1. It is convenient to distinguish between the following six ‘components’ or ‘ingredients’ or ‘parts’ of norms which are prescriptions: the character the content the condition of application the authority the subject(s) and the occasion.

From a complete statement to the effect that such and such a prescription has been given it should also be clear which are its six above-mentioned components.

There are two more things which essentially belong to every prescription without however being ‘components’ of prescriptions in the same sense as the above six. These two we call promulgation and sanction.

The character the content and the condition of application constitute what I propose to call the norm-kernel. The norm-kernel is a logical structure which prescriptions have in common with other types of norm. There may however exist specific differences between the kernels of norms of different types. Here we are directly concerned with the kernels of prescriptions only.

The authority the subject(s) and the occasion seems to be specific characteristics of prescriptions which do not belong to the other types of norm.

The formal theory of norms or Deontic Logic which we are going to develop in later chapters of this work is essentially a theory of norm-kernels. Since the kernels are the common ingredients of all or nearly all types of norm this formal theory may with some caution be regarded as a ‘basic logic’ of norms in general.

2. The character of a norm depends upon whether the norm is to the effect that something ought to or may or must not be or be done.

For the ‘ought’—character of norms we shall introduce the symbol O and for the ‘may’—character the symbol P. Norms of the ‘ought’—character can also be called obligation-norms and norms of the ‘may’—character permissive norms. We shall also speak of the O-character and the P-character of norms and of O-norms and P-norms.

Later in this chapter we shall discuss in some detail the mutual relations between the three norm-characters. It will be seen that the ‘ought’—character and the ‘must not’—character are interdefinable. This is the reason why we have not introduced a special symbol for the second. It may be suggested that the ‘may’—character and the ‘must not’—character are interdefinable too. The question is open to debate and we shall not attempt to decide it. This is the reason why we retain a special symbol for the permissive norm-character.

If a prescription is to the effect that something ought to be done we often call it a command or order. If it is to the effect that something may be done we call it a permission. If finally it is to the effect that something must not be done we call it a prohibition.

Advice counsel prayer recommendation request warning are related categories to command permission and prohibition. We shall not however call them prescriptions or norms. We restrict the field of meaning of ‘prescription’ and ‘norm’ to things of the O-character or P-character. This seems in good accord with ordinary usage.

3. By the content of a norm we mean roughly speaking that which ought to or may or must not be or be done. The content of a prescription in particular is thus the prescribed (commanded permitted prohibited) thing.

From the point of view of their content norms (other than ideal rules) can be divided into two main groups viz. norms concerning action (acts and forbearances) and norms concerning activity. Both types of norm are common and important. ‘Close the door’ orders an act to be done. ‘Smoking allowed’ permits an activity. ‘If the dog barks don't run’ prohibits an activity.

It seems that prescriptions (and maybe other norms too) concerning activity are in an important sense secondary to prescriptions (norms) concerning action. Let us ask: What does the regulation ‘Smoking prohibited’ require us to do? The answer is: If we are engaged in the activity of smoking the regulation orders the act of ceasing to smoke (e.g. by throwing the cigarette away); and if we are not smoking it prohibits the act of starting to smoke (e.g. by lighting a cigarette). Similarly the command not to run if the dog barks orders the act of stopping should we happen to be running and prohibits the act of starting to run should we be walking or standing still.

Thus at least in some cases prescriptions (norms) concerning activity may become ‘translated’ into prescriptions (norms) concerning action. Whether this is always possible we shall not discuss.

4. The norm-contents with which we shall be concerned in our Deontic Logic are the meanings of df-expressions df-expressions it will be remembered are molecular compounds of atomic d-and/or f-expressions i.e. of sentences which describe generic acts and/or forbearances (Ch. IV Sect. 1).

It is convenient to divide norms into positive and negative depending upon whether their content is an act or a forbearance or strictly speaking: whether their content is the meaning of a molecular compound of atomic d-expressions or of atomic f-expressions. But it should be remembered that this division is not exhaustive. A norm whose content is the meaning of a df-expression with both atomic d- and atomic/expressions among its constituents falls in neither category. Such norms might be called (norms of) mixed (content).

‘One must not open the window nor shut the door’ enunciates a positive prohibition. ‘The door may be left open’ has the form of a negative permission ‘open the door’ that of a positive command. ‘Close the window but leave the door open’ illustrates a mixed prescription.

By an elementary norm we shall understand a norm whose content is an elementary act or forbearance.

As we know there correspond to one given state of affairs four elementary types of change and eight elementary types of act or forbearance (Ch. III Sects. 7 and 8).

Every one of the eight elementary types of act or forbearance can be the content of a O-norm or of a P-norm. The total number of types of elementary norm which correspond to one given state of affairs is therefore sixteen.

By an elementary O-expression we shall understand an expression formed of the letter O followed by an elementary d- or f-expression.

By an elementary P-expression we understand an expression formed of the letter P followed by an elementary d- or f-expression.

By an atomic O-expression we understand an expression formed of the letter O followed by a df-expression (elementary atomic or molecular).

By an atomic P-expression we understand an expression formed of the letter P followed by a df-expression.

By an OP-expression finally we understand a molecular compound of atomic O-expressions and/or P-expressions.

p-expressions T-expressions and df-expressions we have said are sentential symbols or sentence-schemas. They express propositions. They describe generic states of affairs changes and acts or forbearances respectively.

OP-expressions too may be regarded as schematic representations of sentences. Whether these sentences or ‘norm-formulations’ as we shall call them (see Chapter VI) express propositions is however a question which will have to be discussed later.

5. The condition which must be satisfied if there is to be an opportunity for doing the thing which is the content of a given norm (other than an ideal rule) will be called a condition of application of the norm. As will be seen presently this can be the sole condition of application of a given norm. But it need not be the sole condition (Section 6).

The conditions of application of elementary norms are simply the conditions for performing the corresponding elementary acts. Let p describe a state of affairs. Consider an occasion on which this state neither obtains nor comes into being independently of action. This constitutes an opportunity for effecting or leaving uneffected through action the elementary change described by ~pTp. Effecting this change can be ordered or permitted. Leaving it uneffected can similarly be ordered or permitted. The symbolic expressions for these four elementary prescriptions are: Od ( ~pTp) Pd( ~pTp) Of( ~pTp) and Pf( ~pTp) respectively.

As the reader will easily realize there are four elementary types of norm whose conditions of application are that a given state of affairs does not obtain but comes into being unless prevented through action; four elementary types of norm whose conditions of application are that a given state of affairs obtains and does not vanish independently of action; and four elementary types of norm whose conditions of application are that a given state of affairs obtains but vanishes unless prevented through action.

Every norm-content or strictly speaking proposition expressed by a df-expression is a truth-function of elementary acts and/or forbearances or strictly speaking of the propositions expressed by the df-constituents of this df-expression. The condition for doing the thing which is described by a given df-expression is a truth-function of the conditions for doing the things of which the proposition expressed by the df-expression is a truth-function. The condition for doing this thing moreover is the same truth-function of the conditions of the elementary acts and/or forbearances as is this thing itself of the corresponding elementary acts and/or forbearances.

Let p mean that the door is closed and q that the window is open. O(d( ~pTp) & f(qT ~q)) then is a symbolic expression of the command to close the door but leave the window open. The condition of application of this command is that both the door and the window are open and do not close ‘of themselves’ i.e. independently of action. If the window closes of itself but the condition otherwise remains the same a command which aims at the same result as the first would have to be formulated ‘close the door and keep the window open’.

6. From the point of view of their conditions of application norms can be divided into categorical and hypothetical.

We shall call a norm (other than an ideal rule) categorical if its condition of application is the condition which must be satisfied if there is going to be an opportunity for doing the thing which is its content and no further condition.

We shall call a norm (other than an ideal rule) hypothetical if its condition of application is the condition which must be satisfied if there is going to be an opportunity for doing the thing which is its content and some further condition.

If a norm is categorical its condition of application is given with its content. From knowing its content we know which its condition of application is. For this reason special mention of the condition is not necessary in a formulation of the norm. It is for example understood from an order to shut a window that it applies to a situation when a certain window is open.

If a norm is hypothetical its condition of application cannot be concluded from its content alone—if ‘content’ is defined as we have done here. Mention of the (additional) condition must therefore be made in its formulation. An example would be an order to shut a certain window if it starts raining.

As previously (Ch. I Sect. 7) stated it is important not to confuse hypothetical and technical norms. How the two types of norm shall be distinguished is not however quite easy to tell. We shall return to this question in Ch. IX Sects. 2 and 3.

Our symbolic notation is so far adequate only for expressing the norm-kernels of categorical norms. For dealing with hypothetical norms we shall have to make use of an embellished symbolism. The embellishment will not however be made until later in Ch. IX.

7. By the authority of a prescription I understand the agent who gives or issues the prescription. The authority orders permits or prohibits certain subjects to do certain things on certain occasions.

By calling the authority of a prescription an agent we indicate that prescriptions come into being as a result of action. For the peculiar mode of action which results in the existence of prescriptions we coin the name normative action.

Prescriptions which are thought to flow from a super-empirical agent as their authority we call theonomous. They are also called the commands or the law of God. We shall not here discuss the problems connected with theonomous norms. We need not even take it for granted that the notion of a super-empirical agent itself makes sense or that there are any theonomous norms. It seems to me that the idea of theonomous prescriptions is an analogical or secondary notion which is modelled on the pattern of norms which flow from human agents as their sources. We cannot therefore understand the concept of theonomous rules of human action until we have a clear understanding of the primary concept of human rules of human action.

Prescriptions the authorities of which are empirical agents may be called positive. The authority of some positive norms is a personal agent the authority of others an impersonal agent. (On the types of agent see Ch. III Sect. 4.)

The laws of the state the by-laws of a magistrate the statutes of a corporation are examples of positive prescriptions which are (normally) issued by an impersonal authority. It may be thought that impersonal authorities of positive prescriptions are but ‘logical constructions’ of human beings acting individually or collectively as norm-authorities. If this view is correct there would be ground for saying that the concept of a positive norm with an impersonal norm-authority unlike the concept of a theonomous norm is not an analogical idea.

The concept of an impersonal norm-authority is intimately connected with the concept of an office. This again is partly at least a normative notion. An office confers upon its holder certain rights and/or duties e.g. rights and/or duties to make laws and issue regulations for others. We shall not here discuss the concept of an office in detail. Some remarks bearing on the notion will be made in Chapter X.

A personal norm-authority can be either a human individual or a human collectivity. The latter case is by no means uncommon. Within a group of ‘equals’ prescriptions of the form ‘We command…’ (addressed e.g. to a member who is unwilling to participate in the work for a common end) are probably more common than prescriptions of the form ‘I command…’. When an adult human individual issues a command or gives a permission to another adult he is usually either acting in the capacity of holder of some office (e.g. as officer in the army or as policeman) or speaking in the name of a group of men. There are some reasons why this should be so. They are connected with the nature of normative activity. We shall examine them in some detail in Chapter VII.

8. The concept of a norm-authority has some bearing on the well-known division of norms into heteronomous and autonomous.

The concept of a heteronomous norm is relatively unproblematic. A prescription is heteronomous we shall say if it is given by somebody to somebody else. Heteronomous prescriptions have different authority and subject(s).

The idea of an autonomous norm is more problematic. One way of understanding the idea is to regard those norms or some of those norms as being autonomous which are not given or issued by any authority at all. Perhaps moral principles could be regarded as autonomous norms in this way. We shall not discuss the question here.

Another way of understanding the idea of autonomous norms is to call those prescriptions autonomous which are given by some agent to himself. On this view autonomous norms are self-commands self-permissions and self-prohibitions.

The question may be raised: Can an agent give prescriptions (commands permissions prohibitions) to himself? That is: Is this logically possible? It must not be taken for granted that the answer is affirmative. My view is that an agent can sometimes correctly be said to command or give permissions to himself but only in an analogical or secondary sense. The attribute ‘autonomous’ moreover does not seem to me ill-suited for such self-reflexive prescriptions.

If only norms which lack a norm-authority are called autonomous then no prescription can be autonomous. If norms with identical authority and subject are called autonomous and if such cases are possible then some prescriptions are autonomous. My view is that in a primary sense prescriptions are heteronomous. Only in a secondary sense (of ‘prescription’) are there autonomous prescriptions.

9. By the subject (or subjects) of a prescription I understand the agent (or agents) to whom the prescription is addressed or given. The subjects are commanded or permitted or forbidden by the authority to do and/or forbear certain things.

There are as many kinds of norm-subject as there are kinds of agents who are capable of human action. If acts of impersonal agents are ‘reducible’ to acts of personal agents and collective action to the action of individuals then prescriptions whose subjects are individual men hold a basic position relative to all other prescriptions. As said earlier (Ch. III Sect. 4) whether such a reduction is possible will not be investigated in this work. Here we shall consider only individual men as subjects of prescriptions.

We shall say that a prescription is particular with regard to its subject when it is addressed to one specified human individual. This is the case for example with the command addressed to N.N. to open the window.

(The prescription can also be given to several i.e. a finite number of specified subjects. This case will not receive special attention here. I shall regard it as being resolvable into a plurality of cases of the first kind mentioned.)

We shall say that a prescription is general with regard to its subjects when it is addressed either to all men unrestrictedly or to all men who satisfy a certain description.

The laws of the state to the extent that they are concerned with the conduct of individuals provide examples of prescriptions which are given to men of a certain description. The laws are made for the citizens of the state and not for all mankind. Prescriptions of the type ‘Children under 12 must not work the lift’ are also addressed to agents satisfying a certain description.

The question may be raised whether there are (can be) prescriptions which are given to all men unrestrictedly. ‘Thou shalt not kill’ ‘Never tell a lie’ ‘Love your neighbour as yourself—are not these examples of such prescriptions? We can regard the three sentences as formulations of moral principles. The principles surely ‘apply to’ or ‘concern’ all men unrestrictedly. This kind of generality may moreover be regarded as a characteristic of moral principles. But this does not show that they are prescriptions for all men without restriction. If moral principles are prescriptions (on our understanding of the term) we should be able to answer the question ‘Who gave the moral law?’ Some think that God gave the moral law. The formulation of moral principles by means of sentences in the imperative mood may be said to ‘hint at’ this conception of morality. If however we do not subscribe to this conception we cannot instance moral principles as examples of prescriptions addressed to all men unrestrictedly. Our doubts about the possibility of such prescriptions have to do with considerations of the nature of normative action and of relationships of power or strength between norm-authority and norm-subjects). We shall return to these questions in a later chapter.

Consider a captain's command to the passengers ‘Someone ought to leave the boat’. To whom is it addressed? The answer could be: to all the passengers. Yet it would not be true to say that every passenger has been ordered to leave the boat. In fact no one passenger has been commanded to do this. The passengers are ‘disjunctively’ under an obligation to do something. How shall this be understood?

The captain counts the passengers he knows that the vessel cannot carry them all safely to the destination and says to himself ‘Someone ought to leave the boat.’ If this is an order at all it is one which the captain addresses as it were to himself and not to the passengers ‘disjunctively’. The self-command as it were emerges as the conclusion of an argument: ‘If there are as many people as this on board the passage will not be safe; therefore I must see to it that someone disembarks before we travel.’ Having reached this conclusion the captain may turn to some one (or several) of the passengers and order him (or them) to leave the boat. In this case there is no obligation imposed on the passengers ‘disjunctively’. But he may also turn to all the passengers and order them to see to it that one of them leaves the boat. If this is what the captain does I shall say that the order is addressed to the passengers collectively or that the subject of the prescription is not individual men but the collectivity of passengers. This collective agent has been ordered to perform an act which results in one of the members of the collectivity leaving the boat. The passengers may for example discuss the situation among themselves and agree that the heaviest of them must leave or the one who embarked last. This decision may be construed as a prescription addressed by the collectivity to a particular one of its members.

Thus on our suggested analysis a prescription which addresses (commands permits or forbids) all agents of a certain description disjunctively is not a prescription which is general with regard to its subjects. The prescription is particular with regard to its subject this subject being a collective agent.

10. The contents of prescriptions which we here study are certain generic acts and/or forbearances. These contents the norm-subjects are commanded or permitted by the norm-authorities to realize in individual acts and/or forbearances on certain occasions.

Mention of the component which we call ‘occasion’ in the formulation of a prescription is usually mention of a location i.e. place or span in time. ‘Now’ ‘next Monday’ ‘within a week’ ‘once every second year’ ‘sometime(s)’ ‘always’ are words and phrases which may be used to make clear the occasion(s) for which the prescriptions are made (given).

A prescription which is for one specified occasion only we shall call particular with regard to the occasion. ‘Open the window now’ is an example. ‘If it starts raining shut the window immediately’ is another.

A prescription which is for a finite number of specified occasions we also call particular. This case is not of independent interest.

A prescription which is for an unlimited number of occasions we shall call general with regard to the occasion.

A prescription is conjunctively general with regard to the occasion if it orders or permits the realization of its norm-content on all (every one) of this unlimited number of occasions. ‘Shut the window whenever it starts raining’ would be an example.

That a prescription is disjunctively general with regard to the occasion will mean that it orders or permits the realization of its norm-content on some (at least one) of this unlimited number of occasions. As in the case of generality with regard to subject the question may be raised whether prescriptions can be genuinely disjunctively general with regard to the occasion.

Sometimes the temporal specification of the occasion is such that several occasions for realizing the norm-content may arise within the time specified. For example: The stranger who on arriving in a country is asked to report to the police within a week can comply with this requirement either to-day or to-morrow or… If he chooses to report to-morrow he can do so either in the morning or…

The question may be raised whether there can be an unlimited number of occasions for doing a certain act within a limited time-span such as a day or a week or a year. If the answer is affirmative it may be thought that an order or permission to do a certain thing within such a time-span is disjunctively general with regard to the occasion.

It may however also be thought that even if there can be an unlimited number of occasions for doing a certain act within a limited time-span an order or a permission to do a certain thing within such a time-span is particular with regard to the occasion. We then regard the time-span in question as one occasion. This one occasion is so to speak ‘disjunctively constituted’ of a (finite or infinite) number of occasions of a shorter duration. The conception of the time-span as one occasion is somewhat similar to the conception of a collectivity of men as one agent.

I shall accept the view that a command or permission to do a certain thing within a limited time-span is particular with regard to the occasion. An order to me to do something say within the present year is particular and not disjunctively general with regard to the occasion even if it could be truly said that there are an unlimited number of occasions within the time specified for doing the thing in question. An order to do something once every year e.g. to make an income-tax return is general with regard to the occasion—but ‘conjunctively’ and not ‘disjunctively’ general.

11. As noted in the two preceding sections considerations relating to subject and occasion lead to a classification of prescriptions as particular or general.

We shall call a prescription particular if it is particular with regard both to subject and to occasion.

‘N.N. open the window now’ enunciates a particular prescription.

We shall call a prescription general if it is general with regard to subject or to occasion (or both). If it is general with regard both to subject and to occasion we shall call it eminently general.

As noted in Section 9 a prescription which is general with regard to subject and the subjects of which are human individuals need not be addressed to all men unrestrictedly. Its subjects can be all men who satisfy a certain description e.g. that they are British citizens.

We raised some doubts as to whether there can be prescriptions which are addressed to all men unrestrictedly. These doubts be it observed do not concern the possibility of prescriptions that are eminently general. A regulation which concerns all British citizens but no others can be eminently general.

The question may be raised whether a prescription which is addressed to all men of a certain description could not be regarded as a hypothetical prescription which is addressed to all men unrestrictedly. For example: Could not a regulation which concerns all British citizens be regarded as a prescription which orders or permits all men unrestrictedly to do a certain thing if (in case) they happen to be British citizens?

No doubt we can so define ‘hypothetical prescription’ that prescriptions whose subjects are agents satisfying a certain description may be called ‘hypothetical’. But then we must distinguish between prescriptions which are hypothetical in the sense that their subjects are restricted to agents satisfying a certain description (such as ‘British citizen’) and prescriptions whose conditions of application are restricted to certain contingencies (such as ‘if it starts raining’).

We shall decide not to enlarge the scope of the term ‘hypothetical prescription’ in such a way that it covers also prescriptions whose class of subjects is restricted to agents of a certain description.

Laws of the state we have said before (Ch. I Sect. 5) are a species of prescriptions. It may be thought to be of the ‘essence’ of a law that it must be a general and cannot be a particular prescription.

If we accept the view that laws must be general we may raise the further question whether they must be eminently general or whether generality with regard to subject or with regard to occasion is sufficient.

This question recalls a difference of opinion between two famous names in jurisprudence Blackstone and Austin concerning the meaning of ‘law’ or ‘rule’. Blackstone seems to have held that a law is distinguished from a particular command by being general with regard to its subjects. A law obliges generally the members of a given community or the persons of a given class (description). Austin again saw the distinguishing feature of laws in the generality of the occasions for which they are issued or made.

On Austin's view an order may address all citizens of a given state and yet not deserve the name of a rule. ‘Suppose’ Austin says ‘the sovereign to issue an order enforced by penalties for a general mourning on occasion of a public calamity. Now though it is addressed to the community at large the order is scarcely a rule in the usual acceptation of the term. For though it obliges generally the members of the entire community it obliges to acts which it assigns specifically instead of obliging generally to acts or forbearances of a class.’1

On the other hand an order may according to Austin be given to one specified agent only and yet deserve the name of a rule because of the generality of occasions for which it is given. ‘A father may set a rule to his child or children; a guardian to his ward; a master to his slave or servant.’2 ‘If you command your servant… to rise at such and such an hour on such and such a morning… the command is occasional or particular.… But if you command him simply to rise at that hour or to rise at that hour always or to rise at that hour till further orders it may be said with propriety that you lay down a rule for the guidance of your servant's conduct.’3

Most of the laws of the state Austin thought4 are what we have here called eminently general i.e. general both with regard to subject and with regard to occasion.

As far as the term ‘rule’ is concerned it seems to me that Austin was right in thinking that generality with regard to occasion and not generality with regard to subject is the distinguishing mark of prescriptions which deserve to be called rules.5

As far as the term ‘law’ is concerned the dispute between Austin and Blackstone seems uninteresting. It could nevertheless be an interesting question of political philosophy and the philosophy of jurisprudence whether it is not of the essence of law (or of the state) that laws of the state are general. The question would then have to be related to ideas concerning the purpose of laws and the raison d&être of a state.

12. In the remaining sections of this chapter we shall discuss the various norm-characters. We begin with a discussion of the relation between the ‘ought’—character and the ‘must not’—character. Speaking of prescriptions the question concerns the relation between command and prohibition.

It is obvious that the two characters in question are interdefinable (cf. Section 2). That which ought to be done is that which must not be left undone and vice versa. That which ought to be left undone is that which must not be done and vice versa. Every positive norm of ‘ought’—character is identical with a negative norm of ‘must not’—character and conversely. Every negative norm of ‘ought’—character again is identical with a positive norm of ‘must not’—character and conversely. Speaking of prescriptions: a command to do (positive command) is a prohibition to forbear (negative prohibition) and vice versa; and a command to forbear (negative command) is a prohibition to do (positive prohibition) and vice versa.

We shall list these identities below for the eight types of elementary O-norms:

Od(pTp) says that the state of affairs described by p ought to be preserved or which means the same that one must not let it vanish.

Of(pTp) says that the state of affairs described by p must not be prevented from vanishing or which means the same that one ought to let it vanish.

Od(pT ~p) says that the state of affairs described by p ought to be destroyed or which means the same that one must not let it continue.

Of(pT ~p) says that the state of affairs described by p must not be destroyed or which means the same that one ought to let it continue.

Od( ~pTp) says that the state of affairs described by p ought to be produced or which means the same that one must not let it continue to be absent.

Of( ~pTp) says that the state of affairs described by p must not be produced or which means the same that one ought to let it continue to be absent.

Od( ~pT ~p) says that the state of affairs described by p ought to be suppressed or which means the same that one must not let it come into being.

Of( ~pT ~p) finally says that the state of affairs described by p must not be suppressed or which means the same that one ought to let it come into being.

Although ‘ought to’ and ‘must not’ are interdefinable it is convenient to retain the use of both phrases and also to retain the use of both terms ‘command’ and ‘prohibition’. But there is no point in having different symbols for the two norm-characters in our formalism (cf. Section 2).

When the content of a prescription is mixed a compound of acts and forbearances (cf. Section 4) it is usually not more natural to call the prescription a ‘command’ than to call it a ‘prohibition’. Such prescriptions can be said to be both commands and prohibitions or partly the one and partly the other. Does for example ‘Close the window or leave the door open’ enunciate a command or a prohibition? It does not matter what we call it. But it is interesting to note that the same prescription which we expressed by a sentence in the imperative mood could also have been expressed by the ‘ought’—sentence ‘You ought to close the window or leave the door open’ and by the ‘must not’—sentence ‘You must not leave the window open and close the door’.

13. We have distinguished (Section 2) between obligation-norms or O-norms and permissive norms or P-norms. When the norms concerned are prescriptions obligation-norms are broadly speaking commands or prohibitions. Permissive prescriptions are also simply called ‘permissions’.

The independent status of permissive norms is open to debate. The problems in this region are it seems more urgent to a theory of prescriptions than to a theory of other types of norm. Therefore these problems are also more relevant to legal and political philosophy than to moral philosophy.

We here limit our discussion of permissive norms to a discussion of ‘permissions’ i.e. permissive prescriptions. The main problem before us is this: Are permissions an independent category of prescriptions? Or can they be defined in terms of command and prohibition? Strictly speaking: Can prescriptions of the P-character be defined in terms of prescriptions of the O-character?

There are two ways in which it has been attempted to deny the independent status of permissions. The one is to regard permissions as nothing but the absence or non-existence of ‘corresponding’ prohibitions. The other is to regard permissions as a peculiar kind of prohibitions viz. prohibitions to interfere with an agent's freedom in a certain respect. These two views of permissions must be sharply distinguished and kept apart.

The view that a permission to do a certain thing is the same as the absence or lack of a prohibition to do this thing is common. I have accepted it myself in previous publications. It seems to me however that this view is in serious error for a variety of reasons. Here I shall state one reason only.

One cannot make an inventory of all conceivable (generic) acts. New kinds of act come into existence as the skills of man develop and the institutions and ways of life change. A man could not get drunk before it had been discovered how to distil alcohol. In a promiscuous society there is no such thing as committing adultery.

As new kinds of act originate the authorities of norms may feel a need for considering whether to order or to permit or to prohibit them to subjects. The authority or law-giver may for example consider whether the use of alcohol or tobacco should be permitted. In the case of every authority personal or impersonal there will always be a great many acts about the normative status of which he never cares.

It is therefore reasonable given an authority of norms to divide human acts into two main groups viz. acts which are and acts which are not (not yet) subject to norm by this authority. Of those acts which are subject to norm some are permitted some prohibited some commanded. Those acts which are not subject to norm are ipso facto not forbidden. If an agent does such an act the law-giver cannot accuse him of trespassing against the law. In that sense such an act can be said to be ‘permitted’.

If we accept this division of acts into two main groups—relative to a given authority of norms—and if we decide to call acts permitted simply by virtue of the fact that they are not forbidden then it becomes sensible to distinguish between two kinds of permission. These I shall call strong and weak permission respectively. An act will be said to be permitted in the weak sense if it is not forbidden; and it will be said to be permitted in the strong sense if it is not forbidden but subject to norm. Acts which are strongly permitted are thus also weakly permitted but not necessarily vice versa.

Roughly speaking an act is permitted in the strong sense if the authority has considered its normative status and decided to permit it. But this must not be understood to mean that the authority is necessarily aware of having permitted the act. The permission may also be a logical consequence of other norms which he has issued. What this means will be explained later.

Weak permission is not an independent norm-character. Weak permissions are not prescriptions or norms at all. Strong permission only is a norm-character. Whether it is an independent norm-character remains to be discussed. We return to the question in Section 16.

14. Here a few words must be said about the famous principle nullum crimen sine lege. Can this principle be quoted in support of the idea that permission consists in mere absence of prohibition?

The principle in question can I think be sensibly interpreted in two different ways neither of which however supports the view mentioned about the nature of permission.

According to the one interpretation the principle lays down a rule not about the subject's freedom to act but about the authority's right to punish. The principle under this interpretation is also often worded nulla poena sine lege.

According to the second interpretation the principle is to the effect that anything which is not forbidden within a certain normative order (system hierarchy) i.e. within a totality of prescriptions which flow from one and the same supreme authority is permitted within this order. (The concepts of normative order and supreme authority will be discussed in Ch. X.) This is not a definition of the concept of permission but a permissive norm with a peculiar content. Its content is so to speak the ‘sum total’ of all acts and forbearances which are not already forbidden.

A nullum crimen rule permitting all not-forbidden acts and forbearances may or may not occur within a given normative order. If it occurs within a normative order then relative to this order all human acts are subject to norm. Such an order without ‘gaps’ we shall call closed.6 Normative orders which are not closed will be called open.

The question may be raised whether a normative order could not also be closed by means of a ruling to the effect that anything which is not permitted within the order is forbidden?7

Consider a possible result of action such that there is no prescription in the normative order in question which permits the doing of this thing (result) nor any prescription which permits the forbearing of it. Then by the suggested ‘converse’ of the nullum crimen principle both the doing and the forbearing of this thing would be forbidden. But this—as we shall see presently—is a logical impossibility. Therefore a normative order cannot be closed by means of a norm prohibiting all not-permitted acts and forbearances unless there is in the order for any possible result of action a permission to achieve this result or to forbear achieving it. If it is thought that a complete inventory of all human acts is not possible (cf. Section 13) then this condition cannot be satisfied. And then the suggested way of closing a normative order must be rejected as absurd.

It would however be logically possible to close a normative order by means of some weakened form of the principle that anything which is not permitted is forbidden. The closing principle could for example be that any act the doing of which is not permitted is forbidden. It would then by virtue of the laws of deontic logic follow that it is also permitted to forbear anything the doing of which is not permitted. An alternative way of closing an order would be by means of a principle to the effect that any forbearance which is not permitted is forbidden. Of such an order it would hold true that it is also permitted to do anything the forbearance of which is not permitted.

Closing a normative order by means of the nullum crimen principle faces no such difficulties and is subject to no such restrictions as closing it by means of the suggested ‘converse’ principle. The reason for this is that whereas both the doing and the forbearing of one and the same thing cannot be forbidden (or commanded) without contradiction both things can perfectly well be permitted without contradiction. The closing of a normative order by means of an unrestricted nullum crimen principle moreover does not presuppose a complete inventory of human acts.

15. It seems possible to distinguish between various kinds of strong permission—permissions as it were of increasing degree of strength.

In permitting an act the authority may only be declaring that he is going to tolerate it. The authority ‘does not care’ whether the subject does the act or not. The authority is determined not to interfere with the subject's behaviour as far as this act is concerned but he does not undertake to protect the subject from possible interferences with his behaviour on the part of other agents.

Any (strong) permission is at least a toleration but it may be more than this. If a permission to do something is combined with a prohibition to hinder or prevent the holder of the permission from doing the permitted thing then we shall say that the subject of the permissive norm has a right relatively to the subjects of the prohibition. In granting a right to some subjects the authority declares his toleration of a certain act (or forbearance) and his intolerance of certain other acts.

To prevent an agent from doing (or forbearing) a certain thing is to act in a way which makes the doing (or forbearing) of this thing impossible to this agent. Preventing from forbearing is also called compelling or forcing to do.

We ought to distinguish between not making an act impossible (for someone to perform) and making an act possible. The second is also called enabling (someone to do something). It is the stronger notion. Enabling entails not-hindering but not-hindering does not necessarily amount to enabling.

If a permission to do something is combined with a command to enable the holder of the permission to do the permitted thing then we shall say that the subject of the permissive norm has a claim relatively to the subjects of the command. It is understood that any claim in this sense is also a right but not conversely.

Assume that it is part of a country's constitution that every citizen has a ‘right to work’. Assume that no employer has a job for Mr. X. They cannot be accused of hindering Mr. X from taking up an employment. But Mr. X nevertheless cannot exercise his right. It may then be argued that the constitutionally granted right is ‘empty’ if it is not a right in the stronger sense of a claim. The claim could for example be instituted in the form of a command addressed to all employers disjunctively (or collectively cf. the discussion in Section 9) to take care that a job is provided for anyone who wants to work.

For the sake of avoiding misunderstanding it must be mentioned that what is here being said about rights and claims does not pretend to be a complete analysis of the notions of legal rights and legal claims. We use ‘right’ and ‘claim’ in a technical sense adapted to our purposes. I think however that our concepts of rights and claims are relevantly related to the legal notions; that the former catch hold of essential though not exhaustive logical features of the latter.

Rights and claims unlike tolerations are not concerned with the individual permission-holder only but also with the normative status of his relations to his fellow creatures. Rights and claims are thus social in a sense in which mere toleration is not.

It may be suggested that it is inherent in the nature of permissions to entail rights and/or claims.8 If this is accepted we should have to say that ‘mere’ toleration does not yet amount to a ‘full’ permission. It may also be suggested that the only sense in which laws of the state are permissive is by prohibiting interference with the behaviour of agents in certain respects.9

16. We shall now return to the question raised in Section 13 whether the norm-character of permission can be defined in terms of the (mutually interdefinable) characters of prohibition and/or command.

We have in the previous section seen that the specific characteristics of the two species of (strong) permission which we called rights and claims can be accounted for in terms of prohibition and/or command. It follows from this that if there is an element in permissions which is not reducible to the other norm-characters this element is identical with what we called toleration. Thus what is characteristically ‘permissive’ about permissions would be the norm-authority's declaration of his toleration of a certain behaviour on the part of the norm-subject(s). ‘Permissions are essentially tolerations’ we could say.

In order to see whether permission is an irreducible norm-character or not we must thus examine the notion of toleration. It seems to me that a declaration of toleration can be understood in two different ways:

A declaration of toleration can be a declaration of intention on the part of the permission-giver not to interfere with the permission-holder's freedom in a certain respect. Or it can be a promise of not-interference.

It may now be suggested that a declaration of intention is not a normative concept at all whereas a promise obviously is. If this is accepted permissions as ‘mere’ declarations of intention not to interfere would not be regarded as norms at all. Only permissions as promises of not-interference would be norms. Then the question whether permission is definable in terms of the other norm-characters would be reducible to the question whether the normative character of a promise (or at least of a promise of non-interference) can be accounted for in terms of ‘ought’ and ‘must not’.

That the answer to the last question is affirmative would probably be universally conceded. I shall myself accept the view that permission as a species of promise can be defined in terms of the other norm-characters. If when permitting something to somebody the norm-authority has promised not to interfere with the norm-subject's freedom in a certain respect then the authority must not interfere with this freedom.

If however permissions which are tolerations are regarded as a peculiar kind of promise the question will arise whether such permissions are norms of the kind which we have here called ‘prescriptions’. Prescriptions we have said require an authority and a subject; they are someone's prescriptions to someone. The mere fact that permissions are given by someone to someone does not ensure however that permissions are prescriptions. For if the normative element in the permission is a prohibition to interfere then although the norm-authority of the permissive norm is the giver of the permission the norm-subject is not the receiver of the permission. The norm-subject is the receiver of the prohibition. This is the norm-authority itself.

Thus on the view of permissions as promises permissions would be self-reflexive prescriptions viz. self-prohibitions. But the question whether there are (can be) self-reflexive prescriptions is open to debate (cf. Section 8). If we think that such prescriptions cannot exist we should have to conclude that permissions are not prescriptions.

What kind of norm shall we then say that permissions are? That is: What kind of norm is it that says that promises ought to be kept or that this or that ought to be done on account of its having been promised? That promises ought to be kept would ordinarily be thought of as a typically moral norm and the obligation to do this or that because one has promised to do it would be called a moral obligation. The status of moral norms is problematic (see Ch. I Sect. 8). Some think they are a kind of prescription viz. the commands and prohibitions of God to men. (On this view if permissions are a sort of promise it is God who has prohibited that the givers of permissions subsequently interfere with the receiver's freedom.) Others think of moral norms as related to technical norms concerning means to ends.

Thus the conception of permissions as promises would give to permissions a peculiar moral flavour.

The conception of declarations of toleration as promises can be said to supplement the conception of rights as prohibitions to a third party. In granting somebody a right the norm-authority issues a prohibition to any third party to interfere with the right-holder's freedom in a certain respect (see Section 15). By at the same time promising to respect (tolerate) this freedom he as it were prohibits himself too to interfere with it.

On the question whether permission is or is not an independent norm-character I shall not here take a definite stand. The view that (all) permission is mere absence of prohibition I reject. The view that it can become defined in terms of prohibitions of noninterference with a person's freedom in a certain respect I find attractive. But I do not know exactly which form this view should take and how certain objections to it should be met.

In the Logic of Norms which we are here going to develop we shall therefore retain permission as an independent norm-character.

• 1. The Province of Jurisprudence Determined (1832), Lecture One, in the section called ‘Laws or rules distinguished from commands which are occasional or particular’.
• 2. Op. cit.
• 3. Ibid.
• 4. Ibid.
• 5. The use of ‘rule’ to cover a class of prescriptions must be distinguished from other uses of the word to cover norms which are not ‘prescriptions’ in our sense of the term. Cf. Ch. I, Sect. 4.
• 6. Kelsen seems to me to be in error when he argues that the legal order cannot have any gaps. See, e.g., his General Theory of Law and State (1949), pp. 146–149. Kelsen's argument hinges on the assumption that anything which the law does not prohibit it ipso facto leaves the citizens free to do. There is an interesting discussion of these problems from the standpoint both of deontic logic and of legal theory in the work by Amedeo G. Conte, Saggio sulla completezza degli ordinamenti giuridici (1962).
• 7. Aristotle seems to have had this possibility in mind in Ethica Nicomachea 1138^a 6–8.
• 8. This position has been argued by Professor K. E. Tranøy. See his paper ‘An Important Aspect of Humanism’ in Theoria, 23, 1957.
• 9. This appears to be the position, e.g., of Kelsen, op. cit., p. 77: ‘The legal order gives somebody a permission, confers on somebody a right, only by imposing a duty upon somebody else.’