A Probative Argument of Intention

Introduction

This article develops the basic argument for proving intention in criminal adjudication within common law and non-common law jurisdictions around the world. The evidential proof of an accused's state of mind is a very old problem that arises out of an even older moral quandary, namely: whether, and in what ways, should we distinguish between unlawful killings that are accidental from those that are intentional. ¹ For as long as human beings have been killing one another, ² the moral inclination to punish intentional unlawful killings is much harsher than those that are perpetrated accidentally has endured. ³ For example, the Code of Hammurabi (circa. 2270 B.C.) ⁴ punished intentional killings with death, while compensation was payable for unintentional killings. ⁵ Similarly, the common law distinction between murder, which was punishable by death, and manslaughter, for which compensation was payable, ⁶ is traceable back to the sixteenth century. ⁷ However, the manner in which intention has been proven throughout the long history of criminal adjudication has varied. For example, at a time when it was widely believed that ‘[t]he thought of man shall not be tried, for the devil himself knoweth not the thought of man,’ ⁸ and that ‘secret things belong to God,’ ⁹ medieval European jurists believed that intention was provable only through the alleged killer's confession. ¹⁰ Moreover, premodern indigenous African fact-finders relied on expert marabout (diviners or traditional healers) to reveal the opponent's ¹¹ intention. ¹² In the contemporary era, the dominant mode of proof, particularly among common law jurisdictions, is characterised as the ‘Rationalist Tradition’, in terms of which the proof of all facts in dispute (facta probanda), including intention, is undertaken through logical argumentation. ¹³ Led principally by Jeremy Bentham and John Henry Wigmore, and with strong connections to the 18th century European Enlightenment ideals, ¹⁴ fact-finders within this tradition hold the view that intention is provable through inferential reasoning. ¹⁵ According to Williamson JA, the inferential proof of intention within this tradition requires that: ‘a trier of fact should try mentally to project himself into the position of that [opponent] at that time. He must of course also be on his guard against the insidious subconscious influence of ex post facto knowledge’. ¹⁶

The proof of intention is complicated at two dimensions within the contemporary rationalist era. Firstly, the necessity of proving intention arises from an opponent's denial of this particular factum probandum. The conventional denial of intention, namely: ‘I did not mean to do it’, is a symptom of a constraint in the opponent's freedom to make the right choice of action. ¹⁷ This bare denial is also known as a ‘failure of proof defence’ ¹⁸ (or an ‘evidential defence’) in that the opponent denies having acted intentionally, without advancing a particular affirmative defence or excuse. At this level, the probative argument reflects a linear one-to-one relationship as shown in Figure 1.

OPEN IN VIEWER

Figure 1.

Linear one-to-one probative argument.

This basic model is complicated by two prima facie contrasting ideas. On the one hand, the intention is commonly understood to be ‘subjective’, that is, as having to do with an opponent's ‘state of mind’ ¹⁹ or as an inner ‘mental element’, ²⁰ whereas the evidentiary facts from which intentionality inferences are drawn and the probative standards against which they are evaluated are ‘external’ and ‘objective’. ²¹ In other words, the fact-finder is required to prove inferentially ‘what actually went on in the mind of the [opponent]’ ²² while deploying the objective or external criteria and mechanisms of the Law of Evidence. Without suggesting that this contrast implies an intractable (Cartesian) problem, ²³ my argument is that the kind of inference involved under the circumstances at the very least begs for some explanation or clarification. This is particularly given the danger of drawing inferences that are too far stretched (not immediate) between antecedents and consequents that are too remote (i.e. without an inferential universal or common ground). ²⁴ The core aim of this article is to investigate the nature of the probative inference of intention in criminal adjudication.

The second dimension of the complication for the proof of intention arises from certain semantic features that have the effect of widening or raising the prosecution's burden of proof. In the first instance, the conventional ‘I didn’t mean to do it’ denial can be disambiguated into two cognitive and volitional sub-categories. ²⁵ The denial could be based on a defect of knowledge about the factual circumstances that prevailed when performing her intended action (cognitive denial), ²⁶ or it could be a volitional error involving the opponent choosing the wrong action as a result of an extraneous hindrance on her freedom (volitional denial). ²⁷ For example, an opponent that genuinely, yet erroneously, thought that the gun that she fired was unloaded denies intention for cognitive reasons, ²⁸ whereas the opponent that misfires the gun through a muscular spasm or defect owing to her intoxication ²⁹ ; being placed under duress ³⁰ ; or through some other physical accident, ³¹ denies intention for volitional reasons. The further disambiguation of the opponent's denial into further sub-categories in this way gives the prosecution the additional hurdle of having to refine the specific contours of the opponent's denial. In the second instance, the proof of intention is complicated by the fact that the basic linear one-to-one relationship of probative intention described earlier is not at all frequent in the practice of criminal adjudication. The more frequent case in disputes of intention entails a multivalued one-to-many relationship of the kind represented in Figure 2.

OPEN IN VIEWER

Figure 2.

Multivalued one-to-many probative argument.

Here the opponent has performed a single action (e.g. discharging or wielding a weapon) that has resulted in multiple prohibited consequences. As depicted in Figure 2, the range of prohibited consequences is not coextensive with the scope of the opponent's intention. In fact, the further downward escalation of prohibited consequences (into prohibited consequences 2B, 1B, 3B, and so forth) reduces the scope of the opponent's intention. The multifaceted nature of both the opponent's disambiguated (cognitive and volitional) denial of intention and the multivalued nature of intention in the more common one-to-many probative cases are significant (theoretical) challenges for the proof of intention in the context of criminal adjudication around the world. This article addresses this challenge by developing the basic components of the general probative argument of intention. This general argument is explicated more fully in the last part of this article. The first two parts discuss the consequent (i.e. the material element of intention as defined in criminal law doctrine) and antecedent (i.e. the evidentiary facts from which intentionality inferences are drawn) of the overall argument. The last part of the article then introduces a set of intermediate facts (knowledge, desire and control) that bolster the inferential strength of the argument by connecting the antecedent and the consequent.

The Consequent: the Legal Meaning of Intention

The basic structure (E to F) of probative arguments ³² entails inferring material conclusions (consequent) from admissible evidentiary facts (antecedent). ³³ This section of the article explicates the consequece of the probative argument of intention, whereas the next section discusses the antecedent of the argument. The last section then concludes the discussion by analysing the nature of the inference that connects these two basic components. Wigmore pertinently pointed out that the ‘first step’ ³⁴ in any process of proof is to answer two ‘prime questions serving as the key,’ ³⁵ namely: ‘[w]hat is the Proposition desired to be proved’ and ‘[w]hat is the Evidentiary Fact offered to prove it.’ ³⁶ Although the identification and definition of the consequent (factum probandum) is largely a semantic function of criminal law doctrine, an incorrect diagnosis of this element can potentially lead to logical (and evidential) errors. ³⁷ The purpose of this section of the article is to explicate the nature of the factum probandum of intention by briefly locating its development within the broader history of criminal fault and then proceeding to unpack its various disaggregated sub-components across various jurisdictions.

The development of most criminal justice systems follows a similar trajectory in at least one respect, namely: crimes are punished strictly at the beginning and the role of fault on the merits (as distinguishable from sentencing) only emerges much later on. ³⁸ Maitland pertinently bemoans the ‘utter incompetence of ancient [European] law to take note of the mental elements of crime’, ³⁹ whereas pre-modern African legislators were concerned that requiring anything less than strict liability for punishing unlawful killings ‘would result in increased bloodshed’. ⁴⁰ However, the extent of an opponent's fault could be a factor of sentencing (as distinguishable from the merits) in certain cases. ⁴¹ For example, Roman law distinguished between deliberate (dolus) and negligent (culpa) killings, ⁴² and the Code of Hammurabi differentiated killings with and without intent, ⁴³ but these distinctions pertained to the sentencing (type of punishment) to be imposed, and not the merits.

The role of fault on the merits only emerges much later in the development of most systems of criminal law. It was only through the influence of the medieval Catholic Church that merit-based fault, as we understand the term today, began to emerge. ⁴⁴ The medieval Catholic Church was ‘the first modern state’. ⁴⁵ It established its own ecclesiastical courts staffed by judges that were priests. ⁴⁶ The Church influenced the development of merit-based fault in two ways. Firstly, responsibility would no longer be understood in collective terms, such that the members of a household or family would be held responsible for the crimes of one or more of its individual members. ⁴⁷ Each individual would be held responsible for their own unlawful deeds. It is also in this context, for example, that the Catholic confessional was introduced in that Canon 21 of the Fourth Lateran Council of 1215 imposed a mandatory annual ritual of confessing and repenting for one's sins. ⁴⁸ Secondly, the conceptual baggage of the Christian notions of sin and guilt were interposed into the meanings of crime and responsibility. ⁴⁹ Some of this conceptual baggage included sin as an ‘internal’ or ‘mental’ violation. ⁵⁰ The Bible is replete with related references to Jesus teaching His disciples about sin originating ‘from within’ and ‘from the human heart’. ⁵¹ Furthermore, the concept of sin was also enmeshed with Cartesian distinctions between ‘body and soul’; ‘heart and mind’; and ‘intellect and will’ ⁵² . Therefore, it came as no surprise that this conceptual machinery that accompanied the Christian notion of sin would later be incorporated into the contemporary criminal law concept of mens rea. In fact, the maxim actus non facit reum nisi mens sit rea (‘the act is not guilty unless the mind is guilty’) originates from a sermon (the Epistle of James, 5:12) given by an African theologian named St. Augustine of Hippo in the 4th century. ⁵³ The maxim would then later appear in Europe in Henry I's Leges Regis Henrici Primi (1118) ⁵⁴ and in Coke's discussion of the mental aspect of the crime of treason in the Third Institute (1628). ⁵⁵ A similar spread of the influence of medieval Christian ideas (the individualised Cartesian conception of fault) is noticeable among Civilian jurisdictions too. Early medieval Italian jurists ⁵⁶ and late Enlightenment German theorists ⁵⁷ were the protagonists in these parts of the world.

The substantive commitment towards conceiving of fault in these ‘subjective’ terms, that is, as referring to the ‘individualised’ ‘mental’ aspects of an opponent, continues to persist in the contemporary era. Contemporary fact-finders, unlike their pre-modern predecessors, do not see themselves as being hindered by the medieval belief that one ‘cannot look into a man's mind.’ ⁵⁸ The injunction is rather that: ‘you must look into it, if you are going to find fraud against him; and unless you think you see what must have been in his mind, you cannot find him guilty of fraud.’ ⁵⁹ In other words, fault is ‘subjective’ because it consists in ‘what actually went on in the mind of the [opponent]’. ⁶⁰ This subjectivist conception is instantiated in the criminal law doctrine of several jurisdictions across the world. For example, this includes: the principio de culpabilidad (‘principle of culpability’) in the Spanish-speaking parts of South America; ⁶¹ the Chinese ⁶² and Russian ⁶³ ‘psychological’ or ‘subjective’ components of crime in the far east; ⁶⁴ the common law presumption of mens rea; ⁶⁵ and the Civilian culpability principle ⁶⁶ (e.g. the French culpabilité ⁶⁷ and the German Schuldprinzip). ⁶⁸ African approaches, largely because of their European colonial heritage, to criminal fault reflect a similar ‘subjectivity’. ⁶⁹

However, it is one thing to hold a substantive commitment towards an individualised and subjective (in the senses described earlier) conception of fault, but quite another to be able to prove it. This difficulty is also particularly acute with respect to intention than any other form of fault. One of the earliest judicial interventions with regard to the proof of intention took the form of logical presumptions. These were persuasive (logical) generalisations that emerged from the collective experience of fact-finders having reasoned in similar ways in certain classes of cases. ⁷⁰ In the context of intentional murders, this type of presumption enabled fact-finders to bridge the logical gap between an opponent's unlawful actions and their degree of responsibility: (1) the opponent fired a dangerous weapon at the victim; (2) therefore, the opponent fired the dangerous weapon intentionally. Among common law jurisdictions, it was then widely accepted that: an opponent is presumed to have intended all the natural and probable consequences that arise from their unlawful actions. ⁷¹ The versari in re illicita doctrine is the Civilian analogue to this common law presumption. ⁷² Being a product of the late Enlightenment period, this particular heuristic later became obsolete during the 20th century. Apart from its logical flaws, common law jurists became concerned about the effect it had in lowering the prosecution's burden of proof. ⁷³ Contemporary practices of fact-finding remain committed, sans the aforementioned logical presumptions, to the effectiveness of inferential reasoning in making probative judgments about intention, but criminal law theorists have argued that we now need a broader range of concepts to be able to develop plausible arguments of criminal intent. ⁷⁴ As a result, several jurisdictions have adopted disaggregated approaches that break down intention into several smaller concepts such as knowledge, foresight, willingness, belief, hope, desire, purpose, suspicion, awareness and the like. These smaller intermediate concepts are also typically conceived of in hierarchical terms. For example, while civilian jurisdictions are known for their binary models (dolus and culpa) of fault, ⁷⁵ dolus is sub-divided further into three hierarchical levels, namely: dolus directus, dolus indirectus and dolus eventualis. ⁷⁶ Among common law jurisdictions, the U.S. model consisting of purpose, knowledge, recklessness and negligence is widely embraced. ⁷⁷ In either case (common law or civil law), intention in its ‘purposive’ or ‘desirous’ form sits at the very top of the hierarchy of fault. ⁷⁸

Therefore, for purposes of the probative argument developed later in this article, the nature of the consequent intention consists at least of the desire or purpose of an opponent (deliberateness). However, problems such as the marketplace or aircraft bomber ⁷⁹ ; the ignorant arsonist ⁸⁰ ; the accidental shooter during a competitive duel ⁸¹ ; and the skier that accidentally collides with another person to their death, ⁸² necessitate the expansion of the ordinary (qua ‘desire’) meaning of intention. ⁸³ The second expanded meaning of intention consists of acting with the knowledge or foresight of certain probable, yet undesired, consequences (non-deliberateness). ⁸⁴ This secondary meaning is sometimes known as ‘oblique intention’ ⁸⁵ or dolus indirectus. ⁸⁶

The Antecedent: The Evidentiary Facts

This section of the article briefly outlines some of the pertinent considerations for determining from which kinds of evidentiary facts (antecedents) can the corresponding consequent of intention (in one of its two variants that were discussed earlier) be inferred. Evidentiary facts typically emerge in the form of ‘brute facts’, ⁸⁷ namely: (i) a murder weapon; (ii) a clothing item; (iii) bullet holes on a door; (iv) a written or oral witness statement. These examples track the common typologies of evidence that include: direct, circumstantial, real, collateral, testimonial and non-testimonial evidence. ⁸⁸ To be able to construct probative arguments of intention, each of these ‘brute facts’ needs to be formulated in the form of a ‘proposition’ of fact; that is, ‘a statement of assertion containing a factual predicate.’ ⁸⁹ Our reformulated examples of ‘brute facts’ thus become: (i) the murder weapon was found at the opponent's home; (ii) a clothing item belonging to the opponent was found at the crime scene; (iii) the bullet holes on the door reflect a close grouping; (iv) the eye-witness testified about the identity of the perpetrator. Propositions of fact are a lot more enlightening than their constitutive ‘brute facts’.

There are two potential dangers that could arise from relying on antecedent evidentiary facts; that is, reasoning from evidentiary facts that are either inauthentic or unreliable (authenticity) or from antecedent facts that are too remote or irrelevant to forge a logical connection with the factum probandum of intention (relevance). The authenticity danger arises from the manner (form) in which the evidentiary facts were obtained or presented in court, whereas the relevance danger pertains to their propositional content. ⁹⁰ As indicated earlier, probative arguments typically take the form, E to F. ⁹¹ Inferences about intention (F) that are drawn from inauthentic evidentiary facts or without an evidentiary basis at all is what has been referred to as ‘speculation or conjecture’. ⁹² Drawing an inference under these circumstances would make the argument a classic non-sequitur (the conclusion would not follow from the premise). ⁹³ An illustrative example of this is the following form of hearsay evidence: (i) the opponent's plan to kill the deceased is revealed in a written document compiled by an anonymous source; therefore, the opponent intended to kill the deceased. ⁹⁴ Although the Civilian fact-finder would not be precluded from accessing this kind of evidentiary fact as a common law fact-finder would be, ⁹⁵ the potential logical errors that could arise from authenticity dangers are germane to all forms of probative argumentation. Similarly, the potential error that could emerge from the second relevance danger is the construction of a non-sequitur. However, this would not be on the basis of relying on inauthentic or non-existent (baseless) evidentiary facts, but rather that the logical connection between the evidentiary antecedent and intention consequent is far too remote to be able to draw an inferential link. In common law jurisdictions, irrelevant evidence is irredeemably excluded as a matter of law. ⁹⁶ Civilian fact-finders would similarly be careful not to rely on irrelevant evidence, ⁹⁷ although they would not be precluded from gaining access to this evidence by a categoric exclusionary rule. Therefore, the overall aim of probative argumentation with respect to intention should be rely, as far as possible, on the ‘best evidence’ ⁹⁸ to forge the most immediate or proximate logical connections in order to avoid the respective dangers of authenticity and relevance.

A Basic Model of the Inference of Intention

The two preceding sections of this article focused on the basic components (the antecedent evidentiary facts, E, and the consequent factum probandum, F) of the probative argument of intention. The nature of the inference that connects these two components will be the focus of the current section. It is worth also clarifying that the focus of the section is to develop the basic structure of the argument without discussing the evaluation of its probative weight, which is a subject for a separate project entirely. Inferences may be in the form of generalisations (All Es are Fs) ⁹⁹ or conditionals (If E, then F). ¹⁰⁰ The distinction between these two forms bears no implications for the development of our basic model of the probative argument of intention. However, inferences can be classified further into three logical categories that affect the structure and cogency of intention arguments. An inference can be deductive (If E, then F always follows) ¹⁰¹ ; inductive (If E, then F probably follows in a specified portion of cases) ¹⁰² ; or defeasible (If E, then F plausibly follows, subject to any exceptions). ¹⁰³ Deductive inferences are generally regarded as being inappropriate for the peculiar context of fact-finding in a court of law. ¹⁰⁴ One of the main reasons for this is that the antecedent (E) in probative argumentation typically is disputed and can only ever be proven in probabilistic terms. ¹⁰⁵ As a result, the inferences of probative arguments generally have to be hedged (probably, plausibly or generally). ¹⁰⁶ Although inductive reasoning was shown support by scholars in the early twentieth century, ¹⁰⁷ deploying these types of inferences without statistical data leaves them vulnerable to the criticism that they are insufficiently robust. ¹⁰⁸ Although one cannot classify all forms of probative argumentation into one exclusive (non-pluralistic) category, ¹⁰⁹ defeasible logic seems to fit most forms of probative arguments, including that of intention. Defeasible inferences are identifiable by nonmonotonic quantifiers such as ‘plausibly’ or ‘generally’ because they are accepted tentatively, but subject to change with the addition of new premisses (new evidence). ¹¹⁰ In the absence of new evidence, the effect of this tentative commitment is to shift the burden of proof onto the opponent in the broader context of a dialogue. ¹¹¹ Therefore, a basic iteration of defeasible probative argumentation bears the following form:

If the pertinent evidentiary facts (E) are established, then the factum probandum (F) has been plausibly established.

The inference for both the deliberate and non-deliberate (oblique) forms of intention is presumptive (using the ‘plausibility,’ →, quantifier). The major difference lies in the nature of the antecedent evidentiary facts. The inference for the deliberateness model (desire or purpose) of intention is drawn from knowledge (K) and desire (D), whereas for the expanded non-deliberateness model (oblique intention) the inference is drawn from knowledge and control (C). These intermediate propositions (knowledge, desire and control) are in turn inferred from the kinds of evidentiary facts (antecedents) discussed earlier. The remainder of this part of the article will first lay out the preliminary arguments from the antecedent brute evidential facts to the consequent three intermediate facts. The first epistemic intermediate argument entails that:

Evidential facts that form part of an opponent's general experience (E) plausibly will be known by the opponent (K). ¹²⁰

Conclusion

Publications on the evidential proof of intention are relatively rare in comparison to the separate categories of the doctrinal meaning of intention and the general methods (procedures) of fact-finding in criminal trials. The subject matter of this article has combined these two strands of scholarship. The article has developed the overall probative argument of intention in the context of criminal trials. The basic components of the argument, namely the antecedent evidentiary facts and the consequent factum probandum of intention, are discussed in the first two sections of the article. The last section then proceeds to discuss the nature of the inference that connects these two components. The overall argument is a basic modus ponens (affirming the antecedent) that is adapted from the general model of evidential argumentation (E to F; E; therefore, F) that has been developed by Floris Bex and other legal theorists. The argument distinguishes between two basic variants of criminal intention, namely: one that consists of deliberateness and another that is characterised by non-deliberateness. The main aim of the article was to pull together our basic intuitions and ordinary arguments about intention into a clearly identifiable structure of probative argumentation. To that extent, the article has not attempted to introduce much that is new, except to structure and polish everything that each one of us already knows and thinks about intention. Apart from this, I know no other purpose for jurisprudence.