' and further defines `welfare interests' in a commonness way by reference to personal integrity in physical and psychological senses. The problem with Feinberg's approach is that it is unclear how one is to distinguish on a theoretical level that which relates to a welfare interest sufficient to warrant protection from that which does not. Joseph Raz (Raz 1988, p.414) makes this same point in a different way when he points out that the concept of harm is by itself merely formal or empty. "Since `causing harm' entails by its very meaning that the action is prima facie wrong, it is a normative concept acquiring its specific meaning from the moral theory within which it is embedded. Without such a connection to a moral theory the harm principle is a formal principle lacking specific concrete content and leading to no policy conclusions". Thus, `harm' needs to be defined by reference to some underlying moral theory.

Raz himself supplies the underlying moral theory by identifying people as having the capacity for autonomous action, thus people can formulate their own plans of life. A setback to a person's ability to formulate a worthwhile and coherent plan of life thus may amount to harm. Raz argues that the state should strive to provide the conditions for autonomy but argues that coercion to achieve that aim is rarely to be justified. Andrew Ashworth (Ashworth 1995, p.22 et seq.) takes this as a concession that coercion may be permissible in some circumstances and argues that `respect for personal autonomy means protecting individuals from being forced to act or debarred from acting in certain ways by the conduct of others and this may mean penalising those who act towards others in coercive ways.' This position is capable of two sorts of interpretation. Firstly, one could criminalise those who restrict the autonomy of others or, at least, make a credible attempt so to do. This is the objectivist approach. The touchstone of such a basis of liability is in the degree of harm suffered or apprehended by the victim; here the degree to which one's freedom of action is impaired by the criminal. The focus is upon the victim. This amounts to a refinement of Feinberg's setback to a welfare interest, a setback to the interest in autonomy. In the alternative, one might criminalise those who abuse their own status as an autonomous moral agent. Here the focus is upon the criminal. This is Alan Brudner's suggestion, advanced in the formulation of his `agency paradigm' (Brudner 1993). Here Brudner focuses upon the subjective mental state of those who behave in a morally reprehensible fashion by breaching the Kantian categorical imperative of treating another or himself as a means to an end rather than as an end in their own right. Actual resulting harm according to Brudner is irrelevant; thus Brudner would criminalise the person who (e.g.) shoots at, but misses a sleeping victim where the victim herself never discovers that she has been the target.

A number of authors have argued that the harmful consequences of a defendant's actions are entirely irrelevant since the true evil is contained in the defendant's wrongful intention to do harm (Gobert 1993; Ashworth 1987). Jim Gobert addresses the matter in the context of sentencing and argues that the degree of harm resulting from an assault is mere chance and should not found a distinction between different offences. It will be argued that this suggestion is particularly appropriate in cases involving the transmission of disease because the degree of harm is so difficult to measure and predict. Furthermore, because of the special nature of sexual activity, where consent and subjective intention are inevitably interlinked a refinement of Brudner's analysis seems preferable. This argument will be set out in detail below.

The notion of ascribing culpability according to the subjective mental state of the actor has been the subject of a number of criticisms. Anthony Duff provides the most powerful critique (Duff 1990, pp 149-157). However even Duff allows that the subjective view of an actor is important when ascribing criminal responsibility. Firstly, he argues that there are two ways in which we might understand subjectivism. He distinguishes between that which might well be called `fact subjectivism' and that which might be called `value subjectivism'. Actors hold a set of factual beliefs about the circumstances that surround an action in a given situation. Furthermore, actors make a set of value judgements surrounding acting in that situation. Having drawn this distinction Duff then argues that actors should be judged according to their subjective (or private) view of the facts, but according to the objective (or public) standards of morality embedded within the criminal law. Two points must now be made, firstly; Duff's account of subjectivism, his acceptance of fact subjectivism as exculpatory and his rejection of value subjectivism are consistent with the following analysis. Secondly, it is accepted that both Raz and Duff require that an underlying public moral theory is provided, and it is to this our attention must now turn.

Top | Contents | Bibliography

The concept of associative autonomy.

Brudner, Ashworth and Raz agree on one essential point. They primarily conceive of the individual as making a worthwhile or valuable life for him or herself within the framework of society. This ignores an important feature of human nature familiar even to casual observers and well recognised within the law (see Macneil 1983, on the theory of contracts). We argue that people may more properly be viewed as existing primarily within relationships. This relationship has its own special sort of autonomy, which we term `associative autonomy', the parties agree to come together to form a relationship bounded by its own consent in which some aspects of individuality are subsumed. The plan of life set out by this association is joint, it is neither a simple addition of two `individual' plans of life, neither is it a mere compromise.

Part of the core notion has been pithily set out by Douglas J in his discussion of marital privacy in Griswold v Connecticut 381 U.S. 479, 1965 at 486:

"We deal with a right of privacy older than the Bill of Rights - older than our political parties, older than our school system. Marriage is a coming together for better or for worse, hopefully enduring, and intimate to the degree of being sacred. It is an association that promotes a way of life, not causes; a harmony in living, not political faiths; a bilateral loyalty, not commercial or social projects."

Since Douglas J is dealing solely with marital privacy which leads him into a description of the marital relationship as one which `promotes a way of life rather than discrete causes'. The argument set out in this paper is that no consequences hang upon the fact that the relationship is long-term or emotionally close, relationships in which the parties operate within the bounds of associative autonomy can be short-lived. Douglas J then makes the important point that privacy attaches to the association. This point needs further elucidation, for the question is ... why? Lustgarten and Leigh (1994 p.40) in their book on national security and parliamentary democracy give us a paragraph that quickly moves the questioner to an obvious answer. On the ability to control information about oneself they write:

"More fundamental however, is the sense of mental and emotional security that this control entails. Imagine being unable to draw the curtain in your bedroom, so that others can see you naked at any time of their choosing. The fear and revulsion that this image evokes has little to do with the beauty or otherwise of one's body, but everything to do with one's sense of self. If I have no control about what is known about me, I am seriously diminished as a person both in my own eyes and in those which are capable of intruding upon me."

It should be noticed that Lustgarten and Leigh do not refer to the hostility or otherwise of the onlooker. They are correct in this. The onlooker could be actuated by some other motive; desire or concern immediately spring to mind. If lack of privacy destroys self, lack of privacy when afforded to a relationship destroys that relationship. Indeed one can see this doctrine operating, in a sense, in two directions. When one forms a relationship and thus when one approaches another person closely, one's own privacy disappears -- for the other person can see one more closely - in that sense one's self is diminished. From the opposing perspective when the parties of the relationship are joined in that relationship, an `associate-self' grows up around them and this must be afforded the protection of privacy if it is to survive.

Further insights into the core notion and its practical application in the courts may be obtained by a consideration of the implied or overriding term of the maintenance of trust and confidence in contracts of employment. Such relationships cannot be fully expressed in contractual terms because such relationships are dynamic. In these circumstances the courts have found little difficulty in distilling the essence of the relationship from the particular facts of each case. (See for example the judgment of Browne-Wilkinson V-C in Johnstone v Bloomsbury Health Authority [1992] 1 QB 333). The introduction of the concept of associative autonomy would require no more than a similar exercise, with the courts assessing the boundaries of the relationship in the case of each particular partnership. Respect for the autonomy of the resultant associative self explains why the state needs special justifications for interference in the privacy of consensual sexual relationships. An example of such a justification is where the health of the population is endangered by the activities within the relationship. This argument is set out below.

Consensual sexual activity is generally accepted to be of great personal and social benefit Support for this view may also be gleaned from the writings of a liberal in this matter. H.L.A. Hart (Hart 1962) wrote that sexual impulses form an integral part of each person's day-to-day life such that their suppression can affect `the development or balance of the individual's emotional life, happiness, and personality'. and it is one of the most significant reasons for forming associations. Its special nature should shape any attempt to criminalise an aspect of it. Perhaps more than any other activity it is one for which mutual responsibility ought to be acknowledged. Since individuals may only enter this mutuality by consent the current law recognises that individuals must be sufficiently mature to exercise their own judgement when consenting to intercourse, (ss 6&7 Sexual Offences Act 1956, Law Com CP no 139 Part V, and R v Fletcher (1859) Bell CC. 63) and must be fully able to consent. People also have a responsibility to ensure that, where there is a weaker party, the stronger at least genuinely believes that the other party has consented (see DPP v Morgan [1976] AC 182 where one party was physically restrained by others).⁽⁵⁾

Nicholas Bamforth (Bamforth, N., Law Com. CP No139, paras 10.42 - 10.51 and (1994)) argues further that

"Social meanings should normally be assessed from the standpoint of the participants in an activity, particularly within the field of sexual activity, given the social sensitivity surrounding the area and the sheer range of activities which possess sexual meanings for different people."

This suggests that the activities which take place solely within the relationship, in circumstances where the effects are strictly limited to the members of the association, ought to be absolutely privileged; whilst activities undertaken by the association which bear upon outsiders ought to be judged by the standards of the public law.

The position at the breakdown of the association must now be considered. Where one party to the relationship subverts the plan of the association, it is quite clear that the relationship has, in fact, ended. These circumstances may arise, for example, where there is a deceit as to HIV status. This is examined in detail below. Following the argument of Brudner set out above, it seems that an apparent partner who, in their own mind, abandons or subverts a joint plan commits the moral evil of using another person as a means to their own end.

Top | Contents | Bibliography

The Role of Consent in the Formulation of Offences

Moving from justification for criminalisation to the formulation of offences, it is necessary to consider the precise role of the `victim's' consent. In this respect the present authors depart from the approach adopted in the Law Commission's Consultative Paper no 139 where the defence of consent was regarded as separate from a defined offence. Its approach has been criticised most forcefully by Shute (1996) who emphasises the distinction between desired and undesired behaviour. His argument is that where society approves of the behaviour, consent should be viewed as integral to the definition of the offence, whilst only in a case of socially undesired behaviour should consent be viewed as a defence. Thus, sexual intercourse should not be prohibited per se with consent being a defence; rather lack of consent is an integral part of the offence of rape. The implications of this argument in relation to a general defence of consent will be examined later in this article.

Top | Contents | Bibliography

Stepping Outside Relational Autonomy

Two crimes may be formulated to reflect a situation where the defendant has stepped outside the boundaries of relational autonomy. These are (1) where the defendant deceives the victim by representing that he is HIV-, and (2)`deliberate infection' with the HIV.

(1) Deception

Normal principles of criminal law would require that any deception should be by a positive act unless there are considerations based upon the special vulnerability of the victim, which require the imposition of a duty to disclose on the defendant. Apart from the fact that imposition of positive duties by the criminal law has always caused difficulties, For example the parameters of the positive duty to care in R v Stone and Dobinson [1977] 1 QB 354 have always remained obscure. There are particular reasons for not imposing such a duty in these circumstances. In particular health professionals' perceive that the best way to control the spread of the virus is to encourage anyone who is at risk to test for the virus. A general duty to disclose might well be counterproductive in this respect. Ian Weller (Weller 1997) refers to the significant number of gay men who refuse to be tested and whose first contact with medical care is when they arrive at hospital casualty departments with Pneumocystis carinii pneumonia (see also CD, para 3.16). In the absence of deceit or disclosure the partners together take the risk of transmission, it would be interference with the inherent notion of associative autonomy to prevent them from so doing.

Ormerod and Gunn (1996b) argue both that a change in the nature of the act should be grounds for negating consent and that (1996a) the criminalisation of the transmission of AIDS during consensual sexual intercourse is justified where "one person is deliberately withholding information from the other with a desire to infect that other." The Law Commission's second Consultation Paper (Law Com No 139, paras 6.8 et seq.) considers whether there should be a difference between deceiving a person into the belief that a potential partner is not HIV+, exploiting a mistaken belief (such as information given by a friend), and the situation which would arise if there were a positive legal duty upon a person living with HIV/AIDS to correct mistakes and disclose all relevant facts. The Commission correctly notes that it would, in the latter circumstances, `be necessary to define exactly what must be disclosed' and then continue by asking, [w]ould it make a difference if intercourse were not unprotected?'(Law Com CP No 139, paras 6.30 & 6.31).

In its earlier paper the Law Commission had proposed that the `test should be, and should be no more than, whether the fraud or misrepresentation induced the victim's consent'. (Law Com CP No 134, paras 25.1 & 25.2). The responses to this proposal caused the Law Commission to resile and return to the position at common law where the victim must be deceived as to the nature of the act or the identity of the person involved in the act (Law Com CP No 139, para 6.18).

The case law has traditionally sought to define the nature of the act to which consent has been given. In R v Clarence (1888) 22 QBD 23 consent to marital intercourse was the issue. During intercourse gonorrhoea was transmitted to the wife. The court held that no assault had taken place; the wife consented to sexual intercourse as part of the marriage contract and the fact of the husband's infection did not change the nature of the act so as to negate that consent. Smith (1998) argues that "it seems reasonably clear that Clarence was wrongly decided. (1998 at p319.) While it is clear that the parameters of joint consent involve a consideration of the acts to which express or implied consent has been given, it may be argued that a focus on the objective "nature" of such actions in the absence of an understanding of the relationship causes a distortion of vision. Laying down particular categories of facts which "change the nature of the act" takes the focus away from the complex nature of individual relationships. But for a case when the distinction was used to good effect see R v Richardson (1998) CA, judgment of 25^th March 1998. A dentist continued to practice after being removed from the dental register. The Court of Appeal quashed convictions for assault on the ground that her patients were not deceived as to her identity or the nature or quality of the acts she performed. It is suggested that the same result would follow from investigating the parameters of the consent between the defendant and the victims in this case. It is of course possible to argue that the nature or quality of the act changes when the partner is infected with the HIV. However, it is submitted, this is the wrong way to address the issue. The concern should not be to define the nature of the act to which consent was given but the parameters of the joint consent. In the absence of deceit, the assumption would be that the risk of transmission is borne jointly, it comes within the relational autonomy of the parties. It is the deliberate denial of the victim's choice of giving or withholding an informed consent which is the mischief to be condemned, not the sexual act.

Reference must also be made to the very recent case of R v Cuerrier in the Canadian Supreme Court (unreported, September 3^rd 1998, file No.25738.). The accused was charged with two counts of aggravated assault because he had unprotected sexual intercourse with two women without informing them that he was HIV+. Both complainants had consented to unprotected sexual intercourse with him, but they testified at his trial that they would not have engaged in unprotected intercourse had they known of his seropositivity. The Supreme Court held unanimously that Cuerrier had defrauded the women and this fraud vitiated their consent. A majority (Cory, Major, Bastarache and Binnie JJ held that it was no longer necessary to consider whether the fraud related to the "nature and quality of the act" but instead whether the dishonesty was related to the obtaining of consent. They said that the dishonest act consisted of either deliberate deceit respecting HIV status or non-disclosure of that status. While it is to be welcomed that the Canadian Supreme court disengaged the enquiry from the nature and quality of the act it is suggested that their formula merely broadens the definition of facts which negate consent. They have failed to address the real issue of the true parameters of consent. If the women had testified that they would have consented even knowing that the defendant was HIV+, would this have made a difference?

The core issue is the parameters of associative consent. What problems does that raise in these circumstances? Clearly when giving consent to intercourse a party need not be informed about everything to do with the perpetrator. The real issue in legislating on this subject would be whether to identify AIDS as a single issue and treat it wholly separately from all others or to enter the potential minefield suggested by the Law Commission in its first consultation paper (and see Law Com CP No 139, paras 6.15-6.19) by legislating for a general offence of obtaining consent to intercourse by deception. It would clearly be impossible to require the defendant to give a full and frank account of his whole circumstances before consent was valid even if one aspect of his character or circumstances might make him irredeemably repugnant to the victim. It is therefore here proposed that a specific offence should be created, requiring a specific positive deception as to the defendant's HIV status.

Top | Contents | Bibliography

(2) Outside the Limits of Consent

A further crime may be justified where consent is given to sexual activity but where the defendant, unknown to the sexual partner, harbours a desire to transmit the virus. It is arguable that the transmitter in this situation also treats the victim as a mere thing since the consent is no longer properly informed. It is not necessary to argue that the act changes in nature, R v Clarence (1888) 22 QBD 23, but the only true consent would be one by which the victim accepted that the defendant had a desire to infect. In this rare but not unknown circumstance (see the facts of US v Bygrave 46 MJ 491 USCAAF 1997) it can be argued that the autonomy of the victim has not been compromised. No offence would be committed, since the justification for criminalisation has fallen away. Note that this crime would focus on a true subjective desire to transmit the virus, risk taking would not be sufficient. It is argued that consensual risk taking does not involve one of the parties treating the other as a means to their own end. Thus, the notion that the defendant knew that there was a high risk of transmission would be insufficient to found culpability. It can be argued that reliance on anything but an intention to harm outside the parameters of joint consent is an inappropriate mens rea where the transmission of disease is involved. There are two reasons for this. The first is purely technical and relates to the difficulty of risk assessment. In R v Nedrick [1986] 1 WLR 1025 it was made plain by Lord Lane CJ that the mens rea of intent could be inferred by a jury when the defendant knew that death or really serious injury would come about as a `virtual certainty' of the act contemplated and done. The House of Lords held in R v Woollin [1998] 3 WLR 382 that use by the trial judge of `substantial risk' rather than `virtual certainty' was not acceptable.<sup>(6)</sup> It is by no means clear that transmission of the virus is a `virtual certainty' especially where the person living with HIV/AIDS has a low viral load. People with AIDS who receive the new combination therapies (protease inhibitors, nucleotide analogues and non-nucleotide reverse transcriptase inhibitors) frequently do not exhibit detectable levels of HIV nucleic acid in their white blood cells. This does not mean that they are free of the virus, but that they have as few as twenty copies of the viral RNA per millilitre of blood. Recent reports indicate that some treatments may be so effective that people living with HIV/AIDS may be able to refrain from therapy for periods of up to one year without experiencing any rise in viral load and with an apparent recovery in white blood cell levels (Villa et al., 1997)⁽⁷⁾. Typically, people with `full-blown AIDS' may have some millions of virions per millilitre of blood. If a person knows that they do not have detectable virus in their bloodstream, it seems improbable that they subjectively `expose' a person to the HIV by having unprotected intercourse with them. However, public prejudice may well say that once a person has been tested as seropositive for the virus they should remain permanently on their guard. Medical opinion may remain uncertain on this point for some considerable time. Neither is it `virtually certain' that `death or other serious injury will result' even if the virus is transmitted. Weller (1997 at p 10) points out that, even `[w]ithout any drugs, 50% of people will still be perfectly well ten years after they become infected.' The inference of intent in circumstances where the defendant `knows' of a high risk may therefore be said to impose an impossible standard for it is, at the same time, both `too high' and `too low'. It is too high because it can never be `virtually certain that death or really serious injury will result' from the transmission of the virus, and too low because it includes the notion of consensual risk taking.

Top | Contents | Bibliography

Existing jurisprudence concerning individual transmission.

How have courts treated the transmission of AIDS in the situations similar to those posited above? Here it is instructive to look at the United States where the debate has focused partly on the possibility of charges of attempted murder being successful and partly on specific laws of reckless endangerment. There is however some overlap and incoherence in the approaches. The case of State v Hinkhouse 912 P. 2d 921, Mod. 915 P.2d 489, Or. Ct App. 1996 is particularly informative. It raises issues of the appropriate mens rea for crimes of individual transmission. The court appears to accept that the highly subjective mens rea, which we have suggested is appropriate for criminalisation in these circumstances. Hinkhouse repeatedly had unprotected sex with a small number of women. The facts show that he was fully aware of his HIV seropositivity and its implications for his partners. On at least one occasion he promised to use a condom during intercourse but failed to do so. Whilst he knew that he was HIV+, his partners were unaware of his status. If they asked he would lie, otherwise he would say nothing about his HIV status. In denying that he was HIV+ he said that if he had been HIV+ he would try to spread the disease. Some of the women with whom he had intercourse became infected with the HIV. Hinkhouse also had protected intercourse with a woman whom he hoped to marry. He made her aware of his HIV status and always used a condom during intercourse with her. Hinkhouse had told his probation officer that he was HIV+. He acknowledged that the disease could be spread through a single unprotected sexual encounter, and that to spread the disease would be `murder'. At one time he signed an agreement drawn up by the probation service to the effect that he would not engage in any unsupervised contact with women without the express permission of his parole officer. Hinkhouse was convicted of attempted murder in relation to the unprotected intercourse. A person is guilty under the law of Oregon of attempting to commit a crime when he `intentionally engages in conduct which constitutes a substantial step toward commission of the crime'. The Oregon law further provides that to act `intentionally' is to `act with a conscious objective to cause the result or to engage in the conduct so described'. Furthermore a person `commits attempted murder when he or she attempts, without justification or excuse, intentionally to cause the death of another human being'.

The Court of Appeals of Oregon affirmed Hinkhouse's conviction holding that he had shown by his persistent exploitative conduct that he had not acted on impulse to satisfy his sexual desires, but instead acted deliberately to cause his victims serious bodily injury and death. The present authors argue that the reason Hinkhouse should be criminalised is because he has abused and broken the associativel autonomy he and his partners set up.

This case is usefully compared with that of Smallwood v State 680 A. 2d 512, Md. 1996. Smallwood was an HIV+ robber and rapist and thus consent was not an issue. He knew of his seropositivity and had been counselled to practice safe sex. He told health care workers that he had only one sexual partner and that he always used a condom. He raped and otherwise sexually assaulted three women without taking any precautions against the transmission of the virus. He successfully appealed against his conviction for attempted murder. The Court of Appeals of Maryland held that Smallwood lacked the specific intention to murder his victims. Hinkhouse was cited before the court and was distinguished on the grounds that the court had held that, as a matter of fact, Hinkhouse had provided ample evidence of his subjective intention to kill or seriously harm his victims. This was lacking in Smallwood's case, the court holding that he had only the mentes reae sufficient to rob his victims and rape them. The court held that for Smallwood to be guilty of attempted murder he would have had to show by words or action that he intended to kill or cause really serious injury. The court in Smallwood reviewed the authorities binding upon them and in addition a number of cases from other state jurisdictions. In the judgment the court highlights a number of decisions. In Hinkhouse words of specific intent were used in that he said that he wished to pass the infection. In State v Caine 661 So. 2d 1358, La .1995, the defendant jabbed a used syringe into a victim's arm whilst shouting `I'll give you AIDS' and was convicted of attempted murder. In Weeks v State 834 S.W. 2d 559, Tex. App. 1993 the HIV+ defendant spat at a prison guard in the belief that he could transmit the virus in that way. He said that he `was going to take someone with him when he went'. Similar results were reached in State v Haines 545 N.E. 2d 834, Ind. App. 1989 and Scroggins v State 198 Ga. App. 29, 401 S.E.2d 13, 1990 Ga.App. where the defendants respectively sprayed their victims with HIV infected blood and bit them. All these facts were held sufficient to found convictions for attempted murder. In this group of cases, including Hinkhouse and Smallwood, the US courts are taking an uncompromisingly subjectivist approach to the question of attempted murder. Did the defendant evince a desire, by his words or actions, to kill or do really serious harm to his victims? The emphasis in these cases is clearly upon intention as desire regardless of the consequences. The Weeks case seems to show that even where the means employed was wholly inadequate to achieve the result the court is willing to convict. So far as is known no-one has ever been infected with the HIV by being the victim of a spitting attack, and it is generally believed by the medical community that it is highly unlikely that the virus could be passed in that way.

Conversely, as shown in Smallwood, if the defendant did not independently show that they desired to kill or inflict really serious harm the court would not convict of attempted murder. Whether or not actual physical harm was achieved is not seen as relevant in these cases. The insistence on subjective mens rea as alone sufficient is, as explained above, at odds with the traditional UK analysis of intent which includes the ability of the jury to infer intent from knowledge by the defendant that his behaviour involves a very high risk of harm.

Smallwood is easy to understand if the basis for the decision was desire alone as sufficient mens rea. However, the court seemed to "hedge its bets" by undertaking their own estimation of the level of hazard posed by the administration of semen to Smallwood's victims. Whilst the court affirmed their earlier statement in the case of State v Raines 326 Md. 582 (1992). that `an intent to kill may be inferred from the use of a deadly weapon directed at a vital part of the human body', they denied that a single act of intercourse, (or even, presumably, two exposures to semen) was sufficient evidence on which to found an inference of an intent to kill. This was because there was insufficient evidence on which to conclude that death would or could result from such an exposure. If this risk analysis was an investigation into actual risk in order to discover if there was a desire to kill it is explicable and consistent with an analysis, which adopts direct intention as the only applicable mens rea. If, however, it is a reassessment which would lead to the inclusion of recklessness as defined in R v Caldwell [1982] AC 341 as a sufficient mens rea, it appears to be inconsistent with the line of cases discussed above and the discussion earlier in the judgment.

Top | Contents | Bibliography

Widespread Transmission

It is clear that there is a second, social, evil that needs to be addressed. This concerns the widespread transmission of the virus. The draft Bill makes no distinction between single and widespread transmissions. However, the evil of transmission of the virus, which harms a number of recipients, places the crime of transmission into the public domain. Brudner who sets out a second paradigm for the ascription of culpability, which he names `the welfare paradigm', again affords some guidance. It is based upon the arguments put forward by J.S. Mill (Mill 1989) and propounded most articulately by Joel Feinberg (Feinberg 1984). The core proposition is that actions should only be criminalised where the consequence of the action is harm, either physical or psychic, to identifiable others or to society as a whole. This restates the harm principle together with the derivative offense principle developed by Feinberg, (Feinberg 1984, p26). The validity of the offense principle is challenged by Shute (Shute 1996, p.691). Brudner (Brudner 1994) attaches culpability under this paradigm to regulatory crimes - mala prohibita. According to Brudner, the appropriate mens rea for such crimes may be objective recklessness as defined in R v Caldwell [1982] AC 341 or even strict liability. Here it may be seen that the consequences of the action are essential to the determination of culpability as the emphasis is upon the harm caused.

There are two alternatives; either some harm must directly result from the action and so a penalty must be applied as a corrective or, the risk of harm consequent upon the action must be so great that a criminal penalty is appropriate as a deterrent to future action.

The justification for introducing an offence of this sort for transmitting the HIV to a wide section of the public is that society is entitled to protect itself against widespread damage as a matter of public health. This view also accords with Ashworth's view that society is entitled to act coercively where it acts to protect the lives of its citizens. In the case of widespread transmission it can therefore be argued that the consent of the parties to the intercourse that spread the infection is legally irrelevant because the concern to protect public health outweighs the autonomy of the people involved. Thus, transmission or undertaking a significant risk of transmission should be criminalised where it can be shown that a person has passed the virus on to a section of the public. The definition of `section of the public' is discussed below, and it is suggested that the limits to such a group are well discussed in R v Johnson [1997] 1 WLR 367 and R v Madden [1975] 3 All ER 155. It must be noted that the legal test proposed is based upon the fact or risk of transmission of the virus rather than upon the actual development of HIV disease. In accordance with Brudner's welfare paradigm, any such offence should have a very low mens rea requirement.

Top | Contents | Bibliography

Public Health Law

This suggested criminalisation is in line with the way in which public health law is used for dual purposes: firstly to prevent or penalise the actual transmission of the disease, secondly to deter risk taking or penalise those who take the risk of transmission. In the first case, the issues of causation are essential to the crime, in the second causation is legally irrelevant.

Actual Transmission.

In the UK there is currently no specific offence of transmitting the HIV. Ormerod and Gunn(1996b) consider the possible application of ss.23-24 OAPA 1861 in this context. It is thought doubtful whether these sections are applicable to the transmission of a living aetiological agent or of an infection as opposed to an intoxication. Do any existing offences reflect the view that transmission of the virus should be criminalised where one perpetrator infects a large number of victims?

The common law offence that seems to fit most closely is that of public nuisance. This has been used to convict those guilty of spreading human (R v Vantandillo (1815) 4 M&S 73) or animal disease (R v Henson (1852) Dears CC 24). It was used to convict a surgeon who was suffering from hepatitis B when he passed the disease to a number of his patients (see R v Gaud unreported, Southwark Crown Court 21st July 1994; cf. Ingley, determination recorded in the GMC Professional Conduct Committee minutes, 1st-3rd June 1998). Ingley dripped blood into an individual patient whose hip he was replacing. The single patient died of hepatitis B transmitted by Ingley, no other patient was affected. Ingley was not tried for any criminal offence. Thus, it may be that widespread transmission of the HIV may already be a crime of public nuisance at common law.

Some dicta support the proposition that the offences dealing with public welfare and public nuisance are offences of strict liability. The clearest examples of such statements are to be found in Sweet v Parsley [1970] AC 132, the judgment of Wright J in Sherras v De Rutzen [1895] 1 QB 918, and Lord Scarman's statement of the advice of the Privy Council in Gammon Ltd v A-G for Hong Kong [1895] 1 AC 1 . It is argued that these remarks are obiter because the offences in question were statutory and quite different in nature from those under consideration in this article. However, in the prosecution of Shorrock [1994] Q.B. 279 (CA, per Rattee J at p 289). for causing a public nuisance there is authority for the proposition that mere knowledge that one's act or its natural consequence will give rise to a `real risk' of public nuisance is sufficient to found culpability. The degree of knowledge required does not seem to be `actual knowledge', it seems sufficient that somebody knew or `ought to have known, in the sense that the means of knowledge were available to D'. The facts of R v Shorrock were that a farmer was convicted of causing a public nuisance when he hired a field to the organisers of an `acid house' pay party who concealed their real intention to hold the party by means of a thin lie. This shows that the mens rea requirement for such an offence is very low. Whilst the mens rea requirement is easily satisfied, there is an absolute condition that harm actually resulted.

The offence of public nuisance involves some degree of harm to the community as a whole. It has been used to convict persons who have made a large number of obscene telephone calls. R v Johnson [1997] 1 WLR 367 and the cases reviewed therein by the Court of Appeal suggest that the essence of the crime is that the nuisance caused is such that it is so widespread in its range or so indiscriminate in its effect that its abatement is in the interest of the community at large and therefore a criminal sanction is justified.⁽⁸⁾ Such a rule caused the court to acquit Madden [1975] 3 All ER 155 whose telephoned bomb hoax only affected eight security guards. In these circumstances a single transmission of the HIV does not amount to a public nuisance. There are three reported cases of people having sex with large numbers of partners whilst HIV positive: Stephen Thomas in Finland (with over 100 partners), Giuseppina Barbieri in Italy (with over 5,000 partners) and Darnell McGee in the USA (over 100 partners and passing the virus to at least 18). and it is argued that a conviction for the crime of public nuisance would certainly be appropriate in these cases.

Top | Contents | Bibliography

Risk

Exposing other persons to a risk of catching one of a number of diseases, set out in s.10 Public Health (Control of Disease) Act 1984 as amended by the Public Health (Infectious Diseases) Regulations 1988 (SI 1988/1546), is an offence by virtue of s.17 of the 1984 Act. This offence is set out in s.17(1) of the 1984 Act which provides that exposing another person in a public place to the risk of catching the disease is an offence where someone acts `knowing that he is suffering from a notifiable disease'. AIDS is not a notifiable disease in the context of this section, although it is in other sections of the Act. Could this offence fit within the `public health justification' for criminalisation and therefore be extended to include the HIV?

The answer appears to be in the negative. First, the element of `public' is very easily satisfied for the statutory definition of public place is very wide. For example, whilst an act of intercourse in a person's own bedroom would not be criminalised, the same act of intercourse in a hotel bedroom would be culpable. Furthermore the Act does not seek to criminalise the widespread risk of contagion, but would criminalise a single risky encounter. This does not sound like a measure designed to secure public health.

Further, there is a requirement for mens rea as the statute prohibits knowingly creating the risk of transfer of the virus. This seems to equate to at least Cunningham or subjective recklessness, although some writers (Closen and Deutschman 1990 and McColgin & Hey 1997, p.293) suggest that the mens rea requirement is very much lower. Closen and Deutschman believe that the relevant test is objective recklessness, the standard being whether even unreasonable people would perceive a risk. However, such an analysis soon entangles the mens rea `knowing' with the actus reus `expose'. If the mens rea is to contain some element of 'knowing', it may be that some people would avoid knowingly risking transmission not by avoiding risky behaviour but by avoiding learning of their seropositivity. This is precisely the concern addressed in Law Com CP No 139, at paragraph 3.16. Thus an extension of the law using this formula may, in fact, set back the improvement of public health.

Thirdly there need be no actual transmission - it is the risky act that is criminalised. The problem here is that too often the mental state for the risk taking - D desired to take a risk, D knew that he was taking a risk with a high likelihood of harm and so forth - becomes entangled with the knowledge of the amplitude of the actual risk taken.⁽⁹⁾

There are two ways in which the risk could be measured, in a subjective sense, which is to say, by reference to what D had in his own mind; and in an 'objective' sense, by reference to expert medical opinion or by reference to what the majority of people are taken to have in their minds. The confusion between these two ways of measuring risk creates the worst of all possible worlds and indicates the real difficulty met in public health statutes which is with the interpretation of the word `expose', which delineates the actus reus of a typical offence.

The best available source of guidance is from the various states of the USA, a number of which have specific HIV transmission statutes. The Washington state law (no. 9a. 36. 021.) criminalises, inter alia, those who `expose' others to the HIV. The term `expose' was alleged to be unconstitutionally vague in State v Stark, 832 P. 2d 109 Wash. Ct. App. 1992, but the court held that "any reasonably intelligent person would understand from reading the statute that the term ... refers to engaging in conduct that can cause another person to become infected with the virus". Whilst it is clear that Stark engaged in repeated unprotected intercourse with three different women after he had tested HIV+ and been counselled to practice safe sex, and was thus upon the facts, engaged in risky behaviour, the use of the phrase `can cause another person to become infected' is problematic. Many sorts of sexual behaviour can cause infection; for condoms may rupture during penetration; semen or other body fluids may contact broken skin or other permeable membranes during masturbation or oral intercourse despite the fact that reasonable precautions have been taken. Furthermore people with AIDS may have very low titres of virus in many body fluids, so any conduct which brings others into contact with any body fluid may amount to `exposure'. Certainly it has led some commentators to suggest that language of this sort may `make it a felony for a person infected with HIV to shake hands if his palms are sweaty, or to swim in a public swimming pool' (Closen & Deutschman 1990). There do not seem to have been any prosecutions for mere knowing exposure of this sort in the USA and it is suggested that Closen's analysis is unlikely to be followed in the British or most US courts. Those cases which have been prosecuted have been those where the defendant has deliberately set out to pass on the virus but has chosen an inadequate means for so doing. These clearly fall into the category of deliberate transmission considered above. It would constitute an attempt to commit such a crime.

It is therefore suggested that for widespread transmission the public nuisance approach is superior to adopting the current British statute which criminalises risk or exposure to risk.

Top | Contents | Bibliography

Conclusion

The argument in this paper is that where relationships are concerned, autonomy has to be viewed in a radically different light. Relationships take on a life of their own creating a space bounded by mutual consent, which prima facie should be respected, by the state.

While we have largely confined ourselves to the specific situation of criminalisation of the transmission of HIV by consensual sexual behaviour, we believe that the approach adopted contains elements which could and should inform the general approach of the criminal law to issues of consensual behaviour which arise within relationships. The enquiry should start from the relationship and determine the boundary between the ambit of the mutual consent of the parties and any public impact it may have.

It may well be argued that evidential safeguards (especially where euthanasia is an issue) should be required. No-one doubts that the establishing the relevant facts defining the boundaries of a relationship will be difficult, but the authors argue that a coherent approach to private consensual behaviour can be built on the foundations explored in this essay.

Top | Contents

Bibliography

Ashworth, A (1987) `Belief, Intent and Criminal Liability' in Oxford Essays in Jurisprudence (3^rd series) (Oxford: Oxford University Press).

Ashworth, A (1995) Principles of Criminal Law, 2^nd ed. (Oxford: Oxford University Press).

Bamforth, N (1994) `Sado-Masochism and Consent' Criminal Law Review 661.

Brudner, A (1993) `Agency and Welfare in the Penal Law' in Shute, Gardner and Horder (eds.) Action and Value in Criminal Law (Oxford: Oxford University Press).

Closen, MC & Deutschman, JS (1990) `A proposal to repeal the Illinois HIV Transmission Statute 78 Illinois Bar Journal 592.

Duff, RA (1990) Intention, Agency & Criminal Liability (Oxford: Basil Blackwell).

Feinberg, J (1984) The Moral Limits of the Criminal Law, Vol. 1, Harm to Others (New York: Oxford University Press).

Gobert, J (1993) `The Fortuity of Consequences' [1993] Criminal Law Forum 1.

Hart, HLA (1962) Law, Liberty, and Morality (Oxford: Oxford University Press).

Howe, JM and Jensen, PC (1997) `An introduction to the medical aspects of HIV disease' in Webber, DW (ed.) Aids and the Law (3rd ed.) (New York: Wiley).

Law Commission CP No. 134 (1994) Consent and Offences against the Person (London: HMSO).

Law Commission CP No. 139 (1995) Consent in the Criminal Law (London: HMSO).

Lustgarten, L and Leigh, I (1994) In from the Cold (Oxford: Oxford University Press).

Macneil, I (1983) `Values in Contract: Internal and External' 78 Northwestern University Law Review 340-418.

McColgin, DL and Hey, ET (1997) `Criminal Law' in Webber, DW (ed.) Aids and the Law (3rd ed.) (New York: Wiley).

Mill, JS (1989), On Liberty (New York: Prometheus).

Ormerod, DC and Gunn, MJ (1996a) `The second Law Commission Consultation Paper on Consent: a second bash' Criminal Law Review 694.

Ormerod, DC and Gunn, MJ (1996b) `Criminal Liability for the Transmission of HIV' [1996] 1 Web Journal of Current Legal Issues

Raz, J (1988) The Morality of Freedom (Oxford: Oxford University Press).

Roberts, P (1995) Appendix C to the Law Commission Consultation Paper No 139, Consent in the Criminal Law (London: HMSO).

Shute, S (1996) `The Second Law Commission Consultation Paper: (1) Something Old, Something New, Something Borrowed: Three Aspects of the Project' Criminal Law Review 684.

Smith, JC and Hogan, B (1996) Criminal Law, 8^th ed. (London: Butterworths).

Smith, JC (1998) `Offences Against the Person: The Home Office Consultation Paper' Criminal Law Review 317.

Villa, CP et al (1997) `Absence of Viral Rebound after treatment of HIV-infected patients with didanosine and hydroxycarbamide' 350 Lancet 635-6.

Watt, B (1998) `The legal protection of HIV+ Health Care workers and the Human Rights Jurisprudence of the European Court of Justice' European Human Rights Law Review 301-312.

Weller, I (1997) 56/57 NAM AIDS treatment update (August/September 1997)

Williams, B (1995) `What has philosophy to learn?' in Owen, D (ed.) Philosophical Foundations of Tort Law (Oxford: Oxford University Press).

Footnotes

(1) We use the term sexual intercourse to encompass the range of female with male, male with male and female with female activities in which people engage. It is acknowledged that female with female sex is associated with a very low incidence of transmission and male with male sex with a high incidence.

(2) The difficulties of assessing the degrees of risk, encompassing the inherent uncertainty of transmitting the virus, and the incubation period of HIV related disease and the uncertainty of disease progression are included in the discussion set out below.

(3) Smith makes the point that as it stands the extended definition proposed by the government does not reflect the definition of intent in R v Woollin since the defendant was convicted of murder although he did not intend any purpose other than to vent his anger.

(4) A number of writers have passed unfavourable comment upon this approach. They do not think that the boundaries of criminality can be defined without first reaching an understanding of the nature of consent (Roberts 1995; Shute 1996; Ormerod and Gunn 1996a)

(5) See, furthermore, the discussion of "reckless rape" in Law Com CP No139 para 7.18, "a man who engages in sexual intercourse must take care that any beliefs he forms about consent are accurate" and Shute (1996).

(6) See criticism of the `virtual certainty' standard in Smith and Hogan 1996, at p.58.

(7) A fuller discussion which may be more comprehensible to lawyers is to be found in Watt, B 1998.

(8) See also Attorney General v PYA Quarries Ltd. [1957] 2 QB 169.

(9) The use of recklessness (in the Cunningham sense) in these circumstances merely raises the same complication in the guise of asking whether the risk was unjustifiable?