Effects of international norms: A typology

The constructivist research programme on international norms has demonstrated convincingly that, how, and why norms matter. Norms have been shown to constitute the identity of actors, to guide their behaviour into desired directions, and, altogether, to generate the normative basis of the international system. In the course of this intensive debate, its main concepts, such as the question of what constitutes a norm or different norm types, became fuzzy. Also, while the focus on the intended effects of norms certainly encompasses an essential part of the phenomenon, their unintended effects have been largely neglected. Motivated by these shortcomings, the article presents a new systematisation of effects of norms. The typology developed here discerns two types of intended effects, namely prohibitive and obligative effects, as well as two corresponding types of unintended effects, namely permissive and omissive effects.


Introduction
The constructivist research programme in International Relations (IR) has successfully demonstrated the efficacy of social norms so that 'the once controversial statement that norms matter is accepted by all except the most diehard neorealists' (Checkel, 1997: 473).Yet, the research field is characterised by terminological and conceptual fuzziness that extends both to different types of norms and to their different effects.Are permissions also norms, next to proscriptions and prescriptions?Are norms themselves either constitutive or regulative or does this distinction only apply to their effects?How can we study such effects -and what other effects of norms are there, but marginal to our research?Far from being a purely theoretical exercise, these questions matter in the application: Answers to them focus and guide our empirical research, they define what phenomena we study and how we do so -and they may limit our view regarding not only suitable objects of study, but also potential explanations and conclusions.
With this article, hence, I pursue two aims: (1) In a sorting endeavour, I re-organise and advance some key parts of the conceptual framework of norms research by clarifying and defining central terms as well as their relation to each other.(2) As a result of these considerations, I develop a new typology of norm effects that is based on the distinction between proscriptions and prescriptions as two norm categories and on the distinction between intended and unintended effects as two effect categories.A cross tabulation of these properties yields two types of intended effects, namely prohibitive and obligative effects, as well as two corresponding types of unintended effects, namely permissive and omissive effects.These types of unintended effects are inverse 1 interaction effects, which need to be seen in relation to their intended counterparts and which emerge between different norms.
Building on Tannenwald (1999: 437;2007: 46), the article conceptualises these effect types by using the distinction between constitutive, regulative, and attentional effects.Accordingly, prohibitions produce the following intended -prohibitive -effects: They direct the attention towards certain undesired behaviours and stigmatise these behaviours to motivate the actors to refrain from them.As inverse effects of these, prohibitions also produce unintended -permissive -effects: First, they divert the attention from other undesired behaviours (attentional effects); second, due to the strong stigmatisation of the prohibited behaviours, these other behaviours appear more tolerable (constitutive effects), which is why they, third, continue to persist or even become more frequent by working as (functional or psychological) substitutes for the prohibited behaviours (regulative effects) (DiNardo and Lemieux, 1992;Jervis, 1997: 74).We can conceive of obligative and omissive effects, arising from obligations, analogously: Prescriptive norms direct the attention towards certain desired behaviours and stigmatise their omission to motivate the actors to resort to them.In the shadow of such norms, the omission of other desired behaviours goes unnoticed and remains tolerable so that the actors can continue to not act in a certain way.
The primary function of the proposed typology is descriptive.By identifying and characterising different types of norms and of norm effects, the typology can be applied to empirical research in different ways.It enables us to search purposefully for certain observations, to recognise certain empirical phenomena as instances of an overarching concept and to systematically analyse -describe, map, and compare -them (Elman, 2005).What type of norm are we dealing with?What are its different effects?Are certain types of norms more common in certain policy fields than in others?Do certain types of norms tend to produce more unintended effects than others?
While the typology itself is not explanatory in the sense that the types of effects it contains would serve as predictors of certain outcomes (Raymond, 1997: 226), it can, nevertheless, fulfil some explanatory functions as well (Elman, 2005: 297).First, if the identified norm effects are treated as dependent variables, we may seek to explain variances in their occurrence and prevalence.What makes some norms more likely than others to produce permissive and omissive effects?Why and when do actors opt for a specific type, that is, why, and when do they choose positive obligations (prescriptions) or negative obligations (proscriptions)?Are regulative effects of proscriptions stronger than regulative effects of prescriptions and if so, why?Why are certain behaviours attracting more normative attention than others?Second, if the identified norm effects are treated as independent variables, we may use them to explain certain outcomes.Do permissive and omissive effects of some norms inhibit the emergence, the diffusion, and the enforcement of other norms -and if yes, how?Does deficient compliance with some norms result from regulative effects of other norms?
By generating questions like these and offering conceptual tools to address them, the article highlights and integrates two angles recently emerging in norm research.So far, norms scholarship has mainly concentrated on intended effects of norms, but the first rather new trend concerns the acknowledgement and study of the other side of the coin, namely the unintended, and even contradictory effects that norms yield quasi-unavoidably (D'Ambruoso, 2015;Tannenwald, 1999: 46).The second, even newer, trend concerns the acknowledgement and study of normative complexity (Fehl, 2023).From this perspective, norms are conceived of as elements of systems where changes in one part will have implications on other parts -the focus, thus, shifts from single norms to relationships between multiple norms. 2 This research strand, so far, has been predominantly concerned with substantive conflicts between norms, for example with situations in which one norm requires a behaviour that another norm discourages or where compliance with one norm comes with trade-offs for the other. 3In contrast, by emphasising the unintended legitimisation of certain behaviours through the stigmatisation of others, this article points to substantive interactions between non-conflicting norms.At the same time, by emphasising the competition for attention between norms, it points to non-substantive, that is resource-related conflicts between norms.
The article is organised as follows.The next section conceptualises the typology's two generative properties -the types of norms and the intendedness of effects -and describes how they translate into the four effect types.In the subsequent section, I use the distinction between constitutive, regulative, and attentional effects of norms to specify in detail what prohibitive, permissive, obligative, and omissive effects are.After that, I connect existing work on norms with the conceptual framework developed in this article and suggest an agenda for further research of theoretical questions and empirical applications arising from and enabled through my framework.The conclusion summarises the main insights the article seeks to provide and points to some normative implications of the research topic of unintended effects of norms.

A typology of norm effects
While several types of norm effects have been circulating in the literature on norms, a systematic, unified distinction of these effects does not exist.The most common distinction is between constitutive and regulative effects of norms (Glanville, 2016: 185;Risse, 2000: 5).Sporadically, other effects of norms, for example instrumental or permissive, are mentioned as well (Tannenwald, 1999: 437).Yet, in the absence of a typology of norm effects, it is not quite clear how these types of effects relate to each other: What exactly defines the different types?Are they alternative effects, produced by some norms, but not by others?Are they located at the same level or are they rather first-and second-order effects?Addressing these questions, the typology developed in the following suggests one possible systematisation that reorganises and fills with substance the conceptual space of norm effects.
The typology presented as a two-dimensional property space in Table 1 is constructed as follows.Its two generative properties are (a) types of norms and (b) the intendedness of norm effects.Regarding (a), the typology distinguishes proscriptions and prescriptions.Regarding (b), it distinguishes intended and unintended effects.Through crosstabulation, the typology yields four types of effects, namely prohibitive, permissive, obligative, and omissive effects.Hence, prohibitive effects are the intended effects of proscriptions, while permissive effects are the unintended effects of proscriptions.Similarly, obligative effects are the intended effects of prescriptions, while omissive effects are the unintended effects of prescriptions.

Types of norms: Proscriptions and prescriptions
Various norm types are being formed along different criteria in the research on international norms -these types, however, represent ad-hoc distinctions following specific research interests rather than adding up into one overarching classificatory scheme (Raymond, 1997: 226).The distinction between legal and social norms, for example, is made along the norms' degree of bindingness and formality (Búzás, 2018).The distinction between substantive and procedural norms is made along their objects of reference -with the former denoting behavioural provisions that apply to specific policy fields, and the latter denoting rules that structure decision-seeking processes (Hurrell and Macdonald, 2013: 69-70).To account for varying degrees of abstractness, further subtypes of substantive and procedural norms may be added, such as fundamental norms that entail stronger moral claims while being more general and more significant for the normative system as such (Lang et al., 2006: 287;Wiener, 2014: 36), or such as metagovernance norms (Pantzerhielm et al., 2022: 595) that, in contrast to simple rules of procedure, entail more general principles referring to the 'governance of governance' (Jessop, 2014;quoted in Pantzerhielm et al., 2022: 595).Furthermore, based on the underlying problem structures and problem characteristics, cooperative and moral norms or liberal and non-liberal norms are distinguished (Goertz and Diehl, 1992;Schneiker, 2021: 108).
The typology proposed in this article uses the kind of normative command -the injunction -conveyed by the respective norm to distinguish proscriptions and prescriptions as two norm types.Proscriptions and prescriptions are widely referred to in norm research (Raymond, 1997: 225), but while some authors consider them as the only two relevant injunctions qualifying as norms (see, e.g.Jepperson et al., 1996: 54;Raymond, 1997: 225;Tannenwald, 1999: 436), others consider permissions as a third type next to them (see, e.g.Finnemore, 1996: 181;Goertz, 2003: 39).Since in IR, these different understandings simply co-exist without being debated, it is useful to consult another strand of literature on this matter, namely the deontic school -a legal philosophical tradition that is explicitly concerned with imperative statements and the oughtness of acts (Føllesdahl and Hilpinen, 1970: 1;Von Wright, 1951: 1-5).
Strikingly, the deontic school has been rarely referred to in norms research (but see Goertz, 2003: 39), even though it is closely linked to norms, especially to their central qualities of oughtness and action-guiding (Jurkovich, 2020: 694;Lutz and Sikkink, 2000: 655).Resonating neatly with those norm scholars who include permission in their notion of norms, in deontics, three different imperative modes are distinguished: prohibition, obligation, and permission.Prohibitions convey claims about impossible actions, obligations about necessary actions, and permissions about possible actions (Føllesdahl and Hilpinen, 1970: 8;Von Wright, 1951: 1-2).And yet, when it comes to the question whether all three imperative modes are norms, here too, we encounter a 'problem about permission' arises (Alchourrón and Bulygin, 1981;Lewis, 1979: 95).
To understand why permissions are not norms, we need to look at the three imperative modes -prescriptions, proscriptions, and permissions -in more detail.These modes do not equally fulfil the shared definition of norms as 'collective expectations about proper behaviour for a given identity' (Jepperson et al., 1996).While prescriptions and proscriptions indeed convey expectations, namely that action X is done (required acts of commission), and action Y is not done (required acts of omission), permissions convey much weaker expectations, if they convey expectations at all: Action Z may or may not be done.Hence, unlike prescriptions and proscriptions, which express an unequivocal imperative provision, permissions are indifferent or neutral as the choice between acting and not acting in a certain way is left to the actor, and both opposite possibilities are compatible with permissions (Von Wright, 1951: 3).Hence, if permissions do not imply commands, following them is optional and not obligatory, which means that they lack the action-guiding quality that is the very essence of norms (Alchourrón and Bulygin, 1981: 116;Jurkovich, 2020: 699).
Taking the imperative aspect seriously helps to sort out one related conflation, namely between norms and practices.The question of whether norms and practices are the same -or not -arises not only since the practice turn in IR (Jackson et al., 2015), in the course of which theorists have stressed the inherent normativity of practices (Pratt, 2020: 65).Already earlier, classical norm scholars have occasionally used the terms 'practices' and 'old norms' interchangeably, to denote 'old standards of behaviour' such as colonialism, slavery and slave trade, or the use of anti-personnel landmines, which were then replaced by 'new norms' (Finnemore and Sikkink, 1998: 892;Price, 2004: 110).For the sake of analytical precision, however, it is useful to keep norms and practices distinct from one another and to use the terms more consciously (like, e.g.Bailey (2008) does when referring to whaling only as a practice -not as a norm -that has been changed by new norms against anti-commercial whaling).
But even if we agree that norms and practices are distinct phenomena and practices are not norms (or normative) per se, 4 how do we know when to speak of which -in particular, as practices indeed may give rise to norms (Bode and Huelss, 2018: 395)?The imperative modes just introduced provide an orientation for the decision whether to refer to norms or to practices, or, to put it differently, when practices qualify as norms.The necessary criterion are, once again, the expectations linked to the practices in question, i. e. their oughtness: Is the practice just accepted, tolerated, and permitted, or is it expected?If the practice is expected, is this just an expectation of a non-normative repetition of certain behaviours -or is it accompanied by normative intentions or producing normative consequences (Möllers, 2020: 71-105)?Deviations from practices with a normative character would be met with some kind of negative social sanctions whereas deviations from non-normative practices would not (Cohen, 1980: 131).
Accordingly, some practices like those just listed may have been permitted, accepted, wide-spread, and habitual in certain contexts, but they were not normative, i.e. expected in the sense that those deviating from them had to fear sanctions.You could plant landmines (or hunt whales for commercial purposes), but you did not have to.It is considered immoral and inappropriate to use landmines or to commercially hunt whales now that doing so is prohibited, but it was not considered immoral and inappropriate not to use landmines or not to hunt whales when doing so was permitted (similarly: Jurkovich, 2020: 695-696).If it had been considered immoral and inappropriate, though, planting landmines and hunting whales would have ceased to be permissions and turned into prescriptions -and thus, into norms.
Humanitarian intervention and the responsibility to protect, its successor concept, are further examples illustrating this argument.Martha Finnemore, in her seminal contribution on humanitarian intervention, rightly attests that the then-emerging norms of humanitarian intervention allowed certain actions without requiring them -these norms were 'permissive norms only' (Finnemore, 1996: 181).But from the point of view of the deontic logic, humanitarian intervention did not constitute a norm, but rather a permission -an exception to the prohibition of the use of force.Redefining humanitarian intervention as a responsibility, both of individual states to protect their own population and of the international community to protect all populations, was precisely the attempt to foster the 'emergence of a new norm' (Coleman, 2013: 177).The previous permission (to intervene) was supposed to become a prescriptive norm through increasing its actionguiding, imperative quality: Even though the term 'responsibility' certainly implies a lower level of obligation than the term 'duty', it must be seen as an attempt to express what the actors are expected to do in specific situations, whereas the humanitarian intervention framework rather expressed what the authors were allowed to or even had the right to do (Staunton, 2018: 372-375). 5

Intendedness of norm effects
To form the typology, I use the intendedness of norm effects as the second category in addition to norm types.This section, hence, discusses whether norms can have intentions, and whether and how it is possible to discern their intended from their unintended effects.
Applying the concept of intentionality to norms is not trivial, as intentionality is a concept strongly attached to purposefully acting actors guided by preferences and objectives, and to agency (Schelling, 1978: 17).But can norms have agency?Some authors are highly sceptical of ascribing agency to norms -in their understanding, agency in the context of norms always refers to the agency of the actors, that is, their ability to choose behaviours that might run counter to the normative structure (Bucher, 2014).Moreover, norms are collective products, created by several actors with potentially diverging or hidden intentions.Is the intention of a norm an aggregated intersection of the actors' intentions, or may it diverge from them, which would mean that a norm could acquire a certain autonomous intention?Also, the fact that many norms are taken for granted and being observed almost automatically means that their initial intention is no longer present in the public conscience or no longer relevant (Finnemore and Sikkink, 1998: 904).
These valid concerns notwithstanding, the problem with the intentionality of norms is less tricky than the conceptual discussion suggests, and it can be addressed both theoretically and empirically.Theoretically, I agree that norms themselves do not have intentions as they lack agency.Yet, according to the functionalist understanding of norms, which I share, they do have a purpose (albeit a purpose that may change over time): Norms are supposed to solve certain problems through providing behavioural guidance -accordingly, their primary and obvious purpose is problem solution (Winston, 2018: 640).The legal distinction between the 'letter' and the 'spirit' of law expresses the notion of purposeful norms.The letter of a law refers to its literal, formal meaning, while the spirit of a law refers to its 'general meaning or purpose' -or 'the intention of the law', which is the perceived intention of the lawmakers (Garcia et al., 2014: 480).Thus, we can consider the purposes of norms to be equivalent to their intentions.
Empirically, the intention of norms can be inferred from different sources.As many international norms emerge as a result of deliberate norm-setting initiated by so-called norm entrepreneurs (Finnemore and Sikkink, 1998: 895-896), the discourses preceding norm adoption -observable in campaigns and in negotiations -can reveal (a) how these actors perceived a specific problem, (b) how they understood the norm in question to contribute to the problem's solution, (c) the broader objectives that the actors hoped to achieve with their efforts, as well as (d) the 'spirit of agreements, that is, the unwritten aspect of written agreements' (Cohen, 1980: 132).Once the norms have been adopted, the shared understanding of the international community regarding what those norms are supposed to accomplish can be expected to materialise explicitly in the preambular part of international conventions, resolutions, or declarations.Further evidence can be drawn from the discourses following norm adoption, such as legal interpretations. 6 In addition to their intended purposes, norms can also produce other -unintended -effects.Such effects denote outcomes occurring next to or instead of the outcomes that were originally intended; to put it differently, they denote a discrepancy between the intention that guided certain actions, and some results of those actions (Baert, 1991: 201).Some authors equate unintended effects with unanticipated and undesirable effects (Merton, 1936).In my understanding, however, unintended effects are the overarching category of effects of interest here, and unanticipated effects as well as undesirable effects are its sub-categories.This means that all unanticipated effectsthe effects that were not foreseen when the action was taken -are unintended, but only some unintended effects are unanticipated while others may well be anticipated.Similarly, all undesired effects -the effects deemed negative from a certain normative standpoint -are unintended, but only some unintended effects are undesirable while others may well be desirable.
Unintended effects of norms can be observed in different fields of international politics.Examples include unintended effects of environmental regimes like the observation that the hunting ban on great whales has led to an increased hunting of other whales (Young, 1994: 151, 146).Enforced disappearances can be considered an unintended effect of the increasing legalisation of other human rights norms such as torture, breaches of which disappearances are a way to conceal (Lutz and Sikkink, 2000: 637).International sanctions punishing norm violations have well-known unintended effects such as harming civilian populations.
To sum up, I understand as the intention of a norm its declared purpose, which mostly will be the aspired solution of a particular problem.Accordingly, I understand all effects that a norm produces in addition to this purpose (be they anticipated or not, or desirable or not) as unintended.The typology proposed in this article focuses on particular types of unintended effects, namely on the effects described as 'shadow effects' by Tannenwald (1999: 437): inverse effects of the intended effects that emerge in interactions between different norms.Accordingly, permissive effects are the unintended counterpart to the intended prohibitive effects of proscriptions, and omissive effects are the unintended counterpart to the intended obligative effects of prescriptions (see Table 1 above).The next section conceptualises these effect types.

Conceptualising intended and unintended norm effects
What are these prohibitive, permissive, obligative, and omissive effects and how are they created?How do they relate to constitutive and regulative effects?This section addresses these questions in two steps.First, I discuss the common distinction between constitutive and regulative effects and add attentional effects as a third type of effects located at the same level.Second, I use this distinction to conceptualise the intended as well as the unintended effects of prohibitions and of obligations.

Constitutive, regulative, and attentional effects of norms
Focusing on constitutive and regulative effects, in the following, I make four arguments: First, there are no constitutive or regulative norms, but only constitutive and regulative effects of norms.Second, constitutive and regulative effects are not the corresponding products of the logic of appropriateness and the logic of consequences.Third, we need to add attentional effects as a third type next to constitutive and regulative effects.Fourth, while all norms produce all three kinds of effects, these effects do not (necessarily) occur simultaneously but are more likely to occur sequentially and matter to a different degree in different phases of norm evolution.Do the attributes 'constitutive' and 'regulative' refer to norms or rather to effects of norms?For authors following John R. Searle, they refer to norms.Norms, in this view, can be either constitutive, understood as defining meanings, or regulative, understood as defining what actions are allowed or prohibited (Finnemore and Sikkink, 1998: 891;Searle, 1995: 43-50).In contrast, for authors following Nicholas Onuf, 'constitutive' and 'regulative' refer to effects of norms.They dispute that there are two kinds of norms (some that constitute and some that regulate), but stress that all norms always both constitute and regulate.Accordingly, 'constitutive' and 'regulative' are different, but interdependent and co-occurring, categories of effects of norms (Onuf, 1998: 68;Risse, 2000: 5).While the constitutive function is central to norms, norms cannot be merely constitutive, since the definition of norms as collective behavioural expectations necessarily implies a regulative function (Jurkovich, 2020: 696-697;Winston, 2018: 640-641).Hence, while norms cannot be classified as either constitutive or regulative, it is yet possible and useful to differentiate between their constitutive or regulative effects.
Constitutive effects frequently denote the impact of norms on identities, but the definitional, meaning producing capacities of norms extend to other reference objects such as actions and objects.Norms not only determine which actions are allowed or prohibited; in a more fundamental way, they also determine what counts as a certain action, or, in other words, in which context a certain action acquires a certain meaning (Laffey and Weldes, 1997: 210).Similarly, norms not only determine to which categories certain objects belong, but they also create the very categories by naming them and attaching attributes, values, and meanings to them.Moreover, norms define and qualify who the actors are -by defining their identities, they are constitutive of the actors (Goertz, 2003: 40;Jepperson et al., 1996: 54).Since obeying certain norms is necessary both to assure oneself of one's own identity and to be recognised as a certain kind of actor by relevant others, constitutive effects of norms also encompass inclusionary and exclusionary as well as hierarchising functions (Hausteiner, 2020: 54;Towns, 2012: 180, 188).
Regulative effects encompass the impact of norms on the behaviour of actors.Norms produce these effects through specifying the kind of behaviour expected from the actors in certain situations, and through increasing the social and material costs of alternative choices (Cortell and Davis Jr, 2000: 69).By doing so, norms both instruct the actors how to behave and serve as a yard stick for evaluating the behaviour of others (Kratochwil, 1984: 686).Yet, as social phenomena addressing actors endowed with agency, norms cannot determine behaviour -even though they make conformity more likely, it cannot be taken for granted, and deviations always remain possible (Kratochwil, 1984: 705).Nevertheless, behavioural conformity with either prohibitive or prescriptive injunctions is what we can define as the intended regulative effect of norms (Raymond, 1997: 225).
A case in point to illustrate the arguments concerning the constitutive-regulative distinction is the concept of sovereignty, which is frequently referred to as a (constitutive) norm (Risse, 2000: 5).Since sovereignty in the first place defines -constitutes -what the states are and what rights they have, it is, strictly speaking, rather a property than a norm.But while not being a norm itself, sovereignty gives rise to 'collections of norms' (Finnemore and Sikkink, 1998: 891).Traditionally, these norms were directed towards other states, such as the obligation to respect the territorial integrity or the prohibition to interfere in internal affairs (Schneiker, 2021: 108).In the last decades, however, the meaning of sovereignty has changed: Sovereignty norms are now being increasingly understood to be directed at the states themselves, conditioning their sovereignty claims on obligations like respecting human rights and protecting their citizens (MacFarlane et al., 2004: 978).Sovereignty, hence, at the same time has constitutive effects in that it defines and enables, and regulative effects in that it restricts and obliges (Großklaus, 2017: 267).
How do constitutive and regulative effects correspond to the two types of logics -the logic of appropriateness (LoA) and the logic of consequences (LoC) -that depict two ideal-typical primary motives for compliance and have become another central point of reference in the discussion about norms?The LoA, based on the concept of homo sociologicus, assumes normative and intrinsic motives; the LoC, based on the concept of homo oeconomicus, assumes rational and instrumental motives (March and Olsen, 1998).For some authors, intrinsically motivated compliance is a constitutive effect of norms and instrumentally motivated compliance is a regulative effect (Glanville, 2016: 186-187).According to the definition just introduced, however, compliance -regardless of its motive -always is a regulative effect whereas the two logics are two different, but equifinal paths generating this effect.On those paths, constitutive effects play different roles.On the LoA pathway, constitutive effects on actors' identities precede regulative effects -the actors adhere to the norm because they have internalised it as part of their identity and believe in its oughtness.On the LoC pathway, constitutive effects such as defining what is the right kind of behaviour and what counts as this behaviour occur too -constitutive effects on actors' identities, however, may be lacking.While the actors still adhere to the norm, this is a result of a positive cost-benefit calculation involving the anticipated costs of sanctions, but neither does the norm have to become part of their identities nor do the actors have to believe in the norm's oughtness.
The third argument of this section is that, in addition to constitutive and regulative effects, attentional effects need to be established as a third type of norm effects.Such effects, while hitherto lacking a label and a conceptual position, have been nevertheless implicitly present in the literature on norms quite early, namely through the concepts of 'awareness raising' and 'agenda setting' (Keck and Sikkink, 1998: 22-25;Risse, 2000: 20).But while these concepts rather denote strategies of norm entrepreneurs, I focus on the corresponding effects -attentional effects -and thereby draw on Nina Tannenwald who was the first to explicitly point out that norms affect the allocation of attention (Tannenwald, 2007: 46-47).Such allocation processes happen both in actors and in systems, and may encompass different attention resources, namely cognition, communication, space, and time (Newig, 2004: 153-155).
Finally, my fourth argument is that attentional, constitutive, and regulative effects of norms may occur sequentially.This implies that different types of effects may matter more or less in different stages of norm evolution.Concretely, attentional and (some) constitutive effects might be expected while a norm is emerging, while regulative effects might be expected afterwards.A problem must attract attention, the behaviours causing the problem must be defined as impossible and the behaviours mitigating it must be defined as necessary before a respective norm can emerge, trigger behavioural changes and ideally, become a part of the actor's identity.Consequently, we may expect attentional effects particularly from emerging norms, and not, as Tannenwald suggested, from strong and established norms (Tannenwald, 2007: 47).While for emerging norms, public and political attention is crucial, strong norms, by contrast, are internalised to such a degree that they tend to vanish from the public debate (Finnemore and Sikkink, 1998: 895).At the individual level too, the attentional effects of internalised norms decrease, because norms function as 'cognitive energy-savers' by fostering habitual behaviour that replaces constant decision-making on which option to choose (Florini, 1996: 366).
To be sure, these processes are less neat and linear than they appear theoretically.Constitutive effects might not only follow, but also precede attentional effects: Once a problem has attracted the attention, problem definition processes set in -vice versa, however, a particular problem definition and particular actors' identities have an impact on which problems attract and sustain attention (Keck and Sikkink, 1998: 18-22).Regulative effects can create feedback loops: Norm-conforming behaviour has constitutive effects as it reinforces certain identity conceptions, but it will probably not have attentional effects as compliance is what is expected and thus not particularly noteworthy (Finnemore and Sikkink, 1998: 892).In contrast, norm-violating and norm-contesting behaviour can have attentional and constitutive effects.Not only will it put the problems associated with those behaviours into focus, but it will also call into question the violators' identities and trigger constitutive discussions about the norm's substance that may change the norm's meaning, weaken, or strengthen it (Deitelhoff and Zimmermann, 2020;Sandholtz, 2008).

Prohibitive, permissive, obligative and omissive effects of norms
By distinguishing between two types of norms, namely proscriptions and prescriptions, and between intended and unintended effects, the typology results in four types of norm effects: Prohibitive effects, which are the intended effects of proscriptions; permissive effects, which are their unintended effects; obligative effects, which are the intended effects of prescriptions, and omissive effects, which are their unintended effects.The previous section identified three effects produced by all norms: attentional effects, encompassing the distribution of attention between certain behaviours; constitutive effects, encompassing (substantive and normative) definitions of certain behaviours; and regulative effects, encompassing guidance for and motivation of certain behaviours.I now integrate the discussion of attentional, constitutive, and regulative effects into my concept to specify in detail what prohibitive, permissive, obligative, and omissive effects are.Furthermore, I illustrate the different types through examples, and clarify how my conceptualisation relates to Nina Tannenwald's concept of permissive effects.
Brought together, the different distinctions introduced so far help capturing the nature of different types of effects, as summarised in Table 2.The intended -prohibitiveeffects of proscriptive norms are composed of attentional effects directing our attention towards certain undesired behaviours, constitutive effects defining those behaviours as stigmatised and prohibited, and regulative effects motivating the actors to refrain from them.The unintended -permissive -effects arise as a corollary to all three subtypes.When particular attention is paid to certain behaviours, it may be diverted from others.When those behaviours are particularly stigmatised and prohibited, other behaviours may appear relatively tolerable.When the stigmatised behaviours cease, the tolerated behaviours may continue.
The obligative and omissive effects are defined analogously.The intendedobligative -effects of prescriptive norms are composed of attentional effects directing our attention towards certain desired behaviours, constitutive effects defining those behaviours as appropriate and prescribed, and regulative effects motivating the actors to resort to them.Consequently, the unintended -omissive -effects divert the attention from other desirable behaviours, refraining from which remains tolerable, which means that these behaviours continue not to be enacted.
How prohibitive and permissive effects co-occur and interact can be observed well in the case of international norms protecting civilians and combatants in armed conflict.The constitutive effects of these norms lie in the dichotomous classification of all persons as either civilians or combatants, a key element of the International Humanitarian Law (IHL).The most important regulative effect arising from it is the prohibition to target civilians -and the corresponding permission to do so with combatants.This norm, while falling far short of eliminating civilian casualties, has nevertheless motivated conflict parties to prioritise civilian protection and to adjust their policies in order to comply with the norm (Kahl, 2007).In contrast, combatants are not only legal targets of attack, but these attacks can be legally carried out with weapons commonly considered inhumane such as incendiary weapons (Rosert, 2019: 54-85).The attentional effects in this area have been strong too: The protection of civilians moved into the spotlight after World War II, whereas the protection of combatants -initially, one major motive for the very emergence of IHL -is neither a topic on the international agenda nor mentioned among the topics in need of transnational advocacy (Carpenter, 2014: 34).
The interactions of obligative and omissive effects can be observed in the case of international norms obliging governments and international tribunals to criminalise and prosecute sexualised violence in conflict.First, here too, we see a strong attentional effect, as the predominant focus on women as victims of sexualised violence has obscured sexualised violence against men (Touquet and Schulz, 2021: 213-214). 7The underlying constitutive effect is that sexualised violence is defined 'in ways that often exclude men from the class of potential victims' (Lewis, 2009: 19) or even criminalise the male victim of rape next to the perpetrator through prohibitions of male-male sexual intercourse (Lewis, 2009: 25).The gender-distortedness of this norm also creates regulative effects when perpetrators of sexualised violence against men are prosecuted less, the survivors' access to post-trauma services is limited, and when survivors are silenced (Gray et al., 2020: 208).How does my conception of unintended effects relate to the conception by Nina Tannenwald who introduced this type of effects into the debate on international norms?Initially, Tannenwald considered permissive effects to be a 'subset of constitutive effects'; later, she wrote that 'the 'diverted attention' effect can arise as the shadow of either regulative or constitutive effects' (Tannenwald, 1999: 437;2007: 47).She defines permissive effects as follows: This refers to the way norms -taboos in particular -by serving as focal points, selectively divert our normative gaze.By categorizing weapons in certain ways, such as "weapons of mass destruction", and drawing our attention to associated normative injunctions, norms and taboos may obscure other "facts" about the world and shield other practices from attention (Tannenwald, 2007: 46).
The proposed typology integrates the essential aspects of Tannenwald's ideas but rearranges them.First, it recognises the impact of norms on attention as a separate type of norm effects next to constitutive and regulative effects.Second, instead of viewing permissive effects as a subordinate or second-order effect to constitutive or regulative effects, the typology elevates permissive effects to a higher level by recognising them as a distinct type and an overarching category of effects next to prohibitive and obligative effects, all of which always consist of constitutive, regulative, and attentional effects.Third, while Tannenwald's conception was limited to proscriptions, the typology expands it to prescriptions, consequently introducing omissive effects as a second type of unintended effects next to permissive effects.

An agenda for further research
In the following, I make some suggestions on how insights from this article in general and the typology in particular can inspire and inform further research on international norms.Very basically, supporting the recent conceptual contributions by Winston (2018) and Jurkovich (2020), they may help to provide more precise and comprehensive empirical descriptions of norms.Analytical precision, however, is important not just for its own sake, but also for opening new research perspectives.Acknowledging the non-normative character of permissions, for example, may lead us to the question of whether and how permissions, while not being norms themselves, can nevertheless develop normative effects.When prohibitions are lifted, what are the mechanisms that may turn the nowpermitted behaviour into an obligation?How do permissions to deviate from a norm affect the norm in question?Do regular exceptions perforate norms and undermine their effectiveness, or, on the contrary, strengthen them because explicitly stated permissions reaffirm the norm's validity?The practice of targeted killing (Großklaus, 2017) illustrates possible applications of this perspective: Is targeted killing, whose acceptance the US have tried to establish, becoming not only a tolerated, but also an expected way of dealing with hostile military and political leaders and how does it affect the prohibition of assassination?Moreover, the systematic consideration of norm effects can broaden our perspective on what can be attributed as a norm effect.Staunton and Ralph (2020) illustrate how exploring the interplay of attentional, constitutive, and regulative dynamics might add value.Their study investigates why the EU, despite her commitment to the norm of atrocity prevention, failed to act to prevent the atrocities against the Rohingya in Myanmar.Explaining this non-occurrence of a regulative effect, the study highlights a constitutive effect: Atrocity prevention was defined as a subnorm of the norm of conflict prevention without recognising that both norms require different tools to be effective.Furthermore, an attentional effect was at play when the norm of democracy promotion, while not being consciously prioritised higher, nevertheless directed the attention of the EU decisionmakers towards democracy-related policy goals, obscuring the threat of atrocities.To explore the broader question of whether increasing reactive obligations have omissive effects on preventive obligations might be a promising direction in the study of the complex of accountability and prosecution norms (Fehl, 2019(Fehl, , 2023)).
Moreover, further theorisations of how intended and unintended effects of norms coemerge are needed: What are their causes, the mechanisms producing them, and the conditions under which these mechanisms are activated?For example, regarding the attentional effects, we need to know more about mechanisms driving the allocation of attention, which have been a subject of interest in media and communication research for a while (Hilgartner and Bosk, 1988).Regarding constitutive effects, we can start theorising how unintended effects of norms are enabled through some particular properties, such as the dichotomous structure of prohibitions, or the general dual quality of norms as both stable and flexible (Tannenwald, 2007: 47;Wiener, 2008: 43).In this regard, (post) structuralist theories of meaning production (Herschinger, 2011) as well as the expanding research on norm contestation (Deitelhoff and Zimmermann, 2020;Wiener, 2014) should prove instructive.
Regarding regulative effects, studies exceeding the dichotomy of compliance and non-compliance should prove helpful in grasping how actors react to norms, both individually and at the collective level (Búzás, 2018).To study individual reactions, economic and (social)psychological research on actors' responses to interventions restricting their freedom is promising, as it suggests mechanisms such as evasion, substitution, and reactance to be at work (Jones Ringold, 2002).To study systemic reactions, scholarship on the effects of regulation in complex social systems should prove insightful (Jervis, 1997).Formulating corresponding mechanisms would mitigate the main methodological difficulty of establishing causal relationships between norms and the alleged unintended effects, as doing so requires uncovering how the norm (might have) generated the effect.
To assess the prevalence of unintended effects of norms and to expand the range of examples, empirical research might aim at discovering other cases of norms producing unintended effects, as well as at systematically mapping and studying such cases to reveal common patterns, which, in turn, should produce insights about the conditions under which unintended effects emerge and about whether and how they might be prevented.In addition to being viewed as dependent variables that we want to understand, unintended effects of norms might themselves explain other observations, as suggested above.

Conclusion
By presenting a new typology of norm effects, the article reshuffles the state of the art and adds to it through identifying and defining new norm effects.The article's contribution is as much the typology itself as is the process of developing it, since in the course of this process several central terms, concepts, and distinctions of norm research have been discussed, clarified and systematically ordered.Four main insights gained in these conceptual discussions are worthwhile emphasising again: First, in contrast to proscriptions and prescriptions that constitute two types of norms, permissions are not norms because they lack the quality of oughtness -once they acquire this quality, they turn into prescriptions.Second, norms have not only intended, but also unintended effects, defined as effects that norms produce in addition to or even instead of their original purpose.Third, attentional effects of norms should be considered the third type of norm effects next to their constitutive and regulative effects.Fourth, permissive effects that Nina Tannenwald had recognised as unintended effects of prohibitions have their pendant in omissive effects, which are unintended effects of obligations.
The previous section includes several theoretical and empirical directions that further research on the effects of norms may pursue.Another discussion that I hope to inspire is a normative one.This discussion would have to deal with difficult questions such as when potential negative unintended effects of norms outweigh their intended benefits, whether unintended effects could and should be avoided, and if so, how we can increase the intended effectiveness of norms while reducing their (undesirable) unintended effects.

Table 1 .
A typology of norm effects.

Table 2 .
Types of norm effects.