1 Introduction

1First of all, I wish to thank warmly those who have contributed to this forum, for their comments and criticisms. In what follows, I will try to reply, as best I can, to the latter. I am, however, well aware that my paper is tentative and in many respects weak. What matters most to me, then, is having the opportunity to learn from my critics, and to record and, when possible, follow, their suggestions.

2 The social dimension of pre-conventions

4Arena (2016: 59): “[Celano’s] explanation of embodied conventions needs to be further developed in order to account for the social dimension of conventionality”.

6This gap is filled if, as Brigaglia (2016: Sec. 1) suggests (the suggestion is, however, adumbrated also in Arena’s contribution), we include in the definition of the concept of pre-convention satisfaction of what I have elsewhere (Celano 1995) called the “dependence condition” – i.e., where a convention is in place, each individual who is party to the convention behaves as she does because the others behave in that way.

7This condition, however, may be construed in different ways, according as to how the “because” is understood.

9Which way? The “because” in the dependence condition may be understood – and this is, I think, the main bifurcation (Celano 1995 & Celano 2014) – either as a reason-relation (paradigmatically, strategic reasoning, on the background of shared conditional preferences for conformity) or as a causal non-intentional relation: each individual is caused to behave as he does by the fact that the other individuals who are party to the convention behave in that way. (The main way in which this may happen is by sheer imitation.) In the case of pre-conventions, the latter is, it seems to me (as suggested by Brigaglia 2016: Sec. 1), the proper understanding.

10Thus, the correct front crawl stroke is pre-conventional (to the extent that it actually is) because, although it is wired in the body of any expert swimmer, each expert swimmer swims that way because (training, imitation) the others do. I do not mean that it cannot happen that somebody swims the crawl because he deliberately decides to do this (maybe as a way of pursuing a further end), or that he deliberately swims as he does because the others do. This is, of course, possible. What I mean is that what the correct front crawl stroke is – its identity –, is not determined, if not unintentionally and, thus, obliquely, by such decisions. In this sense the correct front crawl stroke is not the output of a train of thought.

11These considerations lead us to the issue raised by Figueroa Rubio, namely, what kind of explanation is appropriate to the coming into existence of a pre-convention. For reasons that will be apparent in due course, however, I shall discuss this later (Sec. 5 below).

3 Pre-conventions and ordinary habits and dispositions

12One objection is raised by many contributors, albeit in different guises: what I call “pre-conventions”, or at least most of them (and some of the examples I offer are, some of my critics contend, of this kind: riding a bike, marching), are nothing more than behaviours which, although they have now become, in expert practitioners, fluid, effortless, unreflective and automatic (a second nature), are the result of the internalisation, through practice, of sets of rules. This may be an interesting phenomenon – so the objection goes – but there is nothing so terribly new about it, nor does it follow that there is anything inherently normative in these behaviours. They are, trivially, behaviours conforming to given rules. This in no way calls into question the orthodox distinction between fact and norm. In short, what I solemnly call “pre-conventions”, and treat as if they were an interesting and revealing “new” phenomenon, are nothing but, trivially, ordinary habits and acquired dispositions.

13Two different issues, in fact, underlie these complaints.

15Second, my “pre-conventions” are nothing but, trivially, habits and dispositions which are acquired as a consequence of repeated practice of a rule-governed activity.

3.1 On the difference between living in an apartment on the fourth floor and not being an earthworm

17Arena (2016: 59):

• 7 Sánchez Brigido (2017: Sec. 3): “not all the examples that Celano proposes are examples of [‘basic’ (...)

“the relationship between the unintentionality of embodied conventions and the Background is still imprecise, and /…/ this renders the notion of pre-conventions unstable”: “Irrespective of the fact that Celano, at one point, defines pre-conventions as ‘conventions that are mostly in the Background of our activities and thoughts, and that, passing usually unnoticed, delimit their spaces’, it seems to me that, from the examples put forward, it follows that not all embodied conventions are a fragment of the Background. The examples given, such as conventions that establish how to swim the crawl or how to march, are, to my mind, on the surface” (Arena 2016: 62).7

“[Celano] describes [pre-conventions] as involving convergent behaviour ‘which is not a biological regularity, and is the result of learning, but which is also automatic: it is spontaneous (unreflective), rapid, fluid, effortless.’ He also refers to them as conventions that ‘become as natural as breathing, a ‘second nature’.’ If this is intended to distinguish pre-conventions from Lewis-type conventions, then it strikes me as a mis-step. As Celano acknowledges, the following of a Lewis-type convention can also become ‘second nature’, and his attempt to show how this differs from the way in which pre-conventions are second nature returns us to the idea that the former, but not the latter, is supported by a ‘train of reasoning’.”

“[In] the case of those who follow a Lewis-convention which has become for them ‘second nature’ we can say, of course, that what they are doing is implementing a train of reasoning that is ‘within them’, in a tacit and non-articulated way /…/ But it is unclear what that might mean, if not that they behave as if they were doing this. But this appears to be question-begging. When the convention has become ‘second nature’ what guides their conduct is, in some sense, what they actually do”: “one should distinguish between the sense in which a ‘train of reasoning’ supporting a Lewis-convention can originally – i.e., when the convention comes into existence – be called ‘tacit’; and the sense in which it can be called ‘tacit’ when the convention has become ‘second nature’” (Celano 2016: 13–14).

3.2 “Pre-conventions” nothing but ordinary habits and acquired dispositions

21Duarte d’Almeida (2017: Par. 7):

• 16 Chiassoni (2017: Sec. 3): “a pre-convention is a convergent behaviour of some group of people, and (...)

“So do pre-conventions ‘exist’ /…/ ? There do seem to be many actions which are (a) not biological regularities, (b) the outcome of learning processes, and (c) actions that we do come, with practice, to perform automatically and spontaneously. It also seems plausible to think that arbitrary conventions may and do feature prominently in such learning processes, and that, with practice, such conventions do ‘recede’ into the ‘background’ and come to pass ‘unnoticed’. So the answer would appear to be that, yes, pre-conventions do exist”.16

22This, however, is not “particularly striking”, says Duarte d’Almeida. We knew already that such a thing may happen. There is no mystery about these phenomena.

23Or is there? In discussing my claim that pre-conventions are normative facts, Duarte d’Almeida asks:

“what does [Celano] give us by way of argument to support such a view? Not very much, as far as I can see. He does avail himself of several metaphors, and he is particularly fond of Searle’s ‘the body takes over’ phrase. But he goes on, I think, to mischaracterize Searle’s point: ‘The central idea is aptly captured by the phrase the body takes over. What the body is doing, from now on, is not a mere de facto regularity, but something that is in between a norm and a regularity’ (Celano 2016: 15)”.

24Why is this a mischaracterisation? In the relevant passage, says Duarte d’Almeida,

“Searle is offering an account /…/ of how it is that we come to acquire and develop such physical skills and to perform many actions, as we do, automatically: ‘repeated practice and training in a variety of situations eventually makes the causal functioning of [explicit] representation [e.g. of verbal instructions of what one is supposed to do] unnecessary in the exercise of the skill’ (Searle 1983: 151). Searle’s point, then, is an explanatory point about how such types of behavior come to emerge, and about the role – the causal role – played by rules, and by rule-learning, in that process”.

25I will deal with the “no argument” charge later (below, section 10). But what about the mischaracterisation charge?

26It is true that I am particularly fond of this passage of Searle’s. My comment about it, quoted at length by Duarte d’Almeida (and reproduced in the quotation above: my “mischaracterisation” of Searle’s point), was meant as an explanation of Searle’s claim that with practice “the rules do not become ‘wired in’ as unconscious Intentional contents”, and, specifically, as a gloss on some of the lines in Searle’s text, omitted, for reasons of space, in the quotation (that also Duarte d’Almeida reproduces) contained in my paper. According to Duarte d’Almeida, it seems, Searle is claiming, in the passage I am fond of, that, with practice, the very same behaviour that a beginner holds as a consequence of trying to follow the rules is held, by the advanced skier, automatically and unreflectively. I will now restore the relevant lines from Searle’s text, motivating my alleged mischaracterisation.

27This is the quotation contained in my paper, reproduced by Duarte d’Almeida:

“as the skier gets better he does not internalize the rules better, but rather the rules become progressively irrelevant. The rules do not become ‘wired in’ as unconscious Intentional contents, but the repeated experiences create physical capacities, presumably realized as neural pathways, that make the rules simply irrelevant. ‘Practice makes perfect’ not because practice results in a perfect memorization of the rules, but because repeated practice enables the body to take over and the rules to recede into the Background /.../ On my view, the body takes over and the skier’s Intentionality is concentrated on winning the race” (Searle 1983: 150-151. Italics are mine).

28And here are, restored (in italics), the lines my alleged mischaracterisation was meant to be a gloss to:

“repeated practice enables the body to take over and the rules to recede into the Background. /…/ The advanced skier doesn’t follow the rules better, rather he skis in a different sort of way altogether. His movements are flowing and harmonious, whereas the beginning skier, consciously or unconsciously concentrating on the rules, makes movements which are jerky, abrupt, and inept. The expert skier is flexible and responds differently to different conditions of terrain and snow; the beginning skier is inflexible, and when different and unusual situations come up he tends simply to fall down. A downhill racer on the course moves very rapidly, over 60 miles an hour, over a terrain that is rough and uneven. His body makes thousands of very rapid adjustments to variations in the terrain. Now which is more plausible: when his body makes these adjustments, it is only because he is making a very rapid series of unconscious calculations applying unconscious rules; or is it rather that the racer’s body is so trained that these variations in the terrain are dealt with automatically? On my view, the body takes over and the skier’s Intentionality is concentrated on winning the race” (Searle 1983: 150-151).

30So, the reason why I am particularly fond of Searle’s 1983 key passage about what happens, in learning how to ski, when the body takes over, is this. In practising an activity according to explicit rules the body itself, thanks to repeated practice, little by little understands what is the right way of performing the relevant actions. In doing so, the body’s understanding goes well beyond what the rules say. The body develops its own criteria of correctness in performing the action (this is what its “knowing-how” consists in). Rule-guided behaviour becomes norm-guided behaviour (in Brigaglia’s terminology). The relevant norms are there, in the advanced skier’s actual, regular, behaviour. This is also an explanatory point, but it is there, in the advanced skier’s behaviour, that the justification for behaving in precisely that way lies.

32Duarte d’Almeida also very briefly mentions and criticises my – itself sketchy – reconstruction of the main (as far as my purposes are concerned) lines of Searle’s argument in chapter 6 of The Construction of Social Reality (1995), entirely devoted to a discussion of the thesis of the Background, and, specifically, my reading of a second key passage of Searle’s (which I am also very fond of; Searle 1995: 141-142) drawn from it – the topic of section 7 of my paper. Which, I claim there, allow us to conclude that the Background of intentionality, comprises (also) normative facts, as I understand them in my paper.

33This is not the place for engaging in a detailed exegesis of chapter 6 of Searle’s 1995 book. It is a rich, intricate text, implying, I think, more than what is explicitly stated, and raising awkward puzzles. So, I will simply reproduce Duarte d’Almeida’s critical remark, and comment on it. Here is Duarte d’Almeida’s remark (2017: Par. 12–13):

“It may be true that, as Searle points out (and as Celano points out that Searle points out), participants in institutional activities develop tendencies, dispositions and skills that are ‘functionally equivalent’ to the systems of rules of the relevant institutions: ‘The basic idea /.../ is that one can develop, one can evolve, a set of abilities that are sensitive to specific structures of intentionality without actually being constituted by intentionality. One develops skills and abilities that are, so to speak, functionally equivalent to the system of rules, without actually containing any representations or internalizations of those rules’ (Searle 1995: 142). But this, again, despite what Celano suggests, is not the same as saying that ‘here /.../ entities intermediate between rules and regularities, embodied norms /.../ emerge’ (Celano 2016: 29). We should avoid conflating (a) claims about how it is that pre-conventions may come to emerge (and the role that arbitrary rules may play in that causal process), (b) claims about the ‘functional’ role of pre-conventions, and (c) claims about their metaphysical status”.

34My comments:
(1) Duarte d’Almeida, it seems to me, treats Searle’s discussion in chapter 6 of the 1995 book as making the same point as the key passage about what advanced skiers do in his Intentionality (1983: 150-151), discussed above. There is however, a deep difference between these two texts of Searle’s. The latter is a description of the kind of phenomena I am concerned with, the former an articulated theoretical discussion, more sophisticated than the treatment in the 1983 book. What is described in the 1983 passage is extensively theorised in the 1995 book. It seems that, in the passing of time, the topic – the Background – has become more and more intriguing, complex, and theoretically problematic in the eyes of the author, leading him to progressively more complex, sophisticated, and problematic conclusions about it (in this connection, the turning point is, I think, chapter 8 of Searle’s 1992 book). This is why, on the one hand, I included the 1983 passage in the section of my paper devoted to offering examples of the kind of phenomena I am interested in, and, on the other hand, I have chosen the 1995 argument as one of the four non-original arguments which I offer in support of my claims. Neither my sketchy reconstruction, nor Duarte d’Almeida’s critical remark about it, do justice to the complexity of Searle’s 1995 argument, and especially to the puzzles raised by it.

37 (4) Given this, it seems to me wrong, again, to maintain that Searle’s point in this chapter of the 1995 book is only an explanatory one. It is, no doubt, also an explanatory one (the main question is, what the causal role of rules constitutive of institutions is). The Background works as a kind of causation. But what we need, says Searle (1995: 141), is “a type of causation” which has “a rational structure”.

39 (6) Duarte d’Almeida rightly reminds me that “we should avoid conflating (a) claims about how it is that pre-conventions may come to emerge /…/, (b) claims about the ‘functional’ role of pre-conventions, and (c) claims about their metaphysical status”. There is a misunderstanding, here, due to my inaccurate choice of words. Duarte d’Almeida’s warning is prompted, as we have seen (in the quotation above), by this passage of mine, which in turn immediately follows the reproduction, in my paper, of Searle’s 1995 key passage: “here /.../ entities intermediate between rules and regularities, embodied norms /.../ emerge”. The misunderstanding is this: my remark was not meant as a remark about the emergence of the sort of normative facts I am interested in (the Background abilities and dispositions Searle is talking about). It was, rather, a remark about the (so I claim) emergence in Searle’s argument of the notion of such a phenomenon. What I wrote is, in fact, “this [meaning, Searle’s key passage] is the crucial step. Here [meaning, at this juncture of Searle’s argument], once again [meaning, as we have seen already in the case of Lewis, Goodman, and Wittgenstein], entities intermediate [meaning, the notion of entities intermediate] between rules and regularities, embodied norms /…/ emerge”. I grant that this was, at best, equivocal. The clarifications I am now adding, in square brackets, come too late.

40 (7) Duarte d’Almeida rightly reminds me that “we should avoid conflating /…/ (b) claims about the ‘functional’ role of pre-conventions, and (c) claims about their metaphysical status”. This makes good sense, but how is it supposed to impinge on my account? We should avoid forgetting that, sometimes, claims about the functional role of X have implications concerning the “metaphysical status” of X. And this seems to me a case of this sort.

41(8) There is a suggestion, in my paper, concerning Searle’s argument in the 1995 book, which none of my critics, as far as I can see, has commented upon. Here it is in italics: “This is the crucial step. Here, once again, entities intermediate between rules and regularities, embodied norms (or, if you will, the pineal gland), emerge”. The point of this reference to Descartes’ often ridiculed escamotage for relating intentionality (cogitationes) and material causes (res extensa) was precisely to suggest that Searle’s remarks about non-intentional structures, neurophysiological factors, which are “sensitive” to structures of intentionality (specifically, to rules), and as “functionally equivalent” to the latter, are no less puzzling. That is to say, we are still in need of a perspicuous elucidation of this phenomenon. Once again, it seems, we are stuck with the pineal gland, or something akin to it.

4 Mistaken examples

42Many critics complain about some of my examples.

43Some of these remarks relate to behaviour that starts as the conscious, deliberate attempt to follow explicit rules, or Lewis-conventions, that becomes, in the passing of time, second nature. I dealt with this kind of cases in the previous section. There are also, however, objections specifically directed at my choice of examples.

45One of the objections, however, is precisely to the point.

46Smith (2016: 72):

“it is plausible that playing chess well involves both abiding by certain conventions and a certain know-how that cannot be reduced to a set of rules. However, it does not follow that the conventions can be equated with that know-how, so as to entitle us to say that the conventions are not reducible to a set of rules”.

47Smith is right. What I wrote is, at best, badly misleading. It implies that conventions such as those defining the basic moves allowed in the game, and which may easily be listed in the first few pages of an handbook, are – not, ordinary conventions, but rather – pre-conventions. This is a mistake.

48“Indeed [he goes on arguing], it is tempting to say that the conventions of chess are (by and large) confined to the rules of chess, and that the know-how involved in being able to play chess well is non-conventional. For one thing, that know-how need not involve any form of agreement, even in Celano’s extended sense of ‘agreement’. We can imagine a community in which only one person possesses that know-how; everyone else plays chess poorly” (Smith 2016: 72).

5 Reasons and causes

50The notion of a pre-convention as a normative fact, as I more or less precisely (on the concept of normative fact, see next section) characterise it, calls into question the distinction between reasons (from now on, by “reason” I mean, unless otherwise specified, justificatory reasons) and causes. The distinction between facts and norms does not, of course, identify with the distinction between causes and reasons. The two distinctions, however, are tightly connected. Causes are facts; reasons translate into “oughts”. Specifically, pre-conventions (and most likely other items in the Background) explain behaviour, and they also guide it – they, thus, also justify it (when it conforms to the pre-conventional regularity).

51Unsurprisingly, many of my critics find the idea that some causes may be reasons problematic, if not a blatant category mistake.

52Moreso (2016: 78):

“the Ontological Commitment Thesis [i.e., pre-conventions are abstract entities] and the Normative Bite Thesis [i.e., “pre-conventions have a normative dimension, they allow us to assess whether certain human actions are right practices of skiing or riding a bike”] are together both incompatible with the idea that pre-conventions are facts which have causal powers in human behaviour. And if we abandoned the ontological thesis, we could not obtain normative determination. That is to say, either pre-conventions (as part of the Background) are part of our causal explanation of human behaviour or pre-conventions are abstract entities able to determine human behaviour normatively. [Footnote omitted] In the first case, pre-conventions lack normative meaning, while in the second pre-conventions cannot integrate our causal explanation of human actions. Tertium non datur”.

53Sánchez Brigido (2017: Sec. 3):

• 26 This is not meant, by Sánchez Brigido, as expressing his own views. It is rather intended to point (...)

“the problem is how it is possible that there are facts (something that belongs to the aspects of the world governed by relations of causality understood very broadly) which are normative (a domain to which causality /…/ does not belong)”.26

54On this score, I plead guilty. (That the distinction between context of discovery and context of justification cannot be maintained all the way down, however, is no news.) What I, following the lead of my authorities, have tried to show is this.

55In principle, and as a rule, we should distinguish carefully between causes and reasons, to the maximum possible extent. However, in articulating the architecture of our reasons for action, and if we dig deep in the reasons we have for behaving the way we do, we reach bedrock: at the end of the day, we cannot but cite facts (in the case of pre-conventions, regularities of behaviour) which both explain and justify our behaviour. Facts like, for instance, the fact that, often, we happen to notice the same analogies, or that, in inductive inferences, we happen to project the same properties (section 4 of my paper); the “practice” – at bottom, participation in a shared form of life – which underlies our mastery of a concept and our following a rule (section 5 of my paper); the non-intentional behavioural data constituting the Background against which intentionality, and thus normativity, only works (i.e., it determines conditions of satisfaction) (section 6 of my paper). These are, indeed, facts, playing the role of causes in explanations of our behaviour. They are, however, facts which manifest themselves when we follow the thread of our reasons. These facts also guide our behaviour, that is, they also fix, they determine, the right way of doing something. They are, in this sense, normative facts.

56Before discussing this notion (in the next section), however, two further remarks are in place here.

58The alternative Figueroa Rubio poses is “whether what fits better is intentional explanation or causal explanation, or whether perhaps there is some other kind of explanation which awaits us so that we can explain pre-conventions” (Figueroa Rubio 2017: Sec. 1). This distinction should be understood thus:

“on the one hand, intentional explanation appears with agency. When some behaviour is understood to be an action, we attribute a mind to the being whose behaviour has been identified. In doing this, we attribute mental states and thoughts (e.g., propositional attitudes, such as beliefs, expectations and desires) to the agent and we represent the behaviour as governed by reasons. On the other hand, causal explanation appeals to the relationship between events (or facts), focusing on the physiological aspect of behaviour. Causal explanation shows the behaviour to be governed by laws of nature” (Figueroa Rubio 2017: Sec. 2).

59I find this alternative misconceived. On any plausible account, mental states are causes of human, and non-human animals’, behaviour and action. Intentional explanation is a species of causal explanation. For the alternative to have any bite we should, I think, understand by “reasons” justificatory reasons. These, in turn, are to be understood as the possible contents of mental states. Thus, the issue is not, in fact, the distinction between two different kinds of explanation. It is rather, once again, the one raised by Moreso and Sánchez Brigido, namely, whether causes can be reasons (and, whether abstract entities can be causes; below, section 7). That, in providing a causal explanation of behaviour and action we understand them as part of nature, is, to me, uninformative. What else could human behaviour, or action, be, if not part of nature? Are we supposed to believe that minds are not natural phenomena?

60Explaining behaviour, or action, through laws (of nature) is another matter, of course. The notion of a law of nature is, however, so fraught with difficulties that I find it risky, to say the least, to adopt the policy of relying on it for building a dichotomy between two mutually exclusive kinds of explanation (meaning by “explanation”, as I have just determined, pointing to the causes of a given phenomenon). “Laws of nature” are a mixed bunch. Different sorts of regularities, or generalisations, fall under this heading. They are not so safe and reliable, waterproof as it were, as XIX century men took them to be. We shouldn’t, probably, think of them, or all of them, as universally quantified undefeasible conditionals. And, on the other hand, it would be odd, in the light of contemporary developments of psychology, cognitive science, and the neurosciences, to assume that explaining behaviour, or action, by appeal to mental states is different in kind from identifying natural regularities. In sum, we cannot, I think, rest content with the traditional shape of the “explanation v. understanding” (or perhaps Natur- v. Geisteswissenschaften) debate.

61(2) That some causes may be reasons of the behaviour they cause is a disturbing thought. It should be noticed, however, that there is another side – the reverse side – to the reasons v. causes divide: can justification be explanation?

6 Normative facts

63The opening paragraph of my paper (Celano 2016) reads:

“I will argue that there are entities that can be plausibly called ‘conventions’, which are neither mere de facto regularities, nor rules (norms), but that – in a sense to be specified – have both the character of de facto regularities, as well as a normative character: they are, literally, ‘normative facts’. This paper attempts to isolate these entities”.

64In this passage, Duarte d’Almeida detects three claims. The first is the claim “that ‘there are entities’ of a certain sort”. What concerns us now are the second and the third. Duarte d’Almeida apparently finds both of them obscure. Happily, however, the third claim, suitably rephrased, helps us in dealing with the second (it “helps readers to understand the second claim” [i.e., the claim that pre-conventions “are not de facto regularities, and that they are not rules or norms”] more clearly”, he says). The third claim, Duarte d’Almeida writes, is the claim that

“those entities [i.e., pre-conventions] have both a normative character and the character of de facto regularities: which must mean, I gather, that what I called the ‘second’ claim is really to be understood as the claim that those entities are neither de facto regularities as opposed to norms, nor norms as opposed to de facto regularities” (Duarte d’Almeida 2017: Par. 1).

65That’s right. This is why I inserted the word “but” between the second and the third theses. (I do not mean that I did it on purpose, it just came to me naturally.) Duarte d’Almeida (2017: Par. 3) further remarks that

“the claims are also not all equally important. The third is philosophically interesting, and Celano knows it. That is probably why he pitches his essay as an attempt to show that the ‘entities’ he is concerned with, his ‘pre-conventions’, are entities of a ‘peculiar sort’ (as he says in his abstract) that somehow straddle the commonly accepted distinction between facts and norms (Celano 2016: 9-10)”.

67The claim that pre-conventions are, in this sense, “normative facts” obviously raises difficulties. Chiassoni (2017: Sec. 2) rules out, or so it seems, the very possibility of

“mysterious entities, which would simultaneously partake in both the factual and the normative. /…/ We can always and profitably keep apart, for any individual event whatsoever, the factual and the normative”.

70This is plainly true. I rely, in my paper, on an intuitive understanding of what a fact, or a norm, is. Sánchez Brigido is right in suggesting that a fully-fledged treatment of the phenomena I am concerned with requires an extensive discussion, and analysis, of these notions.

72And, building on this, the notion of normative fact which I have used seems to me an intelligible and relatively well-determined one. (Which is not to say, of course, that it is a coherent notion, or that it is intuitively plausible that there actually are normative facts, so understood.) A normative fact, in this sense, is a fact that fixes, determines, the right way of doing something, and in this way guides behaviour.

73“Normative facts – Sánchez Brigido (2017: sec. 3) further claims – can /…/ be understood in two ways. A normative fact is either a fact that engenders normativity in some way or an item with a sui generis nature. Either it has a nature of its own, something which does not belong to the domain of facts or to the domain of norms, or it belongs to one of these domains and is somehow related to the other”. He goes on saying that

“on the one hand, Celano should clarify whether he is claiming that facts determine (or give rise to, ground, fix, etc.) norms, which presupposes that they are different items (facts and norms) related in some way, or whether they have a sui generis nature. Celano is ambiguous about this” (Sánchez Brigido 2017: Sec. 3).

7 Can abstract entities be causes?

76Suppose that X says to Y “Please, take five oranges from the bowl on the table in the kitchen, and seven oranges from the fridge, and bring them all in the dining room”. When Y then shows up in the dining room with twelve oranges, it is difficult to maintain that the number 5, the number 7, and the fact that 5+7=12 should play no role in the explanation of Y’s behaviour. An explanation should, of course, mention Y’s mental states: it is because he has representations of the numbers 5 and 7, and he believes that 5+7=12, that he shows up with twelve oranges. It is exactly because his mental states have these contents, however, that he acts this way. We cannot, it seems to me, eliminate the numbers themselves, and their relations, from the explanation.

77This is just to voice my perplexity about the general claim that abstract entities can play no causal role - specifically, no role in the explanation of behaviour. What concerns us here, however, is this claim as related to reasons for action (under the assumption that reasons for action are abstract entities).

78This is yet another facet of the reasons v. causes issue (above, section 5). Moreso clearly sees the problem. He claims that, since abstract entities can play no causal role, pre-conventions and similar phenomena, being (I claim) abstract entities, cannot explain behaviour. And, since reasons for action, if they are to have any normative bite (i.e., fix the right way of doing something), must be abstract entities, then either pre-conventions (as I also claim) play a role in explaining behaviour, having, however, no normative bite, or they do have normative bite, but in this case they can play no role in explaining behaviour.

79This is Moreso’s dilemma. My reply is twofold.

80First, I shall once again rely on Williams’ point (above, section 5). X is a fit candidate for being a reason for action only if it is in principle capable of explaining action.

81Second, Moreso himself (2016: 80) notes that his considerations “leave the problem of intentionality untouched: how are we, thinking bodies, moved by numbers, meanings, possible states of affairs, reasons for action?”. So, in fact, he claims that we are moved by numbers or reasons for action – which is to say, I think, that such entities play a role in explaining (i.e., they contribute to causing) our behaviour. What Moreso here mentions as a further problem, calling it “the problem of intentionality”, is, then, premised on the negation of one of the premises of his dilemma (i.e., that abstract entities cannot be causes). And it is, in fact, the same problem that his dilemma is intended to solve (albeit negatively), and that, I think, should lead us to the recognition that there are entities (among them, I claim in my paper, pre-conventions) that are both reasons and causes.

82It doesn’t seem to me that Frege’s idea of the mind’s “grasping” “thoughts”, or Husserl’s idea that the object of our mental states are “noemata”, are less mysterious than the idea that abstract entities (as, for instance, the front crawl stroke, or phonemes) can be causes of our behaviour.

8 On the relationship between pre-conventions and Lewis-type conventions

83Smith (2016: 71): Celano “suggests that pre-conventions make the existence of Lewis-type conventions possible” (“pre-conventions are needed for the existence of Lewis-type conventions”); but, Smith argues, first, “if pre-conventions are needed, it is only at the formative stage of the development of the Lewis-type convention”, and, second, they are probably not needed at all, not even at this formative stage.

85According to Smith, the relevance of this fact is confined to the formative stage of a Lewis- convention (“it is important to recognise that the issue here concerns how Lewis-type conventions are established, not how they are maintained once they are established”).

86I disagree. It is true that according to Lewis, as Smith rightly emphasizes, as the number of precedents increases, ambiguity decreases, until finally a “regularity” (1969: 38) of behaviour settles. Does this mean, however, that the fact that “we happen uniformly to notice some analogies and ignore others – those we call ‘natural’ or ‘artificial’, respectively” doesn’t play a role any more, or even that its importance is reduced? I do not think so.

87Generally speaking, coordination by convention is, according to Lewis, coordination by precedent. Coordination by precedent is, in turn, convergence of the involved individuals on a coordination equilibrium which is, for them, salient, and

“salience in general is uniqueness of a coordination equilibrium in a preeminently conspicuous respect. The salience due to precedent is no exception: it is uniqueness of a coordination equilibrium in virtue of its preeminently conspicuous analogy to what was done successfully before” (Lewis 1969: 38; emphasis mine).

88This point, is, in Lewis’ account, perfectly general. Specifically, it doesn’t apply only to the “formative stage of the development of [a] Lewis-type convention”, to use Smith’s phrase.

89A conventional regularity of behaviour exists – it comes into existence, and it continues to exist - when we (the involved individuals) are all

“acquainted with a class of previous coordination problems, naturally analogous to our present problem and to each other, in which analogous coordination equilibria were reached. This is to say that the agents’ actions conformed to some noticeable regularity. Since our present problem is suitably analogous to the precedents, we can reach a coordination equilibrium by all conforming to this same regularity” (Lewis 1969: 39; emphasis mine).

90Or, in other words, “coordination by precedent” is

“achievement of coordination by means of shared acquaintance with a regularity [footnote omitted] governing the achievement of coordination in a class of past cases which bear some conspicuous analogy to one another and to our present coordination problem” (Lewis 1969: 41; emphasis mine).

91And, finally, a convention exists, and continues to exist, when the individuals involved extrapolate from precedents to present and future cases (hence the connection, in my paper, to the argument about induction). How does this work? Here’s Lewis’ answer:

“Given a regularity in past cases, we may reasonably extrapolate it into the (near) future. For we are entitled to expect that when agents acquainted with the past regularity are confronted by an analogous new coordination problem, they will succeed in achieving coordination by following precedent and continuing to conform to the same regularity” (Lewis 1969: 41; emphasis mine).

92So, I do not think that, as Smith (2016: 74) claims, “if pre-conventions are needed, it is only at the formative stage of the development of the Lewis-type convention”.

9 What can pre-conventions do for legal theory?

95On this score, I said nothing in my paper, and I have nothing to say now. Some of the contributors, however, put forward their suggestions.

96Chiassoni (2017: Sec. 3):

“jurists can learn a lot from Celano’s theory of pre-conventions. In my view, the learning mainly consists in getting hold of two points. First, legal experience surely does contain legal pre-conventions, which are worth discovering and bringing to the fore. Legal sociologists, psychologists and anthropologists should set to that valuable task. Second, the practice of law, in its everyday workings, takes place on a background of extra-legal pre-conventions, above all and insofar as language is involved”.

98This is, indeed, what Segatti claims. He puts forward two suggestions. If there are, in fact, pre-conventions (as I understand them in my paper) then

“so-called paradigm cases of law /…/ lose much of their centrality. To study pre-conventions or norm-conforming behaviour [Brigaglia’s notion], we also need accounts of situations in which relevant agents use criteria for identifying relevant legal phenomena that only approximate valid sources of law /…/ To provide such accounts, the traditional informational focus of a philosophical analysis of law also needs, I suspect, substantial broadening: if pre-conventions are in the body, then that is where we need to look if we want to study them” (Segatti 2017: Sec. 5).

10 No argument?

100Duarte d’Almeida (2017: Par. 2) claims that the main thesis of my paper (what he calls – above, section 3 – my “third” thesis: pre-conventions “have both a normative character and the character of de facto regularities” (Section 1 of Celano 2016); they are “entities of a ‘peculiar sort’ /…/ that somehow straddle the commonly accepted distinction between facts and norms”) “is a bold thesis; but it is one for which Celano gives us no good argument”.