Indrek Reiland (University of Vienna), "Rules: Regulative and Constitutive"
Analysis (2023) & Philosophy and Phenomenological Research (2020; 2023)
Based on:
“Regulative Rules: A Distinctive Normative Kind”, Philosophy and Phenomenological Research, 2023
“Constitutive Rules: Games, Language, and Assertion”, Philosophy and Phenomenological Research, 2020
“Squid Games and the Lusory Attitude”, Analysis, 2023
We are everywhere surrounded by rules that constrain us but also make certain things possible. Regulative rules like social rules and laws regulate our everyday life and thereby support its smooth functioning. Constitutive rules make possible playing games, and, on some views, speaking languages and performing speech acts like assertion. One might even think that subjecting ourselves to rules is what makes humans distinctive, or, as Wilfrid Sellars evocatively put it, that without them we would walk on four feet.
But what are regulative rules in general and how are constitutive rules special?
Regulative rules are too frequently run together either with orders or with normative truths like moral, epistemic, legal, or strategic ones. However, they should be sharply distinguished from both. Unlike orders and other things with imperatival content, rules are best thought to have propositional content which attributes some action-type A (e. g. driving on the left, turning right) some deontic status like being required, forbidden, or permissible, sometimes on certain conditions C (there’s a red light). Here are two schematic ways they can be written down together with examples:
(R1) (If/only if/if and only if C), doing A is required/forbidden/permissible
Driving on the left is required. (A traffic rule in the UK)
(R2) One must/can’t/may do A (if/only if/iff C)
One can’t turn right if there is a red light. (A traffic rule in New York)
However, unlike normative truths which can have similar contents, rules aren’t true, but in force. For example, legal rules are contents that are in force because they’ve been enacted by the relevant authority. Social rules like rules of etiquette are in force in a community because they’re generally accepted. And rules of games are usually in force for the players at the time of playing because they themselves accept them.
Whereas normative truths are true, rules are in force. This is an essential difference. Rules, qua things that are in force are, despite their propositional contents, not themselves truth-evaluable. This is because to make a rule, to enact, is not to do something constative, something that has to fit pre-existing reality to be correct. Rather, it is to do something performative, something that seeks to change pre-existing reality by creating a new legal etc. truth or fact. Similarly, acceptance of a rule isn’t a constative attitude, but something that institutes normative reality.
Consider No Right:
(No Right) Turning right on red is forbidden.
This can be used to state the traffic rule that is in force in New York and most places in Europe, but not in the rest of the US, but also the legal fact or truth that in New York turning right on red is forbidden. If an authority enacts No Right in New York then it comes to be in force, it becomes a rule in New York. The rule’s being in force purports to generate a legal fact or truth. And absent conflicting rules, it does generate the legal fact or truth that that turning on right on red is forbidden in New York.
The distinction between rules and certain paradigmatic normative truths can be made even sharper by considering their relation to contributory notions like reasons. Rules are enacted/accepted for reasons. These are what constitute the rule’s justification. For example, No Right is put in force by the relevant authority in New York City because traffic is dense, and it helps to prevent accidents. The important point is that once a rule is in force it is isolated from its justification in the sense that its normative force isn’t defeated if it turns out that in this particular case there’s more reason to do the opposite. If you are legally prohibited from driving through a red light then this legal prohibition doesn’t go out of effect even if there is overwhelming reason to do so. In such cases we usually just break the rule, and the fact that we consider this breaking proves the point. In contrast, many paradigmatic normative truths are contribution-dependent and don’t allow for such a separation between their normative force and their justification. In the above case, if there is overwhelming reason to drive through a red light then this is what you all-things-considered ought to do.
To sum up, regulative rules are general normative contents that are in force due to enactment or acceptance. They are not themselves truth-evaluable and their normative force is isolated from their justification.
Given this conception of regulative rules in general, what makes constitutive rules special? The intuitive distinction between merely regulative and constitutive rules which we have inherited from Rawls, Searle etc. is that while merely regulative rules just regulate our actions, constitutive rules make possible the performance of new actions like playing a game, speaking a language, and assertion. How does this work?
Searle’s influential discussion of constitutive rules mixed together two very different ideas. On the one hand, he thought of some constitutive rules like those of games as a subset of regulative rules, as genuine rules that have some special properties. Such rules regulate antecedently existing actions and thereby somehow constitute, make possible, the performance of new actions. On the other hand, he also used the term ‘constitutive rule’ for things that aren’t genuine rules at all and that can be codified in the formula ‘X counts as Y in C’. Such “constitutive rules” don’t regulate anything, but just assign things a certain status in a certain context. They make such and such pieces of paper count as money, a particular line count as a border etc.
It is paramount in understanding the constitutive rules of games, language, and assertion to focus on the first idea and set the second one aside. The rules of games etc. are clearly a subset of regulative rules. Here are some examples:
(Pawn) A player may move a pawn two squares forward only if it hasn’t moved.
(Ouch!) A speaker s may use ‘Ouch!’ iff s is in pain.
(K-Assertion) A speaker s may say that p iff s knows that p.
The idea is that Pawn not only regulates the antecedent action of moving the pieces called ‘pawns’ two squares forward but, together with other rules governing the movements of pawns, makes possible the action of moving a pawn as a part of playing chess. Similarly, Ouch! not only regulates the use of the sentence ‘Ouch!’, but also makes possible using it to speak English and to express pain. Finally, K-Assertion not only regulates sayings, but also makes possible asserting.
Rules of games, languages, and assertion can all be broken while performing the rule-constituted action. Furthermore, and this is of central importance, they can be intentionally broken. Basketball, football, and hockey all involve intentional fouls. And most people find it intuitive that even cheaters are still playing the game. The typical retort is that games feature two types of rules, some that are central and can’t be broken and some that can and incur penalties. But this is false and depends on a bad view of what it is to perform rule-constituted actions. In boxing it’s illegal to kick, headbutt or hit below the waist. These proscriptions are all equally central in determining the nature of the game, even though the first one can’t be broken while continuing boxing, while the latter two can and merely incur penalties. The reasons why the first rule can’t be broken while continuing playing is rather that one couldn’t break it by accident and breaking it signals that one has stopped accepting the rules and participating in the game.
The possibility of intentional breaking is even clearer in the cases of language use and assertion. When you utter ‘Ouch!’ when not in pain to deceive someone into thinking that you are or when you lie, you intentionally break the relevant rules. But your use of ‘Ouch!’ still counts as speaking English and expressing pain, and your lie wouldn’t be one unless your saying counted as an assertion. To speak or assert incorrectly is still to speak or assert.
It follows that, contrary to standard Searlean lore, to engage in these types of rule-constituted activities one doesn’t have to act in accordance with the rules, follow them, or even try to do so. Rather, to play, speak, or assert is just for the rules to be in force for you. This is the essence of what I call the Governance view of what it is to perform a rule-constituted action:
Governance: To perform a rule-constituted action you have to perform the antecedent action while the constitutive rule is in force for you.
This is entirely compatible with your intentionally breaking the constitutive rule at the same time.
We’ve seen that constitutive rules are a subset of regulative ones, and found out what it is to perform rule-constituted actions, but we do not yet know what makes constitutive rules special. What sets constitutive rules apart from merely regulative rules are two further properties. First, unlike merely regulative rules which can just require, forbid, or permit doing something on certain conditions, constitutive rules must specify necessary and sufficient conditions for the action to have the relevant deontic status. (It follows that Pawn above is part of the full constitutive rule governing pawns). Otherwise they wouldn’t fully define a new action. Second, constitutive rules are distinctive because of the primary reasons they are put in force. Regulative rules like traffic rules are put in force for reasons that pertain to the extrinsic benefits gained from the regulation of some existing action. No Right is put in force by the relevant authority in New York City because traffic is dense, and it helps to prevent accidents. But constitutive rules are put in force for the primary reason that doing so makes possible performing the new action. For example, when chess players start a game they accept the rules for the primary reason that it makes possible playing it. Thus, a rule or a set of rules is constitutive if:
a) Content: it specifies necessary and sufficient conditions for the antecedent action to have the deontic status.
b) Justification: it is put in force for the primary reason that doing so makes possible performing the new action.
You can test this by coming up with a set of rules specifying necessary and sufficient conditions for an action to be required, forbidden, or permissible and considering a case where someone puts the rules in force for the primary reason that this makes possible performing the new action. The result will be that it starts looking like the person is doing something very similar to playing a game.
In the case of games, the rules usually come to be in force for the players at the time of playing because they voluntarily accept them. Suits famously built this into his definition on which to play a game is to engage in a “voluntary attempt to overcome unnecessary obstacles” (Suits 2005: 55). But this is a mistake. As Netflix’s recent hit show Squid Games vividly illustrates, players can be forced to play games on pain of punishment or death. And if players can be forced to play a game, then it can’t be essential to playing it that the rules come to be in force for them because they voluntarily accept them. Rather, what is necessary for players to play is that someone has put the rules in force for them so that they could play. This doesn’t have to be the players themselves, but can be a Frontman or umpires.
Here’s the whole picture, once again. Regulative rules are general normative contents that are in force due to enactment or acceptance. The constitutive rules of games, language, and assertion are a subset of regulative rules which regulate antecedently existing actions and, by being in force, constitute the possibility of performing a new actions. They are special in always specifying necessary and sufficient conditions for the anteceden action to have a deontic status and in that they’re put in force for the primary reason that it makes possible the performance of the new action. In the case of games, the players usually voluntarily put the rules in force for themselves. But it could also be that someone else puts them in force for the players so that they could play.
Suits, B. 2005. The Grasshopper: Games, Life, and Utopia, 2nd Ed. Ontario: Broadview Press