Dissecting Norms

Levi recently launched a couple new salvo’s in the debate over normativity (here, here, here and a bit earlier here), and although he hasn’t mentioned me, I think his reference to ‘transcendentalists’ who are concerned with guaranteeing normativity is probably aimed in this direction, especially after our earlier exchange over Latour (here and here on deontologistics, which petered out in a comment exchange here on larvalsubjects), and his reference to the ‘howler’ that norms don’t exist.

The major thrust of Levi’s argument still seems to be that concern with transcendental normativity precludes the possibility of first analysing the real social conditions (and their causes) that underlie undesirable political states of affairs, and then acting upon these analyses in strategic ways to undermine these and potentially produce new and better social configurations. This is put in a slightly more inflamatory way in his comparison of philosophers of normativity with the kid in the playground that thinks shouting at the top of his voice that the bully is in the wrong is enough to stop the bully. I’ll try to take this jab in good spirit.

Now, I’ve elsewhere tried to explain why I think this is very wrong, and why I’m very much in favour of basing political action on the kind of detailed analysis of social conditions that Levi says my position excludes. Indeed, as I’ve explained in detail elsewhere (here), I think that without making the proper distinctions between the normative (what ought to be) and the natural (what is) we are unable to properly provide the kinds of detailed analyses of social structures we all seem to want (and the point is that we all seem to want them). I’m not a classic Kantian deontologist insofar as I don’t think that all norms of action are transcendental and universal, but, unfortunately, the name of the blog tends to give a different impression.

However, I think there is an important difference between my portrayal of the Latourian position (which Levi endorses to some extent) and his portrayal of the transcendentalist position (which I endorse in a very specific way). I’ve upfront claimed that Levi doesn’t endorse neo-liberalism in any way, but that aspects of the Latourian position (that he seems o endorse) nonetheless commit him to something which is at the heart of neo-liberalism, which tends to undermine arguments against neo-liberalism (if not necessarily strategic political action against it). However, Levi hasn’t attempted to demonstrate why my position leads to consequences I wouldn’t endorse (i.e., pure Kantian deontology and an inability to motivate political action on the basis of detailed social analyses), as much as intimate that this obviously follows from a very general position of which my own is a species. This is slightly frustrating given that I’ve been very explicit in both wanting to base political action on such analyses, and in trying to work out some methodological constraints on such analyses (here again).

Moreover, Levi’s argument seems to rest on the idea that the analysis of normativity is of itself meant to be some kind of political strategy, and that, because this is on its own an obviously ineffective strategy, those who engage in the analysis of normativity are thereby pathetic and miserable figures unable to contribute anything  to politics. This just seems like an obvious straw man. There is  no good reason to think that the analysis of normativity (what I am calling fundamental deontology) is supposed to be a direct political strategy any more than ontology is.

1. Meta-Political Preliminaries

To put my position in as simple terms as possible: just as there is no way for us to work out what is true about anything outside of the ordinary processes of inquiry into and argument about that subject matter (e.g., the actual practice of science), so there is no way for us to work out what the specific goals of our action should be, and more importantly the means of achieving those goals, independently of the real practical disputes about the specifics of a given situation (i.e., concrete rather than abstract political discourse). Brandom would say that Truth is something that is not available wholesale, but can only be gotten retail, one bit at a time. I would add that the same is the case with the Good (what we should do, both in the sense of the goal of action and the means of achieving it). Ethics and politics aren’t wholesale matters.

However, there are norms governing how one should carry out the retail process of getting to the truth, even if they don’t give you the truth on a silver platter. This is just the normative structure of theoretical rationality as such. I would add that there are similarly norms governing how one should carry out the retail process of getting at the good, even if these norms don’t give you the good on a silver platter. This is the normative structure of practical rationality, and the important point about it is that it extends beyond personal instrumental reasoning about how to satisfy my own preferences, and into the domain of collective reasoning first upon what we should do, and then how we should do it.

Uncovering this ideal structure of collective practical reasoning is not a political strategy. Although it should have some weight in non-strategic debate, there is always the concern that any political action that proceeds by public discourse will be undercut by those who engage in the discourse in a strategic rather than a sincere manner. This is an entirely legitimate worry, and there must be strategic considerations with regard to how one deals with these (e.g., for left-leaning Americans, the question of how to deal with the genuinely distortive effect that media outlets like Fox have upon the public discourse is obviously a huge strategic problem).

However, this does not thereby mean that we should engage in debate only in a strategic manner. If we treat all discourse as a strategic matter, then we are never in a situation where what is right, be it in terms of what is true, or what is good, is really at stake in the discourse. We need to have some honest and open discussion, lest we abandon the ideal of truth and goodness entirely, but what it is to have such a discussion is a normative matter.

On the question of precisely what political strategies we should adopt, the transcendental norms of rationality are mostly silent. I don’t thereby think the inquiry into these norms thereby offers nothing of value to political thinking. There are at least two things it can offer:-

1) It can offer use ideals, or something to aim at in our attempt to change society for the better, insofar as it gives us some idea of what it would be for a society to determine its actions on the basis of genuine collective reasoning. This is not to say that it gives us the blueprint of some utopia. There is no perfect system, but this provides us with ways of assessing the imperfections and pathologies involved in current systems of collective rational determination (i.e., modern democracy). It gives us something to aim for in making this system better, even if it can never be perfect.

2) A grasp of the ideal normative structure of collective rationality can aid us in analysing the real causal structure of collective reasoning, and thus help us to develop strategies for countering and overcoming the pathologies of democratic politics, such as the endemic manipulation of the public discourse mentioned above. It can do this precisely by allowing us to purify our understanding of the real causal structure of collective reasoning of any misplaced reference to ideal normative notions. It is only by taking the ideal seriously that we can diagnose the confusions it often engenders, and thereby develop the kinds of detailed analysis of the dominant socio-political structures that is necessary to work out any kind of strategy for changing them.

Does this mean that all we can have on this transcendental approach is some abstract commitment to democracy, and maybe some more concrete methodology for developing political strategies? Yes and No. As far as I can see, this is all the transcendental approach provides on its own (maybe discourse ethicists can find more in it, I’m as yet unsure). Does this mean this is where ethics and politics end? Not at all. In fact, I would go so far as to claim that the transcendental norms demand that we push beyond them. We’re obligated to get involved in the nitty gritty retail business of politics, but this obligation does not of itself tell us what to do about healthcare, or privatisation, or any of a hundred other specific political issues. If what you want is my ideas on what we should be aiming to achieve on these concrete issues, and some strategies for trying to achieve them, then that’s a different debate, though one I’m not entirely adverse to, even if I don’t have all the answers.

This purportedly simple characterisation of my position aside, I understand that the rest of my position is quite complex. The justification of the second point above is, for instance, a very complicated matter. I also recognise that my position is delivered in a piecemeal fashion on this blog – especially my ideas about the nature of normativity. In particular, I’ve said several times that norms don’t exist (they are pseudo-beings), and although I’ve cleared up to some extent what I mean by this (e.g., here), I’ve also said that they are nonetheless manifest in the natural realm (which for me is the domain of Being as such). I’ve said bits and pieces about this latter claim, but I should probably try to be more explicit about what I mean by it. All I can do is try to make all of these ideas a bit more clear, and hope that the debate can be put on better footing.

The rest of this post will thus be a lot more technical. I will first try to provide some considerations regarding the nature of explanation and the kinds of concepts involved in it, in order to motivate a dissection of a certain way of thinking about ‘norms’. I’ll then begin this dissection by splitting apart the two different ways in which we talk about ‘norms’ into norms proper on the one hand and practices on the other. I’m then going to show precisely how these two are related, which will involve a more detailed analysis of norms proper. I’ll then try to show some of the problems that the conflation of these two different notions produces for explanations of social phenomena. I will conclude by addressing

This post is thus quite long. In fact it’s the longest post I’ve written so far, clocking in at about 15,600 words total. This is really pushing the limits of a blog post, but it isn’t really anything like a finished article either. These ideas are also still in the process of development, and I have not yet hit on the perfect mode of expression for all of them (this would most likely be a much shorter and clearer post if I had). Nonetheless, it is very useful for me to articulate my position in this way, and I hope it will be of interest to others, even if they don’t necessarily want to read it all in one sitting. I am thus grateful to anyone who takes the time to read this in full.

Now, on with business.

2. Empirico-Normative Hybrids and Quasi-Empirical Explanation

In my post on Eliminativism and the Real, I put forward the idea of empirico-normative hybrids – concepts formed by hypostatizating normative notions that are then deployed in quasi-empirical explanations of certain phenomena, but which can’t actually develop in relation to the experience of those phenomena (hereby understood to include experimental testing) in the way that genuine empirical concepts can. This inability to develop can produce a gap between these quasi-empirical explanations and the more detailed empirical concepts that can also be legitimately applied to the objects  involved in the explanation. This gap becomes apparent in relation to the counterfactual dimension of explanation.

As I suggested in that post, the concepts of belief and desire, when deployed within psychological explanations, are paradigm examples of such hybrids. As such, they can let us get a handle on this idea of a divergence or gap between empirico-normative hybrid concepts and genuine empirical concepts. Take an ordinary folk-psychological explanation of why my room mate goes to the shop to buy milk: (i) He wants milk, (ii) He believes we have no milk, (iii) He believes the best way to get milk is to go to the shop, (iv) Therefore, he goes to the shop to satisfy his desire for milk. This is a very simple reverse engineering of a likely practical syllogism that my room mate might have gone through. Now, despite the fact that most of our actions aren’t directly motivated by working through practical syllogisms, this form of explanation is nonetheless very useful.

Now, the power of a causal explanation consists in the fact that it picks out the factors of some situation that are relevant to the production of some effect, such that we can predictively project that the same combination of factors will produce the same effect, even when the irrelevant factors are varied (time and place being the classic examples). What this means is that any genuine explanation should underwrite certain counterfactual inferences, to the effect that if certain irrelevant factors had been different then the effect would nonetheless have been the same (this is called counterfactual robustness), and that correspondingly, had the relevant factors been different, that the effect would have not been the same. Moreover, the idea is that this should be projectable beyond a hypothetical variation of the given situation, to other similar situations, i.e., there is a certain ideal that the explanation should be generalisable. This ideal of generalisation is cashed out in the fact that the explanation is composed out of general concepts with domains of application beyond the given situation.

For instance, if a programmer debugging a program claims that a compile error is caused by a particular piece of code, then it should be the case that varying other parts of the code in suitably minor ways will not mitigate the error, and that removing that piece of code (or suitably re-writing it) should prevent the compile error. There is also the expectation that were this piece of code to be inserted into other programs, or were the program to be compiled on different machines, that it would produce the same error. Moreover, the features of the code (and the compiler) which are productive of the error should be formulated using general concepts with application beyond the given situation, so that they could describe other similar compile errors in different programs.

The complication here is the ‘suitably’ appended to the idea of counterfactual variation. In truth, most, if not all, causal explanations have limited threshold of variation beyond which the causal result is indeterminate. This means that in practice our concepts aren’t adequate to determine the result of every conceivable counterfactual variation. What if we compile the program with the newer version of the compiler? Who knows? However, this kind of underdetermination of how varying the situation would affect the causal chain is perfectly acceptable. To have an explanation it must only underwrite some counterfactual inferences, but it need not decide whether all such possible inferences are good or bad.

However, there is a difference between the way such counterfactual underdetermination works in the case of genuine empirical concepts and that of empirico-normative hybrid concepts. In the ordinary case, our concepts can potentially develop indefinitely by incorporating further kinds of counterfactual variation, thus developing connections with the concepts which are used to describe those variations. So, to give a terribly crude example, our concept of electrical resistance becomes more determinate when we incorporate within it the general relation between variation in temperature and variation in resistance. This process of further determining our concepts through determining the properly modal relationships they have to one another is just what empirical science is about.

However, this process is not a smooth one of continuous addition to the content of our concepts, but, as I suggested in the post on eliminativism, it also involves the points where conceptual revision becomes necessary (at these points we have to make choices about which aspects of our concept to drop, but we are not forced to make any particular choice). Sometimes this revision is subtle enough that we can still claim to be talking about the same thing (e.g., the concept of ‘light’ has undergone a fair amount of revision (such as becoming one part of the larger electro-magnetic spectrum), but we take it to indicate the same thing across its history), whereas at other times the revisions called for are drastic enough that we dissociate the new concept from the old concept entirely (e.g., splitting the concept ‘oxygen’ from ‘air’).

The important idea is that for a concept to be genuinely empirical, and thus for it to function properly in genuine empirical explanations, it must be able to undergo this process of development. This process might ultimately lead to it being mutated beyond recognition, or abandoned in favour of a better concept, but it must be open to it. This means that we must be able to expand its counterfactual or modal content, and we must be able to do this by putting it into relation with other empirical concepts we have at our disposal. The idea I am putting forward is that empirico-normative hybrid concepts actually block this process of development by (a) giving us no way of additively linking them up with certain other suitable empirical concepts, and (b) preventing us from revising them (or throwing them out) in order to enable such relations.

Let’s return to our earlier example of psychological explanation. The concepts of belief and desire we are using to formulate the explanation allow us to underwrite certain counterfactual claims, for instance: ‘If my room mate had worn a red jumper rather than a white one, he would still have gone out for milk’, and more directly ‘If he had not believed we were out of milk, he would not have gone to the shops’. However, this form of explanation does not connect up with certain kinds of counterfactual variations that we can perform now that we have more elaborate neurological concepts. For instance, if my roomate still had the same beliefs and desires, but his serotonin levels dropped by 12%, would he have still gone out for milk? What about if he his pre-frontal cortex was malformed in x way? Or, do these variations prevent him from having those very beliefs and desires?

In these cases, it seems as if not only do the concepts of belief and desire not give us any resources to adjudicate on those counterfactuals, but that there is no way that they could be made to link up with concepts such as ‘serontonin’ and ‘pre-frontal cortex’. This is because the essential features of the former concepts are the roles they play in explanations of rational behaviour, and the latter concepts are completely neutral in relation to whether behaviour is rational or not (i.e., they play a part in causal explanations of behaviour regardless of whether we take it to be rational or irrational).

Now, its important to note that none of this denies that there is some development in the use of concepts such as belief and desire. For instance, the idea of the unconscious (and corresponding unconscious beliefs and desires) is a way of modifying the standard form of folk-psychological explanation in order to give it more explanatory power (so as to explain apparently irrational behaviour). There have also been radical appropriations of the notion of desire (such as Deleuze and Guattari’s conception of desiring machines in Anti-Oedipus) which attempt to separate the notion from its roots in rationalistic explanation entirely. However, in the former case, increased explanatory power does not equal empirical adequacy, and in the latter case, the appropriation of the concept is metaphysical rather than an empirical development of it (which isn’t to devalue it in any way).

The link between the concepts of belief and desire and the notion of rationality prevent us from ever adequately providing an account of the relation that empirically identifiable variations in brain states (and otherwise described social and physical factors) have upon the causal powers of beliefs and desires, or upon their very constitution. This is because treating one another as rational individuals whose actions are determined by their theoretical understanding of the world and the preferences they acknowledge is something we are obligated to insofar as we attempt to engage in communication, debate and joint action with one another. This obligation does not vary even as our understanding of ourselves and each other as properly causal systems does.

We can thus ‘patch’ our empirico-normative hybrid concepts, to extend their explanatory power some way into the domain of the seemingly irrational, but we can never use them to describe our behaviour as properly a-rational causal systems, in the way the ideal of causal explanation demands, unless we break their connection with the notion of rationality as such, which means rendering them entirely alien to our ordinary notions of belief and desire.

The point that must be emphasized is that all this doesn’t mean that these quasi-empirical explanations are thereby completely useless. On the contrary, they can be very useful despite being empirically inadequate. Moreover, there are many cases in which precisely what we need is a basic or shorthand explanation. The process of science education is as good an example as any, where we are weened upon simplistic explanations of phenomena (which have certain predictive power) which then give way to successively more elaborate and developed theories. Nonetheless, utility and truth are liable to pull apart at some point.

In short then, empirico-normative hybrid concepts are hypostatizations of normative notions which are precluded from developing the way that genuine empirical concepts are, and quasi-empirical explanations are explanations that deploy these concepts in sometimes useful ways that are nonetheless empirically inadequate insofar as they prevent us from linking them up with our broader empirical understanding of the way in which the causal situation may vary.

3. Dissecting Norms

The point of the preceding section was to motivate the idea that the concept of ‘norm’, at least in the way it is often used, is in fact an empirico-normative hybrid similar to that of belief and desire. I’m going to try and perform a dissection of this hybrid, so as to pull apart its two halves. The fundamental idea underlying this is that we often use the term to mean two distinct, but intimately related things: a kind of collective practical commitment and a kind of collective behavioural regularity. We might also characterise this as the distinction between norms proper and practices, respectively. I’m going to try and show what these two things are, why they must be kept distinct, but also why they are so intimately related and thus easy to confuse.

To take the latter notion first, we often use the term ‘norm’ to refer to something like a regularity present in the behaviour of a given community, or a set of practices that the community engages in. For instance, we might talk about the traditions of a certain culture (the example I keep using is the Jewish kosher laws, but we could equally talk about Indian marriage taboos, or English dinner party ettiquette) in terms of the rough patterns of behaviour exhibited by members of the culture and the mechanisms through which these patterns of behaviour are passed on across the years and maintained within certain tolerances at any given time (e.g., the oral and written history of the given traditions, the educational practices through which these are bestowed, the power structures (such as courts or other organisations) which are involved in interpreting the traditions in difficult cases, and the structures of punishment or sanction used to correct deviant behaviour, etc.).

It is important that we talk about both the behavioural regularity and the social mechanisms through which the regularity is maintained (or regulated), because we wouldn’t ordinarily identify all collective behavioural regularities as ‘norms’. It might be that most or all the people living in a certain area go swimming on Friday, and there might even be good causal reasons why they do so, such as the fact that there is a lower price on Fridays, but this doesn’t constitute any local ‘norm’ regarding going swimming. For the regularity to be identified as a norm it seems that there must be a certain special kind of causal reason for its provenance. The simple answer here is that it consists in the regularity being brought about and maintained through the members of a community correcting eachother.

Now, this correction can take all kinds  of different forms. It can be anything from completely explicit debate over the content of some rule (such as a law or protocol) to completely implicit correction of one another’s behaviour without appeal to anything like an explicit principle or rule (such as correcting another’s table manners without appeal to any codified set of rules governing table manners). However, it is the fact that these processes of correction can involve explicit codifications (e.g., rules) of the regular behaviour, and that there can be arguments over the interpretation of these codifications, which produces the intimate relationship between norms proper and practices. It is also this fact which entices us to conflate them in a hyrbid  concept of ‘norm’.

It’s important to note that the notion of practices I’m pushing here is meant to be slightly broader than this restricted notion of collective behavioural regularity that is bound up with the hybrid concept. It is not meant to be restricted to cases of regularities brought about entirely through processes of correction. The reasons for this will become clear later on. For now, we will turn to the other concept of norm.

The other way of using the term ‘norm’ refers to what I above called a kind of collective practical commitment. To understand what this is it is important to understand what practical commitments are in general. Now, I’ve talked a lot about the Brandomian idea of theoretical commitment, but I haven’t discussed practical commitments anywhere near as much. I appologise for the length and complexity of the following explanation of practical commitments, but it is a necessary evil.

4. Practical Commitment and Norms as Reasons

In essence, ordinary practical commitments are just commitments to act in certain ways. Brandom thinks of them as non-inferential outputs of discourse, in a way analogous to his conception of observation as a non-inferential input into discourse. What this means is best understood with reference to the kinds of practical syllogism we mentioned above.

The syllogism that was implicitly ascribed to my room mate above is something like: ‘We don’t have any milk’, ‘I want some milk’, ‘The shop will have milk’, therefore ‘I shall go to the shop’. The idea is that the conclusion here is a practical commitment, a responsibility to do something, even if it is not strictly a responsibility to anyone but oneself. One can of course undertake practical commitments that aren’t the outcome of chains of reasoning (e.g, in a fit of bravado declaring ‘I shall climb everest!’), and one can also undertake them in ways that do make one responsible to others (such as promising someone that one will do something).

However, as in the case of making a promise to someone else, undertaking a practical commitment (even in the case of personal instrumental reasoning) does not causally necessitate that one will fulfill this commitment. It provides a reason to act, rather than the action itself. There are obvious cases in which one is hindered in the performance of the act by external forces (e.g., someone shooting my house mate dead on the way to the shops), but there are also cases in which one consciously ignores a commitment one has undertaken to another (e.g., breaking a promise) or in which one does something other than what one rationally recognises one should do – akrasia – or weakness of the will.

The latter case is most interesting, insofar as it leads to one of Brandom’s more important insights, namely, that part of what it is to treat someone as a rational entity is treating them as disposed to act on practical commitments they undertake as a consequence of practical reasoning. This need not be an unshakable disposition (someone need not be perfectly rational), but we could not count someone who never acted upon such practical commitments as a rational agent.

Now, all of the above examples of practical commitments are examples of acknowledged practical commitments. However, practical commitments are subject to the same difference between acknowledged commitments and real commitments as the theoretical case (which I’ve talked about elsewhere here and here). This means that one can be committed to do things that one does not take oneself to be committed to.

However, what one is really committed to can exceed what one takes oneself to be committed to in two different ways. On the one hand, one’s grasp of the content of the commitment one has explicitly undertaken can be deficient, so that one does not fully understand what it would be to fulfill the commitment. For example, one might undertake a commitment to climb Everest, but not understand that Everest is in Nepal, and thus that one has thereby undertaken a commitment to go to Nepal. This is exactly the manner that the gap between acknowledged and real theoretical commitments works. On the other hand, one can have done something which thereby commits one to do something else, even though this action did not involve an explicit undertaking (i.e., an acknowledgment) of a further commitment. Here one has not misunderstood what the action being performed consists in, but rather what the wider social significance of performing that act is. For instance, when one uses a swimming pool, one undertakes to abide by the rules of that swimming pool. Most of us understand this even if we do not actually understand what the precise content of all these rules is. In this case we have committed ourselves to something that exceeds our grasp of it, but not by explicitly acknowledging anything.

These points aside, it is crucial to recognise that although practical commitments play this role in producing outputs from reasoning, via a causal disposition to act on our commitments, they are not simply the conclusions of practical reasoning, but can also be premises. This is obviously the case in instrumental reasoning, where we aim to work out the appropriate means for acting upon a commitment we’ve already undertaken. To extend the above example, once my room mate has undertaken the commitment ‘I shall go to the shop’, he may then want to reason about the best way to carry out the commitment. For example: ‘I shall go to the shop’, ‘The shop will close soon’, ‘The bus is the quickest way to the shop’, therefore ‘I shall get the bus to the shop’.

Now, it’s pretty obvious that it can’t be a condition of carrying out a practical commitment that we then perform instrumental reasoning which breaks it down into other practical commitments, because this would end up in a horrible regress, whereby we could never do anything until we’d worked out how to do all the individual steps required to do it, and we couldn’t carry out all the individual steps without working out the steps required to do them, and so on. There must be some level of practical understanding or know-how that we can tap in carrying out a commitment without having to proceed via further instrumental reasoning. Nonetheless, although there must always be a point at which instrumental reasoning gives way to such know-how, this is not to say that there is any specific point at which how a given practical commitment should be carried out cannot be rationally analysed.

Finally, it is important to understand the difference between ‘should’ and ‘shall’. An assertion like ‘I shall go to the shop’ indicates that one acknowledges a practical commitment to go to the shop. The reason this is indicated by ‘shall’ rather than ‘should’ is that there is the implicit assumption that by acknowledging that commitment that you will act upon it (as explained above). However, when we ascribe a practical commitment to someone else we use ‘should’, because they do not necessarily acknowledge it. So, we might say of someone who has explicitly stated ‘I shall climb Everest’ that ‘He should go to Nepal’, or that ‘He’s obliged to go to Nepal’. What all this indicates is that the content of a practical commitment (e.g., me climbing Everest, or my room mate going to the shop) is independent of the way it is ascribed to or acknowledged by the relevant person. This content is just the content of the proposition the person is committed to making-true (e.g., ‘Pete has climbed  Everest’, or ‘Pete’s room mate has gone to the shop’).

Now that we have gone over what practical commitments are in general, we can start to get a grip on what specifically collective practical commitments are. At root, they are commitments that groups of individuals undertake. When I say that norms are a kind of collective practical commitment, I mean to exclude a certain very straightforward kind of commitment that a group can undertake, e.g., a housing co-op undertaking a commitment to build a new set of houses, or a nation undertaking a commitment to wage and win a war against another nation. These kinds of commitments are commitments to joint action, and they set about to accomplish some specific task. Now, although we can institute additional social norms as ways of organising such action towards achieving this task (e.g., the social norms which kick in in war time – ‘one should keep mum’, ‘one should not waste resources’, ‘one should support the troops’, ‘one should join the army if one meets x conditions”, etc.), we need not necessarily do so. Instead, we might simply break down the process of accomplishing the overall task into smaller tasks, which can ultimately be broken down into individual practical commitments (e.g., ‘I’ll handle the process of getting planning permission, you hire the architect, and you guys pitch in with the building work’).

What distinguishes norms from the above kind of collective practical commitments (which we might just call joint tasks), is that they do not involve each member undertaking a commitment to bring about some overarching state of affairs (albeit through different and co-ordinated actions), but instead involve each member undertaking to follow some conditional task responsibility or maxim. We can best explain this in terms of the kinds of inferences one can draw from ascriptions of the two kinds of practical commitment.

For example, say that ‘The residents of Ryhope village are committed to not littering’, and also that ‘Pete is a resident of Ryhope village’. What does this imply Pete is committed to? If we read this as a joint task, then on the face of it Pete is committed to bringing about the state of affairs where no person in Ryhope litters, or performing some specific action as part of a plan to bring about that state of affairs. On the other hand, if we read it as a statement of a norm, Pete is simply committed to not littering himself. Of course, not littering could be simply the part Pete plays in a larger plan to bring about a state of affairs where no one litters, but this is a special case of a joint task.

Norms thus function in a certain way in practical reasoning. For example: if ‘Bank tellers should wear ties’, and ‘Jeff is a bank teller’ then ‘Jeff should wear a tie’. This works just as well in first person reasoning: if I am a bank teller, then I should wear a tie (or ‘shall’ wear a tie, if the reasoning is not hypothetical). For a group to be bound by a norm is for each individual member of that group to have the same conditional task responsibility (in this case, wearing a tie when on the job).

It is important to note that the conditional structure of a set of norms can get quite complex, and it can differentiate the tasks that any member of the group takes on insofar as they take on certain roles within that group, for instance ‘If one is a designated driver, then one should not drink alcohol, but if one is not a designated driver, then one should buy the designated driver drinks’. Indeed, the above example of ‘bank tellers’ could easily indicate such a social role. These kind of complex norms which differentiate between roles are precisely the kind of norms often required to organise the kind of joint tasks we talked about above. Moreover, in social institutions that require not just the bringing about of a particular state of affairs, but the maintenance of that state of affairs, some such set of norms is often essential (e.g., the explicit and implicit rules that govern the internal workings of a corporation or public institution like the post office or a government department).

So, to finally end this section: norms are practical commitments or obligations on the part of certain groups, but they are not necessarily commitments for the group as a whole to act in a certain way (even if this is then broken down into more bitesize individual commitments), but are instead commitments for each individual member of that group to act in certain ways under certain conditions. This means that a norm is a reason for each member of a group to act in a certain way, precisely because it entails that each member has an individual practical commitment to act in that way.

5. Rules, Norms and Practices

It’s now time to turn to the connection between norms and practices. To do this it is helpful to consider the classic Wittgensteinian problem of rule following. We won’t go into this in too much detail, as there has already been a mass of literature on this topic that we’d rather not engage with. Nonetheless, a brief overview will bring out the salient points. Wittgenstein’s problem is the following: how is it possible to follow a rule correctly?

The idea here is that any rule governing what we should do can be followed either correctly or incorrectly, and it seems that to be able to follow it correctly we must be able to understand what it would be to so follow it. To misunderstand what the rule requires would be to misinterpret it, and therefore it seems that to understand it properly is to interpret it properly. However, interpretation is something that itself can be done correctly or incorrectly, and so it seems that there must be some rules governing how we are to interpret the original rule. However, this produces a vicious regress, as this means that to properly understand any given rule we must understand some further rule, and so on. As Wittgenstein says, this implies that for it to be possible to follow any rule correctly, it implies that there must be some way of grasping at least some rules that is not a matter of interpretation.

Brandom translates this as saying that there must be some way of grasping a norm that is not a matter of understanding an explicit codification of that norm in the form of a rule. Indeed, it is a condition of us being able to understand and interpret explicit rules that there are implicit norms governing how they are to be so understood (what we might call interpretational norms). This consists in us already having some abilities to just do the things in question (interpretation being one example of something we do). This connection with action is where Wittgenstein’s insight ties in to some of Heidegger’s early insights about the primacy of practical intentionality over theoretical intentionality.

We above said that the Wittgensteinian problem is a matter of how we understand what it would be to follow a rule correctly. We can actually break this up into two parts: the state of affairs that following the rule correctly would bring about, or the end of the action, and the means through which it is achieved. Once we do this, we can see that the Wittgensteinian regress joins up with another regress that we mentioned earlier, namely, that in order to be able to do anything, we must have some understanding of how to do it that does not arise from instrumental reasoning. Heidegger’s real insight (suitably interpreted of course) is that it is the practical know-how which staves off this instrumental regress which staves off the more general Wittgensteinian regress.

The point is that our understanding of the possible ends of action, and thus of the contents of the propositions which we make-true through those actions (and thereby the content of the rules which demand that we make them true), is dependent upon our practical understanding of the means through which these ends are achieved. Moreover, this understanding of means is independently grounded in the social realm, in how ‘One does’ perform the task, rather than being itself dependent on instrumental reasoning about how a task should ideally be performed. The latter claim needs justifying in more detail, and to do so we need to get to the heart of this relation between practice and understanding.

For Heidegger and Wittgenstein, it is in virtue of always-already being thrown into some culturally articulated world, or form-of-life, wherein we are given over to a certain understanding of ordinary tasks and how they are performed, that we can understand non-ordinary tasks and unusual means. To put this in more Heideggerian language, it is in virtue of being given over to some understanding of things that we can interpret and develop that understanding. This Heideggerian notion of interpretation is precisely not a matter of instrumental reasoning, but is meant to be something prior to this kind of reasoning which makes it possible. What it indicates is an ability to consider, take apart, and adjust the ways of doing things that we have inherited. Putting this in a different way, we need to have some dispositions for doing certain things in certain ways, in order for us to have the possibility of doing anything in a novel way, and at least some of these dispositions must be acquired, or at least maleable, so that we can alter and adjust them in an interpretational manner.

How then does the social enter into this story?

It is the fact that these dispositions are largely shared which makes possible the explicit use of words to express the same contents (at least putatively ‘same’, but this is another issue). One of Heidegger’s less well understood ideas is that discourse or talk (Rede) is primarily a matter of organising joint action, and thus that it is through our practical understanding of what is involved in joint action, and our development of words for articulating it, that these words gain the possibility of being used in so called apophantic discourse, which lets us make assertions about the way things are in abstraction from what we are doing with them. This ‘accruing of words to significations’ and the development of practices of assertion out of this, is only possible on the basis of some commonality between our dispositions for action. This is to say on the basis of something like the kind of collective behavioural regularities or practices we discussed in section 3.

This is where we see why we can’t restrict the notion of practices to those regularities produced through social processes of correction. This is because, although there must be some commonality amongst our dispositions, it need not be caused by social correction. There can be more or less hardwired capacities for doing things that we all share, and non-social environmental factors that cause us all to develop certain capacities or dispositions for doing certain things in certain ways. Now, this doesn’t exclude social processes of correction as a cause of this commonality, and there is very good reason to think that such social processes play a very important role in establishing and maintaining this commonality. However, if this commonality does not essentially consist in social processes of correction, what is the relation between them?

Let’s think of two different ways in which we can correct someone’s action in relation to some end. On the one hand, we can correct their choice of means, or the way in which they are trying to realise the end (e.g., ‘You shouldn’t hold the hammer like that!’). Whether the mistake being corrected is a matter of incompetence (which might prevent the task from being completed at all) 0r simple inefficiency does not matter. On the other hand, we can correct their understanding of the state of affairs they are meant to be bringing about, or the end of the action (e.g., ‘That isn’t how you make a bookcase’, or ‘A bookcase shouldn’t be like that’).

Here we have two different senses of ‘should’: an instrumental sense and an obligatory sense. Although the former sense gets its force from some practical commitment that the person corrected has undertaken (i.e., to build a bookcase), what provides the relevant reason for the correction is the instrumental matters involved in bringing about the end (e.g., one shouldn’t use a hammer in a certain way with a certain type of nail that one happens to be using).  The latter sense gets its force directly from the practical commitment they have undertaken, insofar as it indicates that they have misunderstood the end itself, or the content of the comitment.

This becomes more complicated when we apply this distinction to the ways in which we can correct norm-guided actions, because the ends of such actions are conditionals. Thinking of it this way, we can see two ways in which one could correct the same action, for example, the use of a hammer. One might, as above, say that ‘You shouldn’t use a hammer that way’,  to indicate that there is a better way of bringing about some further end. However one might also use it to indicate that one is thereby breaking some norm of hammer usage (e.g., ‘if one uses a hammer, one should use it thus‘).

It is this fact that produces a certain amount of ambiguity in the reading of ‘das Man’ in Heidegger, insofar as sometimes it is unclear whether the world is constituted out of genuine norms for the proper use of equiptment. The response to this reading is usually that one is not obligated to use the hammer in the way ‘One does’ use a hammer, even if for the most part we understand hammers in terms of the way ‘One does’ use them. The correct interpretation is to recognise that norms can contribute to what ‘One does’ in a big way, but they do not constitute it in its entirety. It is thus the case that even if some of our practices are maintained through processes of correction that are more than just matters of instrumental correction, but veer into the territory of obligatory correction, it need not be the case that all of them are this way, and indeed it could not be the case.

Nonetheless, what the above considerations indicate is that, in the case that we undertake a practical commitment (be it a norm, a commitment to a joint task or an individual practical commitment), there is always some sense in which we can not only use inappropriate means, but rather fail to aim at the right end. This is because even though the force of a given commitment might only come from the fact that it is a responsibility to oneself, the content of the commitment is something one does not have complete authority over. So, if I announce a commitment to climb Everest, I might be equally able to renounce the commitment, insofar as I have not made the commitment to anyone (whereas one cannot take back a promise without the approval of the promisee), but whether or not getting a third of the way up the mountain counts as ‘climbing Everest’ is not something that is completely up to me. I can renounce my commitment and take up a new one, but I cannot change the content of my original commitment.

Thus, although the possibility of understanding and following an explicitly formulated rule is dependent upon a background of common practical dispositions which ultimately fix the content of the terms used in its formulation, there is always a legitimate question as to whether one has correctly followed the rule, and argument about this must always ultimately appeal to implicit norms governing how the content of the commitment is to be interpreted (e.g., norms governing what is implied by ‘Pete has climbed Everest’, such as ‘Pete has gotten to the peak of Everest’). In essence, although it cannot be that it is necessary to interpret a rule to follow it, there must always be the possibility of explicitly interpretating it, and engaging in interpretational debates about its content.

This kind of interpretation is different to the kind of Heideggerian practical interpretation we talked about earlier (what he would call circumspective interpretation), insofar as it is a matter of working out the content of the commitment rather than how best to execute it. The important point is that insofar as the content of a norm is fixed by our practices, the norm is implicit in those practices, and similarly, that insofar as our the terms a rule  is specified in are fixed by our practices, the specific norms of interpretation we appeal to are implicit in those practices.

This means that arguments about how to interpret a rule ultimately regress to arguments about how to make explicit the norms implicit within our practices. These arguments then appeal to facts about what has been done to make claims about what should be done. The irreducibility of norms to practices consists in the fact that the former facts always underdetermine the latter. This is not to say that there are no good reasons for picking one interpretation over another, but rather that this process of interpretation is inherently open ended. There will always be facts which give us reason to exclude possible interpretations, but there will always be points at which we have to make decisions about how to apply the rule by making appeals to some selective reading of what has previously been done.

I’ve talked about this elsewhere (at the end of this post), by giving the example of case law, wherein a judge must work under the assumption that the law, in conjunction with the weight of previous cases, is determinate and thus dictates what must be done in a given case, and yet must cherry pick individual cases in order to build up an interpretation of the law which justifies why one thing must be done rather than another. This is not a matter of making stuff up, but is an eminently rational activity. It is simply the case that this activity is as much a matter of determining the norms, as it is a matter of working out what they dictate we should do.

However, it is a different kind of rational process than the process of description which goes on in the sciences. Even though experimental evidence always underdetermines which scientific theory is right, we can always develop more tests, and get more evidence. The world is fully determinate, and it will always help us choose between theories as long as we can work out the right question. Interpretation on the other hand does not have the kind of plenitude of evidence which is bequeathed to description. For instance, in interpreting the intentions of the writers of the US Constitution, we will always be confronted at points where we have to make a best guess, by marshalling an interpretational narrative gleaned from the historical facts available.

The reason I have been claiming that norms do not really exist, or that they are pseudo-beings (here), is precisely because the truth of interpretations of them cannot be assessed in objective terms. This indicates that there is no real, fully determinate thing underlying our talk about a given norm. This isn’t to say that talk about norms is not a matter of truth, simply that the kind of truth in question is not objective truth.

This isn’t to say that there is anything indeterminate about our practices. The collective behavioural regularities present in our communities, along with the social mechanisms and other factors which maintain them, are fully objectively describable. It is simply the case that one cannot simply read the norms that we follow off of the way we actually behave. This is the same argument made against teleological approaches to nature, insofar as we cannot straightforwardly read norms of proper functioning off of the way things (such as internal organs) actually function. This is not to say that we can’t get some good regulative principles out of how the world actually works (I’d like my heart to keep on beating the way most people’s hearts do, and I’d like my doctors to make sure that happens), just that there will always be troubling cases where we have to make choices (such as when organs or organisms adapt in ways which diverge from their purported ‘role’ but in ways that are apparently beneficial).

To conclude this very long section, we thus have a three levelled structure: explicit rules, implicit norms, and objectively describable practices.

The essential difference between explicit rules and implicit norms is that the former are codified and the latter are not (the process of codification is a matter of interpretation).

The essential difference between norms and practices is that norms are just reasons for action, whereas practices consist in the actions and dispositions to act of a given group, and as such are causes of actions.

6. Social Networks and the Structure of Practice

Now we’ve described the basic structure of the interface between norms and practices, we can see why it is easy to conflate them in a hybrid concept. However, it is important to flesh out this interface further, because it is only on this basis that we can demonstrate how the hyrbid concept of norm that some deploy in social explanations makes those explanations quasi-empirical in the same way as folk-psychological explanations of human behaviour.

What we showed above is that for there to be anything like norms proper, there must be the possibility of theoretical interpretation of those norms (rather than just Heideggerian practical interpretation), but that this theoretical interpretation is grounded in our practices, insofar as there are specifically interpretational norms, and we must simply be disposed to just follow them in certain ways. Moreover, although these interpretational practices are liable to be maintained through social processes of correction (in the obligatory rather than instrumental sense), they are ultimately grounded in what we might call our instrumental practices of doing certain things in certain ways, and these need not be maintained through correction (although they can be in part).

In addition to showing that norms are grounded in practices in general, we have also shown (in brief) how it is that specific practices become involved in the interpretation of those norms. However, what we haven’t discussed is how these processes of interpretation play a role in the causal story regarding the production, maintenance, and development of our practices. Put differently, we haven’t shown how our talk about norms plays a role in the social machinery that underlies our collective behavioural regularities. This is our most pressing task.

In order to do this it is important to recognise that we can equally view human discourse from two perspectives. On the one hand, we can view it from a normative perspective, wherein we assess what is meant by the claims made within it, the justification of those claims and their relation to one another. This is to take up what Brandom calls the attitude of deontic scorekeeping in the game of giving and asking for reasons (I’ve got another post brewing explaining this in more detail). On the other hand, we can view it from a naturalistic perspective, wherein we assess the interactions between the participants in a debate in causal terms, and try to provide an explanation of how it is that they affect one another’s behaviour that is not beholden to the notion of justification. These two perspectives are what I have elsewhere called the internal and external perspectives toward rationality, respectively.

It is important to note that engaging in discourse properly requires adopting the former perspective, but that it can also be adopted toward discussions one is not involved in. However, it is difficult to adopt the second attitude to discussions one is involved in, and there is a good sense in which doing so precludes taking up the former stance properly. This is what I was getting at in the earlier posts on Latour, insofar as the Latourian position seems to imply that we always adopt the latter stance (see here and here). Regardless, it is perfectly possible to take up the latter stance in relation to discussions that one is not involved in, and in relation to the general structure of discourse as such. This external perspective is precisely the perspective from which we must address discourse if we are to understand what role it plays within the real causal networks that make up the social domain.

Now, before we look at how the interpretation (or explication) of norms plays a role in the mechanisms underlying practices, it is helpful to understand how these mechanisms function in a case where there little or no explicit interpretation. Let’s take dinner party ettiquette as an example. This is a set of implicit norms governing how one should act at a dinner party in a given society. Now, if we look at this from the side of practices, we see that this consists in a certain set of collective behavioural regularities that are displayed by guests at dinner parties across the society in question. However, as noted earlier, we are not just interested in the bare fact of the regularity, but with the systems through which that regularity is maintained or regulated. What is good about the example of dinner party ettiquette is that because the norms are for the most part implicit, explicit interpretations of the norms and discourse about them should play little if any role in these systems.

The first thing we should note is that in talking about a collective behavioural regularity we are not thereby implying that the dispositions to act of the group in question are thereby homogeneous. They are rather necessarily heterogeneous. Different groups of diners at different dinner parties will be disposed to act in different ways. Rather, the idea is that there is a certain amount of commonality between their practices that is perpetuated despite the fact that there are many differences, and importantly, despite the fact that many individuals within the group will be judged to be erring, or deviating from this rough common practice.

The next thing to note is that the group is constantly gaining new members and losing old ones, and as such there must be a mechanism for incorporating those new members in the group so as to sustain it across generations. Although in other cases there will be specific social mechanisms in place to do this that are separate from the ordinary mechanisms through which the rough commonality is maintained (e.g., educational systems which induct children into the practices of interpretation required to read), in the case of dinner party ettiquette these are mostly the same (although there are of course finishing schools). These ordinary mechanisms are just the ways in which we influence one another’s behaviour in actual instances of practice, in this case attending a dinner party. Here, we mimick one another’s behaviour, and at times we give signals regarding (and sometimes explicitly correct) what others are doing wrong.

In short, all dinner party guests are part of a social network through which they reciprocally influence (or correct) one another’s behaviour in the context of dinner parties. Although this is a matter of reciprocal correction, it is not for that matter symmetrical. Some nodes in the network exert more pressure on others than is exerted on them. This is indicative of power relations and hierarchies of status. Similarly, some nodes make more connections in the network than others, giving them more influence over the general pattern of behaviour than others (dependent upon how much influence they can exert).

The social network thus has a very complex topological structure, and it is the complexities of this structure that constitute the above mentioned heterogeneity. For instance, there might be significant divergences between the behaviour of those in different areas of the network (possibly corresponding to geographical region, but not necessarily), such as variations in what constitutes a proper cheese course. Such nascent divergences can ebb and flow, and potentially dissolve or produce massive rifts within the network itself. A different example of this is the development of languages. Sociolinguists have done some very good work on the way that tiny divergences of usage within local social networks of speakers can balloon to create dialects, and ultimately how dialects can drift apart and produce separate languages (thus splitting the original network into parts).

Now, if we understand the basic structure of practices in these terms, we can see how easy it is to show how non-human factors can have a causal role on their creation, maintenance, and development. For instance, the differences in norms governing what constitutes a good cheese course between parts of the network might have to do with regional differences in what cheeses can be easily procured. We can thus happily endorse ANT style analysis of how the farms, farmers, cows, milkmaids, soil conditions, working conditions, and other factors take up roles in the network of causes that is productive of a particular practice. Moreover, we can understand how these various other non-human actors interact with the social network the practices consist in through their interactions with the individual humans that make it up.

So, we now have the outline of a picture of how the mechanisms underlying the practices which implicit norms are implicit in function, and how these mechanisms can involve non-social factors. We now need to show how the processes of interpretation through which implicit norms are made explicit, and explicitly instituted norms are applied to new cases, are bound up in these mechanisms. The short answer is that they are bound up with them in innumerable different ways. There is no clean cut opposition between completely implicit norms and completely explicit norms, but there are in fact many ways in which the talk about norms as explicit rules of action enters into determining our patterns of collective behaviour. Part of the importance of separating out norms and practices is so that we do not overlook the complexities of these social structures by assimilating them to a model borrowed from our understanding of norms as reasons for action.

This point about complexity aside, we will try to show how interpretation can enter into practice in at least two different ways. On the one hand, we will suggest how our description of the social network underlying the implicit norms of dinner table ettiquette could be altered by the introduction of a widespread discourse on the interpretation of these norms. On the other, we will return to the example of case law as a system within which the stipulation of explicit rules, along with the process of interpreting them, is an essential structural feature of a set of practices.

Taking the first example, it is perfectly possible to imagine there being a point at which a set of norms, like those appealed to in hosting and attending dinner parties, which are for the most part implicit in practices, could become the object of a discourse that tried to make these norms explicit, which itself then affected the way the practices are regulated and developed. When we talk about these norms being ‘for the most part implicit’, what we are recognising is that there can be exchanges between dinner party guests which do involve arguing over the explicit interpretation of some aspect of the norms, without this argument thereby playing a large role in the mechanisms through which the practices are regulated. If such instances are isolated from one another then there is no sense in which a proper discourse on the interpretation of the norms has been constituted. Such instances must open up onto one another, such that reasons given for interpretating the norm in one way in a given exchange can potentially be transferred, deployed, and potentially criticised in others.

This isn’t to say that such discourse must be egalitarian, and that every dinner party conversation about the norms must be able to have an affect on the wider interpretative discourse, just that there must be some structure through which different instances can connect up, no matter how hierarchical that structure might be. For instance, it’s quite possible for there to ultimately be books published on the matter of how one should act at dinner parties, by authoritative authors, who may indeed disagree with one another and carry out an argument in print. It is possible for this kind of academic debate to form a central part of the social network through which the norms are maintained, the authors thus becoming sort of supernodes in the network. This is not necessary, but it can happen, and moreover it involves this social network intersecting with a bunch of other social and non-social factors, such as the economics of printing books on dinner party ettiquette in a recession (which necessarily involves more of those non-human actor things), and the other social factors determining the authors’ social status in different sections of the population. A public discourse (whatever form it takes) can thus slot into those mechanisms which are productive of the very practices they draw on in order to make its interpretations (and how it slots in depends precisely on the form it takes).

The second example is one in which the public discourse upon the norms plays a integral role in the mechanisms through which the practices those norms are implicit in are maintained and developed. In fact, in the legal system, the discourse has a very rigidly determined structure, with very explicit statuses that confer interpretative authority, and procedures for questioning that authority in light of counter-arguments, and how those counter arguments must be formed and presented. Also, this is a case in which this discourse does not play so much of a direct role in regulating practices, as it does slot into a whole bunch of other mechanisms which do so (the police, the penal system, etc.). Moreover, this is also a case in which there are acts of stipulation (on the part of politicians who write the laws), which are meant to establish explicit rules for action that are not thereby mere interpretations of norms already implicit in our practices. Here we have a case in which a different kind of discourse also inputs into the determination of practices.

Obviously, I’m not aiming to provide a detailed and nuanced analysis of the way in which the legal system slots into our society here, as much as gesture at the way we must think about it if we are to provide one. What we have to see is that having such elaborate processes of interpretation involved in the social mechanisms underlying our practices produces a different kind of self-regulation than the implicit case. It’s always the case that practices self-perpetuate in some way, and it’s also the case that all kinds of external factors can be involved in this process of through which practices regulate themselves.

However, the involvement of processes of interpretation allows these external factors to input in a far different way. For instance, it allows the legal system to involve authoritative scientific testimony (which is itself dependent on certain highly structure social networks) into the process through which a given law is interpreted in a way that is effectively impossible in a more diffuse social network. This also means that practices can adapt and become more complex in a much more rapid way. For instance, the development of social mores dealing with matters of obligation prior to the advent of the legal system was far slower than the rapid development that has happened after it (think of how detailed contract law has become).

7. Quasi-Empirical Explanations of Social Phenomena

Now that we’ve given a bit more detailed account of the interface between norms and practices, we can proceed to articulate the problem with explanations that run these two different things together. What must be maintained from the outset is that such explanations are not useless, and moreover they can seem genuinely informative. The problem is that if we take this usefulness to indicate that they reflect the real structure of the world, then we undercut our ability to integrate our understanding of the phenomena they explain with our understanding of other phenomena. We thereby lose our ability to develop potentially better modes of explanation on the basis of empirical research.

In order to make the claims I want to make about the hybrid concept of norm, I first want to provide an analogy with a different kind of concept that can be used in more obviously problematic kinds of social explanation: ‘the will of the people’. One might crudely appeal to ‘the will of the people’, in explaining the outbreak and ultimate success of the American War of Independence. There is of course a certain amount of value in such kinds of crude explanation. Of course the support of the populace played an important part, but this kind of shorthand reference to that support misses out its actual detailed structure. It misses out that there were some who were more supportive than others (and indeed that some were quite opposed), and that this support was maintained and reinforced through heterogeneous social networks, and that the way in which the different points within these social networks contributed to the overall production and consumation of the revolution was full of complicated chains of causation.

If we take the ‘will of the people’ as just a useful shorthand which is meant to indicate this more detailed causal story, then we’re fine, but if we reify the ‘will of the people’ and treat it as some additional object which is involved in this causal story then we preclude ourselves from providing the more adequate causal story. This is the case even if we treat the ‘will of the people’ as being constituted out of the social networks we were talking about. This is analogous to those who say that we can find beliefs and desires in the brain, as constituted out of neurological states. This is because this idea of a ‘will of the people’ is fundamentally based on an analogy with the more ordinary notion of a ‘will’ or ‘intention’, and as such is meant to incorporate something like a content which specifies what state of affairs the will aims to bring about (these notions themselves being parasitic on the proper notion of a individual practical c0mmitments). In this case the content is something like ‘the peoples’ will was that they be free from British rule‘. Insofar as one takes this content to play some kind of role within the explanation of how the phenomenon occurred, one falls back into thinking in terms borrowed from our normative understanding of the structure of practical reasoning.

Specifically, in the case of the ‘will of the people’ we are talking as if a collective practical commitment to a joint task of the kind we talked about above is playing a causal role in facilitating the American War of Independence, and we are thus given to understand how it is causally efficacious in terms of how it enters into the rational action either of the group as a whole (as if we understood it as a single rational individual), or of the individuals who make if up. In either case, we are drawn to think how it is that this content – ‘being free from British rule’ – determines rational action. It should be obvious that thinking of ‘the people’ as a single actor whose action is determined in accordance with some practical commitment is fundamentally flawed (Foucault has repeatedly told us as much). However, we must also emphasize how it is also wrong to understand each individual within the group as rationally acting upon this same practical commitment. In binding the actions of all of this group together under a single content, what we have done is to suppress the heterogeneous character of the various dispositions of the individuals in the network, and the complex structure of the network through which they influence one another.

One can now see how this is parallel to the case of ‘norms’. Rather than ascribing to a social group some state or object which consists in a reification of a joint task, explanations which appeal to the hybrid concept of norm ascribe a state or object which is a reification of a different kind of collective practical commitment, or what I have called a norm proper. This case is equally problematic, because in virtue of ascribing a practical commitment with a fixed content, this kind of explanation implicitly homogenizes the practices which play the real causal role in producing the social phenomena we are interested in. If one wants to maintain that the ‘norm’ is somehow constituted out of the heterogeneous practices, one has to give some kind of account of how this heterogeneity is condensed into a single content. However, the content is in fact grasped through some kind of retroactive operation of averaging the heterogeneous behaviour of the group. This just assumes that there is some way in which a homogeneous content can be produced out of such heterogeneity, rather than demonstrating it.

We are thus drawn into the exact same situation as we were with folk-psychological explanations above. Just as we have no way of linking up changes in serotonin levels with changes in the contents of beliefs and desires, so we have no way of linking up subtle changes in the behavioural dispositions of individual agents, and the social mechanisms that connect and regulate them, with changes in the content of hybridized ‘norms’.

I anticipate that the response to this will simply be to claim that talk of ‘norms’ in the way I’m laying it out never implied anything like the idea of a homogeneous content that I’m describing. If that was genuinely the case, then we would never really have been talking about norms proper at all, but simply about practices all along. This would reveal that there had never really been any dispute (at least up until the question of how to conceive of the ‘force’ that practices exert on things). The problem with this is that people who do appeal to this hybrid concept of norm in providing these kinds of social explanation do not do so on the basis of providing some actual theory of what kind of entity these norms are, but assume that there is a sort of familiar way we can understand the terms usage (which might then be elaborated into a theory at a later date). The problem with this is that our familiar way of understanding it is drawn from our internalistic understanding of norms as collective reasons for action.

The real issue underlying this is that just as we can interpret the behaviour of other humans, and even non-human animals, in intentionalistic terms by extending the ways we are obliged to understand ourselves and others as rational agents (thus producing folk-psychological explanations), we can interpret the collective behavioural regularities of other cultures in terms of ‘norms’ and the like by extending the ways we are obliged to explicitly interpret the norms we take ourselves to be bound by as part of a community of rational agents (thus producing what we might call folk-sociological explanations). This latter approach is what I have elsewhere called an internal perspective on other cultures than our own, as opposed to a genuinely external perspective (here).

To restate the opinion of my previous posts then, I find the Latourian approach so problematic precisely because it takes what are perhaps useful shorthand explanations of social phenomena in terms that borrow from our ordinary understanding of culture and then proceeds to hypostatize these explanations, thereby reifying the terms they borrow. This leaves us with a kind of half-arsed conception of the interaction between culture and nature, rather than a rigorous account of how culture is situated within nature, as a set of complex systems that we don’t necessarily have any better understanding of than we do geological processes.

In short, if we are going to understand what practices are properly, and thus the detailed causal structure of the social realm, then we need to pull them apart from norms, even if we must also make sure to understand the relation between them.

8. The Emergence of Norms

Now that I’ve gone somewhat further towards justifying the claim I made at the beginning, I’d now like to defend against what seems to be Levi’s other persistent objection, namely, that transcendental approaches to normativity (and thus my own) can give no account of the emergence of norms within history. To take a representative quote:-

What is required is an immanent genesis of norms without reference to transcendence. This genesis must be historical in character, rejecting the notion that norms are ahistorical and eternal, and provide an account of how these norms come into existence. Marx’s dialectical materialism was one such account. Marx doesn’t begin from the premise of ahistorical and eternal norms, but rather shows how norms are generated within historical settings or how they become available.

To respond to this objection properly it is necessary to show that it is actually concerned with two different things. On the one hand, it objects to the transcendental approach as ontologically unsound. This stems from the apparently obvious fact that norms have real effects within the world, and thus to say that there are any norms which can both have these effects, but are not part of the world, is to introduce some properly transcendent, element into one’s ontology. On the other hand, it demands that the norms governing how we should act, such as how we should reorganise our society, must be derived from the current historical situation, rather than from some ahistorical ideal. It should be noted that the first of these objections is concerned with norms insofar as they can play a role in our explanations of social phenomena, while the second is concerned with norms insofar as they can play a role in our practical reasoning regarding what kinds of political action we should undertake. These are precisely the two different sides of the hybrid concept of norm that I’ve been talking about.

Now, it should be pointed out that the latter objection stems from a commitment to Marxism that I do not share. Any objection to the effect that my position is incompatible with Marxism is no skin off of my nose. However, I think that there is more that can be said about this, and I will endeavour to do so after tackling the first objection, which rests on a commitment to ontological immanence which I do share.

The response to the first objection involves explaining further the difference between transcendental norms and ordinary norms. The former are those that we  cannot but be bound by insofar as we are rational at all. I have described this elsewhere in terms of what I call the primary bind (here), or our subjection to those norms without which we could not subject ourselves to any. What is important about these norms is that they are not instituted. They gain neither their force nor their content from any act of stipulation analogous to the politician’s making of law or the judge’s interpretation of it. This is because they are the norms governing the very processes of stipulation and interpretation themselves. One cannot interpret any norm, or institute any norm without already being bound to interpret and stipulate in the proper way. This is what the ahistorical character of transcendental norms consists in. One is always already bound by them, and this binding can be traced to no historical point of origin.

Ordinary norms on the other hand, although they come in many different forms (e.g., the norms of dinner party ettiquette and laws are different kinds of norm), are always historical in some sense, insofar as they are linked up with either historically situated practices  (such as the practices of dinner party ettiquette that have evolved in Europe over the past several hundred years) which the norm is implicit in prior to being made explicit, or to acts of stipulation and interpretation which are also historically situated (such as the signing of the Magna Carta, the passing of the US Civil Rights Act, and the subsequent judicial decisions which have interpreted the content of both).

The important point here is that there are two different ways we can approach this historicality. Firstly, we can treat it in a purely temporal fashion, and examine these practices, and the processes of interpretation and acts of stipulation they are bound up with, in purely causal terms. Here the history of the development of human social structures is viewed in precisely the same way as the history of the earth’s geological development. This external view is the kind of view of history I believe DeLanda argued for in A Thousand Years of Non-Linear History. Secondly, we can view historical practices as containing implicit norms, and the connection to our current practices as tying us to those norms, and we can see the various points of stipulation and explicit interpretation as providing material for and positions within a larger historical discourse on the interpretation of those norms. This is the internal view of our cultural history, in which we see our forebearers as interlocutors, and rational agents to be interpreted, rather than causal systems situated within wider social mechanisms whose actions are to be causally explained.

To put this in as simple terms as possible, the former perspective looks at what we have done in order to develop explanatory resources to predict what we will do, whereas the latter perspective looks at what we have done in order to build an interpretative case for what we should do. The awkward point is how to understand what we has been said about what we should do, i.e., the explicit discourses upon norms throughout history. The former perspective treats what has been said as a special case of what has been done, and tries to understand the special mechanisms through which this has a causal impact upon what was done and what will be done. The latter perspective treats what has been said as having a content that must either be interpreted, or as making a move within the continuing interpretative debate which is potentially relevant (or sometimes both).

What we have then, is the outline of an account of how practices within which norms are implicit emerge historically, and how explicit talk about norms  (including interpretational processes that make implicit norms explicit) emerges historically. Moreover, both these things are both phenomena within the world, and play affective roles within the causal genesis of other real phenomena within the world. At no point does anything like ‘norms’ need adding to this causal story, and indeed, as I’ve taken pains to show above, if it did it would mess up the very form of our explanation. Norms don’t exist in any real sense, even though they by necessity play a role in our practical reasoning, and thus in our internalistic understanding of the way we act both as rational individuals and rational groups.

Now, it would be legitimate at this point to ask what is the relationship between ahistorical transcendental norms and the historically situated practices that manifest them. The response to this question is as follows. There are of course practices of rationality (which are most likely maintained through a mix of social correction and common biological capacities) which have emerged within history (through a mix of social and biological evolution). We have dispositions to relate to ourselves and others in certain ways, as well as dispositions to give and ask for reasons for our commitments, and to institute and interpret norms. The important point here though is that direct appeal to these norms, in interpreting them or trying to make them explicit, does not and cannot appeal to the kind of historical facts that other kinds of interpretation can. The link between how we do think and act and how we should think and act is weakest at this point. This fact was grasped well by both Frege and Husserl in their critiques of psychologism – one cannot deduce how we should think from how we do think.

Precisely how the interpretation of these transcendental or fundamental norms of rationality is to proceed, if it cannot proceed in the more ordinary way, is another question. It is a very important question that I am currently working on, as to answer it is to work out the precise methodological structure of fundamental deontology. I won’t say anything more about it here though. Regardless of all this, the important point is that whatever these transcendental norms are, they play no active role in the production of phenomena within the world, even if our practices of rationality do.

9. Conclusion: Against Teleology

In concluding, I will return to the second part of the last objection, namely, that my position cannot allow for the norms governing how we should act to emerge immanently from the given historical situation. As I mentioned above, the motivation for this objection seems to be the assumption of the correctness of a roughly Marxist position. As I also noted, I’m not a Marxist, nor for that matter am I a Marx scholar. The fact that I don’t know Marx particularly well often makes me hesitate when discussing political philosophy, but I will take a stab at it nonetheless.

Now, as I mentioned at the very beginning, I do not think that all norms that are relevant for political action are transcendental. I think that the process of working out what we should do politically is dependent upon our understanding of the historically articulated social situation we find ourselves in. However, it is important to realise that this necessity stems from several different sources. Firstly, if we are to properly interpret the content of the norms that we already take ourselves to be bound by, then we must have as good an understanding of the practices that these are implicit in as possible. Secondly, if we are to work out how it is that we are to act in accordance with these norms, i.e., in instrumental terms, then we must have as good an understanding of the resources and obstacles our situation presents us with. Thirdly, if we are to explicitly revise some of these norms, then it will be through the previous two sources, by providing an interpretation of a more fundamental norm we are governed by on the one hand, and then showing that instrumental factors require that we either  change some other instituted norm, or require the institution of a new one, on the other. This process of revising and extending the instituted norms (as opposed to transcendental norms) that guide our actions is thus doubly dependent upon a proper understanding of the concrete structure of our social situation.

What is interesting about the position from which the second part of the objection arises is that it seemingly demands something more than this sensitivity to the historically articulated social situation I’ve just suggested. This means that it demands more than that interpretation of norms already guiding our action be sensitive to the concrete structure of our practices, more than that our instrumental reasoning upon how to satisfy those norms be sensitive to the real structure of our situation, and more than that the revision of these norms be sensitive to both.

What it seemingly demands is that there be norms of action that are immanent within our social structures other than those implicit within the ways we do actually act, norms which show us how we should reconfigure these very practices. Moreover, it seems that the process through which we uncover these norms is something other than either of the three mentioned above. It is a process through which we read off how things ought to be from how things are. As such, the position underlying the latter objection seems thoroughly teleological.

Now, insofar as this teleology is a historically inflected teleology (a la Marx), it is not as simple as crude natural law accounts of norms (influenced by Aristotle). However, that the norms immanent in social structures change (perhaps through playing a role in the development of the social structures themselves) does not make them any less problematic. The difference is analogous to the difference between Aristotelian accounts of biological kinds which understand telos in a static way, and more modern dynamic teleological approaches to biology which attempt to integrate evolutionary theory. In the end, as Kant shows in the third critique, such approaches still understand biological functioning in terms of an analogy with practical reason. Both the historical and evolutionary forms of dynamic teleology still take the form of ascribing to their object (be it a social system or a biological one) an end, the content of which is achieved through interpretation rather than description. This end can thus at best play a regulative role in explanations of the system’s behaviour.

However, there is also a disanalogy between this kind of historical teleology and the dynamic form of biological teleology. This is the fact that the kind of historical teleology discussed above does not wish to find the norms governing the way the social system currently functions, but those governing how we should act so as to change its functioning. The sense in which they try to derive how things ‘ought to be’ from how things are is thus much more ideal than that of biological teleology, because there is a much greater gap posited between them. By analogy, we might say that this approach is like instead of trying to understand how a heart should function on the basis of how it does, and then trying to aid it in perfecting or optimising its functioning, trying to understand how some superior (or evolved) heart would function on the basis of how the current heart functions, and then trying to change the current heart into the superior one.

In truth, if we were to actually engage in some kind of forced evolutionary redesigning of organs, we’d proceed in the same rational way I’ve been advocating with politics. We’d either take our understanding of what the heart as it currently is should do (say, pump blood around the body at a variable rate of speed), on the basis of what it does do, and via instrumental reasoning work out a better way to carry out this function, or we’d locate some more fundamental norm (say, delivering oxygen from the lungs to cells), and on the basis of instrumental reasoning work out a better way of satisfying this that makes the lower level function redundant. In no case do we leap from norms of current functioning to ideal norms of functioning without proceeding via a more detailed process of interpretation and instrumental reasoning (potentially involving norm revision). To proceed any other way is to act as if the superior evolutionary form is already present in a way that can be easily described. This is as strange a proposition in the study of history as it is in evolutionary biology.

To return to what I said at the beginning of this post, there is no master discourse which might give us the Good on a silver platter, and making the  Good historically relative does not make this idea any more acceptable. We can only get at the Good piecemeal, and that means engaging in practical reasoning about what we should do and how we should do it with regard to specific issues. That the form of such reasoning, both individual and collective, is transcendental, does not provide us with any short cuts in this regard, but it might nonetheless provide us with some interesting resources for pursuing it.