ISSA Proceedings 2002 – E Contrario Reasoning And It’s Legal Consequences
E contrario reasoning is the argument that says that a certain legal rule must not be applied analogically, that is to say: that the legal rule must not be applied to certain facts that are not mentioned in this legal rule. A very nice example is mentioned by Harm Kloosterhuis in his recent dissertation. The example deals with the question whether or not the legal rule that holds persons liable for damage caused by tort committed in a group may be applied to a group of dogs that have caused damage. The Court judged this analogy unsound (NJ 1996, 172; Kloosterhuis 2002, 197).
The argument is famous for it’s logical validity problem. The problem is that at the surface e contrario reasoning seems to produce the fallacy of denying the antecedent. When, in modus ponens, the first premise states the legal rule ‘if p, then q’ – the antecedent ‘p’ meaning the legal facts and the consequent ‘q’ meaning the legal consequence – and the second premise states the concrete case which does not match the legal facts – symbolised by ‘not p’ – the conclusion ‘not q’ – meaning that the legal consequence is not entailed – is a logically invalid inference.
About a decade ago, the International Journal for the Semiotics of Law published a discussion between two jurisprudential scholars about the solution for the logical problem of e contrario reasoning. In this discussion they both tried to analyse e contrario reasoning as a logically valid argument. I agree with them that it is reasonable to suppose that e contrario reasoning is logically valid indeed, because there may be good reasons not to apply a legal rule analogically in specific circumstances and in law it happens all the time without any problem. However, in this paper I want to argue that both scholars, Maarten Henket from Utrecht University and Hendrik Kaptein from the University of Amsterdam, are solving a non existent problem. The reason for this is that two different types of e contrario reasoning have to be distinguished, that have to be analysed very differently. Kaptein and Henket have confused these two types by combining features of both types in one type of argument.
My strategy is as follows. First, I will elaborate on the validity problem and show how Kaptein and Henket have tried to solve the problem. Then I will show which argument types can be distinguished as instances of e contrario reasoning. Finally, I will try to find an explanation for why it is often overlooked that two concepts of e contrario reasoning exist.
2. The supposed logical problem of e contrario reasoning
In the literature the logical problem of the e contrario argument has usually been solved in the following way: suppose that the legal rule on which the argument is based contains a necessary condition for the applicability of the legal consequence. The legal rule is then considered to mean ‘only if p, then q’. When the legal rule is interpreted this way, a position in which the legal consequence is negated might be drawn in a logically valid way.
So it seems that the logical problem could easily be solved. However, the interpretation of the antecedent of a legal rule as being a necessary condition for the applicability of the legal consequence does not reflect reality in a lot of cases. Sometimes different legal rules pose the same legal consequence, e.g. rules that regulate paying damages or rules in criminal law that regulate a certain sentence (Kaptein 1993, 318-319; Henket 1992, 160-161)[ii]. In these rules the description of the legal facts is to be considered to pose a sufficient condition for the appearance of the legal consequence, not a neccesary one[iii]. So, a logical solution for the supposed problem of e contrario reasoning should also include reasoning based on legal rules of which the legal facts only pose a sufficient condition for the legal consequence to follow.
In their explanation of the logical validity of e contrario reasoning both Kaptein and Henket appeal to the adjudicational context in which the legal rule is applied. According to Kaptein (1993, 319, 321) this context carries that in the specific case at hand none of the other legal rules is applicable that eventually could have the same legal consequence aimed for as the disputed legal rule. This view is expressed in the analysis by an extra premise to the reasoning form in which it is stated that no other ground than p is applicable in the concrete case to obtain the legal consequence q. This premise may be ‘if q, then p’ – meaning no other legal rule with the same consequence is applicable – or it may be the two premises ‘if q, then [p v r v s]’ and ‘not p, not r, not s’ – specifying the other legal rules with the same legal consequence that are not applicable. The reasoning now states that no ground for the conclusion ‘q’ exists, so ‘q’ will not occur.
Henket’s appeal to the adjudicational context implies a different method. On the one hand he states that a lot of legal rules might be interpreted as containing a necessary condition of themselves; on the other hand he recognises that some rules cannot be interpreted this way. When such rules must be applied, it must be supposed – according to Henket – that the logical status of a legal rule changes as a result of the facts of the case (1992, 160-161). That means that in the case of reasoning e contrario the legal facts of the rule function as a necessary condition for the following of the legal consequence, whereas in the case of analogy the legal facts of the rule function as a sufficient condition.
In my view, these analyses do not reflect e contrario reasoning in a correct way. Two types of e contrario reasoning have to be distinguished, one of which does not pose a logical problem and one of which the logical problem could easily be solved in the traditional way. The reason that Henket and Kaptein must resort to their analyses is that their concept of e contrario reasoning is a mixture of these two different types.
3. Modern and classic e contrario reasoning
Let’s have a closer look at the example of e contrario reasoning that I mentioned earlier. The court argued that the legal rule that states that damages can be recovered on the basis of tort committed in a group is not applicable to dogs. The argument for this is that dogs do not meet the criteria stated in the rule: they ’re not able to act unlawfully, they’re not able to withdraw each other from actions, and it is not likely that dogs can be taught a sense of values, which is necessary for the applicability of this legal rule.
Let us contrast this example with another example of e contrario reasoning. In order not to complicate things I have chosen an example from daily life instead of a juridical example. Suppose that the organizers of this conference would have decreed that smoking is not allowed in the rooms where lunch is served. This decree can be reformulated as the rule: ‘If X finds himself in a room where lunch is served, then X may not smoke’. Then, the question could come up whether or not someone smoking in another room than the lunchroom, e.g. in the hallway, is allowed. A smoker could argue that the rule only regulates smoking in the lunchroom and that, if it would have been otherwise, the rule would have been formulated more broadly. He then applies the rule e contrario by concluding that smoking is allowed in all other rooms. This can be reformulated as the rule: ‘If X does not find himself in a room where lunch is served, then X may smoke’. This reasoning is based on the inversion of the legal rule, assuming that this rule also stipulates the opposed legal consequences to the opposed legal facts.
Both types of reasoning are instances of e contrario reasoning. In the literature, people often consider these as similar types of reasoning, just like Kaptein and Henket do. However, in my view these types differ in important respects.
The first example concerns argumentation that leads to the decision not to apply a legal rule analogically in a specific case. The phrase ‘e contrario reasoning’ points to a certain result in a certain context: the result not to apply the legal rule in the context that the disputed legal rule does not mention the facts at hand. The decision is a decision for the concrete facts at hand and does not say anything about the (in)applicability of the legal rule to other non settled facts that might ever occur. So, the same rule may be applied analogically the one time and e contrario the other. I will call this type of argument modern e contrario reasoning, because it is the type of e contrario argument described in modern jurisprudential literature.
The second example also shows a concrete case that is not settled by the disputed rule. However, the conclusion that is reached by the smoker does not only apply the facts at hand, but applies all the facts that are unmentioned in the rule. That’s because this kind of reasoning is not concerned with the inapplicability of rules. On the contrary, the rule is considered to be applicable indeed, but in a reversed way. The rule is a ground for a conclusion about the treatment of the opposed facts. I will call this type of e contrario reasoning classic e contrario reasoning, because in Dutch literature it is only described in older texts[iv].
Classic e contrario reasoning is reasoning from contrasts. What is at stake here is the question whether or not the legislator – in this case the organisers of this conference – wrote this rule for a very exceptional case, and therefore may be supposed to have intended to indirectly regulate the unmentioned facts. Could this question be confirmed, then an implicit second legal rule is presumed to be hidden in the explicit legal rule. The explicit and the implicit legal rule are mutually exclusive rules and therefore contrast each other. The implicit legal rule might regulate the facts at hand.
4. The logical analysis of modern and classic e contrario reasoning
So far, I hope that I’ve shown that modern and classic e contrario reasoning are different types of argument. Although they resemble each other in the sense that both can be used in the situation of no legal rule settling the facts at hand and both are an alternative to the possibility to apply the legal rule analogically, their implications are very different. The main and essential implication of the differences between modern and classic e contrario reasoning is that classic e contrario reasoning does, and modern e contrario reasoning does not lead to a stance in which the legal consequence of the rule at hand is negated on the ground that the facts at hand do no match the facts stated in the legal rule. That has implications for the way in which both types of reasoning have to be analysed.
Which logical analysis suits the classic e contrario argument? The answer depends on how the opposite legal rule might be deduced from the explicit legal rule. If we presume that the adjudicator who reversed the legal rule did that with the intention to argue in a logically valid way, the antecedent of the legal rule is supposed to contain both a sufficient and a necessary condition for the applicability of the legal consequence. The logical condition of the legal rule is then said to be that of material equivalence; an adequate representation of a legal rule that is supposed to settle an exceptional case(v). The logical condition of material equivalence and the classic e contrario type go hand in hand, because the legal rule being an equivalence implies necessarily that it contains two legal rules at the same time, the explicit and the implicit one. So, the logical problem of classic e contrario reasoning necessarily has to be solved in the traditional way[vi].
In contrast to classic e contrario reasoning, modern e contrario reasoning does not itself lead to a position in which the consequence of the legal rule is negated. The negation of the legal consequence cannot be concluded on the basis of the non applicability of the legal rule. The claim fails because a legal ground lacks. Whether or not a legal consequence will follow and which consequence that will be, depends on the applicability of other legal rules and on the procedural context in which other rules or may not be applied. The effort of Henket, making the conditionality of a legal rule change according to the result of the reasoning is not only circular, but also superfluous[vii]. The same applies to Kaptein. He offers an elegant explanation for the factual non following of the legal consequent on the basis of the lacking of any other relevant legal rule, but he does not solve a problem, because the legal rule is no basis of any conclusion at all[viii]. Anyway, more important is that this analysis cannot represent the classic type of e contrario reasoning. Classic e contrario reasoning is not concerned with the inapplicability of legal rules to concrete facts, because the implicit, deduced rule is supposed to regulate the opposed facts. So, by classic e contrario reasoning the legal consequence follows directly from the fact that the explicit legal rule does not mention the facts at hand[ix].
5. Why a mixed up concept of e contrario reasoning?
I hope that I’ve made clear that the two types of reasoning that are both called e contrario reasoning have to be analysed in a very different way and that no really difficult logical problem exists for both types of reasoning.
The analyses made above show that for modern e contrario reasoning no logical problem exists and that the logical problem for the classic e contrario argument is easily solved. The problem signalled by Kaptein and Henket is the result of mixing up the two types of argument, namely the broad definition of the modern type – a rule being applicable analogically in the one case and e contrario in the other – combined with the belief that from this kind of reasoning a stance should follow in which the legal consequence is negated.
The observation of two types leaves the question open how it might be explained that in jurisprudential literature both types of e contrario reasoning have been mixed up so often. Part of the answer lies in the general description of e contrario reasoning as the reasoning by which a legal rule is not applied analogically. This description suits both types. Once someone ever uncarefully has described the classic type in the general way, it is not strange that the ignorant reader reads something different than what was originally meant. In this regard I believe that the classic concept of e contrario reasoning has been widened as a result of vague descriptions, examples that suit at first sight both argument types, and a logical analysis, in which ‘not p’ can be interpreted in two ways, namely in the way that it represents the non matching of one of all possible facts which might occur with respect to the legal rule, and in the way ‘not p’ being the class of facts opposite to ‘p’.
Explaining modern e contrario reasoning as the widening of the classic e contrario argument would explain why the expression ‘e contrario’ does not seem to have any meaning for the modern e contrario argument. After all, ‘e contrario’ does not refer to a specific kind of reasoning that would be based on contrasts, like for example the appeal to authority refers to the argument in which an authority is quoted to confirm the conclusion. In contrast, the classic e contrario reasoning type might very well be represented by the phrasing ‘reasoning from contradictions’. My suggestion therefore is that the classic e contrario argument is the original e contrario argument, from which the modern type has arised.
[i] Many thanks to Taco Groenewegen, who not only has been very helpful in writing this paper, but also has been very good company in reflecting on the subject of this article since a very long time.
[ii] Of course it is not the case that legal rules which contain a unique legal consequence always can be analysed as the antecedent posing a necessary condition for the consequent. So, other, intrinsic reasons exist for deciding whether or not a legal rule has this logical condition.
[iii] According to Kaptein (1993, 318), material implication is the standard interpretation of legal rules; whereas according to Henket (1992, 154, 164) a lot of rules might be interpreted as rules of which the antecedent is a necessary condition for the legal consequent.
[iv] Only a few authors make a distinction between classic e contrario reasoning, that they just call e contrario reasoning, and the modern argument, that they call the non appliance of a legal rule or argumentum e silentio or the like. See Van Bemmelen (1891, 17), Bydlinski (1991, 477), Canaris (1983, 52), Klug (1982, 144), Schneider (1965, 180).
[v] A logical analysis enables one to judge the argument, because is has to be decided whether or not the legal rule does in fact state an exceptional case. This evaluative part must be decided on juridical grounds; this is not a matter of logic. Anyway, not many rules of law might be considered as rules regulating exceptional cases (Canaris 1983, 49-50); that’s why classic e contrario reasoning seldom is considered to be a sound argument.
[vi] Henket would actually agree on this part, for in his example of a legal rule that must be considered to contain a necessary condition of itself, he describes – without recognising – classic e contrario reasoning. Kaptein however, who reacts to this example, seems to believe that all e contrario reasoning relies on the inapplicability of other legal rules.
[vii] It is circular in the sense that a logical analysis is supposed to show the premises on which the argument is based, in order to evaluate the argument. In the case of e contrario reasoning this would mean whether or not the premise is interpreted correctly. However, if premises adjust themselves to the result of the reasoning, there’s nothing to evaluate.
[viii] Moreover, Kaptein’s analysis does not specifically represent the form of reasoning in which analogical appliance is negated, for it represents every reasoning form in which it is argued that the claim fails. These reasoning forms have never been considered to pose a problem no legal consequence occurring.
[ix] Cf. Canaris (1983, 51), Klug (1982, 139) and Schneider (1965, 198).
Bemmelen, P. van (1891). Rechtsgeleerde Opstellen I. Leiden: Brill.
Bydlinski, F. (1991). Juristische Methodenlehre und Rechtsbegriff. Wien/New York: Springer.
Canaris, C.W. (1983). Die Feststellung von Lücken im Gesetz. Eine methodologische Studie über Voraussetzungen und Grenzen der richterlichen Rechtsfortbildung praeter legem. [2nd revised ed.] Berlijn: Duncker & Humblot.
Diephuis, G. (1869). Het Nederlandsch Burgerlijk Regt, 1e deel. Groningen: Wolters.
Henket, M. (1992). “On the logical analysis of judicial decisions”. International Journal for the Semiotics of Law 5, 14, 154-164.
Kaptein, H. (1993). “E contrario arguments in law: From interpretation to implicit premisses”. International Journal for the Semiotics of Law 6, 18, 315-324.
Klug, U. (1982). Juristische Logik. Berlin: Springer-Verlag.
Schneider, E. (1965). Logik für Juristen. Die Grundlagen der Denklehre und der Rechtsanwendung. Berlin/Frankfurt am Mein: Franz Verlag Vahlen.