Abstract
The requirements of rationality are fundamental in practical and theoretical philosophy. Nonetheless, there exists no correct account of what constitutes rational requirements. This paper attempts to provide a correct constitutive account of ‘rationality requires’. I argue that rational requirements are grounded in ‘necessary explanations of subjective incoherence’, as I shall put it. Rationality requires of you to X if and only if your rational capacities, in conjunction with the fact that you not-X, explain necessarily why you have a non-maximal degree of subjective coherence.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
I shall use ‘constitutive account’ and ‘analysis’ synonymously in this paper. I am aware that this may be misleading. Analysing X is often taken as setting out the meaning of X. However, I will not use ‘analysis’ in this sense. Determining the meaning of ‘rationality requires’ is not part of my agenda in this paper. I also do not intend to give a full constitutive account of the requirements of rationality. That is, I shall not try to give an account of what a requirement of rationality is. My aim here is more limited. I seek to provide a general account of what makes it the case that one is subject to a particular requirement of rationality. This is what I shall refer to as a ‘constitutive account’ or ‘analysis’. In addition, I shall refer to the activity of establishing a constitutive account as ‘analysing’.
Furthermore, I will assume that a constitutive account or analysis can be stated in terms of necessitated and universally quantified bi-conditionals. For example, a constitutive account of being a brother may read as follows: necessarily, for all X, X is a brother if and only if X is male and X has a sibling.
Constitutive accounts or analyses can be circular or non-circular. They are circular if and only if the right-hand side of the bi-conditional refers to properties and relations that cannot be analysed without referring to properties or relations used in the left-hand side of the bi-conditional; they are non-circular if and only if the right-hand side of the bi-conditional refers to properties and relations that can be analysed without referring to properties or relations used in the left-hand side of the bi-conditional. If a correct constitutive account or analysis of X is non-circular, then we can say that it gives a reduction of X. It is my aim in this paper to draw up a non-circular constitutive account of ‘rationality requires’.
By ‘attitude proposition’ I mean a proposition that ascribes a single attitude or a combination of attitudes (or their absence) to a subject. Examples are: Ingo wants a new job; Rainer does not believe that it is raining; Janice likes ginger if she believes that it comes from ecological farming; Daniel does not want to make plans about living in California permanently unless he believes that he will work on a vineyard; etc.
I say ‘in one way or another’ because X-ing may improve: (i) the overall coherence among your attitudes, or (ii) the coherence among a specific set of your attitudes.
As I explain in detail in Sect. 8, I will assume that the type of attitudinal coherence relevant for rationality based on rational requirements consists in manifesting a disposition that is sensitive to the success conditions and constitutive aims of one’s attitudes. For example, you are attitudinally incoherent if you have a pair of contradictory intentions despite your disposition to avoid contradictory intentions precisely because they cannot jointly fulfil their success conditions. I will say more on this type of coherence in Sect. 8.
For a brief discussion of an analogous ‘necessary-condition account’ of ‘normative requirements’, see Fink (2012: 128–129).
There is a quick, yet ultimately unsuccessful, attempt at rejecting this view. To say that requirements of rationality specify necessary conditions for full rationality seems to suggest that one read ‘requires’ in ‘rationality requires’ as expressing the mere inverse of the necessary-condition relation. That is, if A requires B, then B is a necessary condition for A. However, ‘requires’ cannot be used in this way in ‘rationality requires’. If rationality requires Janice to have another shower if she believes she ought to have another shower, we cannot express this by turning it into a binary necessary-condition relation such as ‘Janice has another shower if she believes she ought to have another shower is a necessary condition for rationality’. Read literally, this makes no sense. Moreover, it is quite evident why this does not work. It fails to account for the fact that requirements ‘belong’ to someone (such as to Janice in the above example). In fact, ownership turns the ‘requires relation’ from a binary into a ternary relation.
This is not an argument against supposing that requirements of rationality specify necessary conditions for full rationality. The relation implied by a necessary condition does not have to be binary. It may be ternary instead. It is relatively unambiguous how to construe this ternary relation. When rationality requires you not to believe a contradiction, then not believing contradiction is a necessary condition for you to be fully rational. In general, at w, if R requires of S that A, then for S to have the corresponding property of R at w, A is a necessary condition at w.
I would like to thank a number of anonymous referees for emphasising this point. It has led to a major amendment of the presented arguments.
Many thanks to an anonymous referee of Erkenntnis for putting forward this analogy.
Again, I say ‘in one way or another’ because one’s X-ing may improve: (i) the overall coherence among one’s attitudes, or (ii) the coherence among a specific set of one’s attitudes.
‘Locality’ is a desideratum of rational requirements because it ensures that rational requirements reflect our ordinary judgments and attributions of irrationality (cf. Kolodny 2005: pp. 515–516). To show how the necessary-condition-for-coherence analysis preserves locality, consider a ‘preface-paradox’ type situation. Suppose you believe the conjunction of all assertions in one of your authored books. Yet, at the same time, you also disbelieve this conjunction. You thus end up with two beliefs whose conjoined contents form a contradiction. Avoiding the combination such beliefs is, I assume, a necessary condition for being fully attitudinally coherent. The necessary-condition-for-coherence analysis thus entails that rationality requires you not to have such a combination of contradictory beliefs.
This holds even when, for instance, you have excellent evidence for both beliefs. Suppose you have checked every assertion in your book and found no error. Yet knowing about your fallibility, you also have excellent evidence that it is not the case that every assertion will hold true. Given your evidential position, it might be more (perhaps even most) coherent for you to continue believing the two contradictory contents. But even so, the necessary-condition-for-coherence analysis does not ‘cancel’, as it were, the requirement not to have the two contradictory beliefs in question.
Raz (2005: 19) puts it like this: ‘Rationality consists in part in proper functioning. People who fail [for example] to pursue the means to their ends display or manifest a form of malfunctioning criticisable as a form of irrationality’.
See also Price (2008: 86). Surely, if you are severely irrational, you may not come to believe this. One may argue instead that you must be disposed to believe that you have conclusive reasons.
‘[W]e ought … to use ‘irrational’ in its ordinary sense, to express strong criticism of the kind that we also express with words like “foolish”, “stupid”, and “senseless”’ (Parfit 2011: 114).
I have discussed this and the previous point in Fink (2012: 129). Mathias Sagdal objects to me that computers, and other organisms or systems, might also be rational. If so, ‘having a mind’, ‘being alive’, etc. may not turn out to be necessary conditions for being rational. But this is not problematic for the point I wish to make here; it just changes what belongs on a list of necessary conditions for rationality.
In his early writings on rationality, John Broome ignored the question of when a particular requirement of rationality applies to particular individuals. Implicitly, the requirements of rationality applied to everything. However, Broome (2007a, p. 38) recognised this as a problem when realising that this entails that requirements can only be satisfied or violated, but not be avoided. In his Rationality Through Reasoning (2013) he changed his view so that all such requirements apply to those who are capable of rationality.
Like moral or legal blameworthiness requires moral or legal culpability, I assume that ‘rational blameworthiness’ requires rational culpability, which, I take it, consists in one’s rationality capacity. That is, you can be rationally liable for having or lacking a certain pattern of attitudes (in virtue of being subject to certain rational requirements) only if you are in possession of a rational capacity that includes some kind of a (perhaps dispositional) understanding why a certain pattern of lack of attitudes constitutes an incoherence among your attitudes.
This example is likely to be oversimplified, as an understanding of means-relations requires some conceptual grasp of truth. But, for the sake of simplicity, I will ignore this complication.
I assume a subject S is fully coherent in a certain context c if and only if S satisfies all requirements of rationality that apply to S in c.
Lewis (1973).
Cf., for example, Broome (2007c, d, 2013), Kolodny (2005, 2007), Raz (2005). This statement simplifies things, however. For example, I assume that a normally developed person is not entirely as rationality requires her to be if: (i) she believes that ought to A; (ii) fails to intend to A; and also (iii) believes that bringing about A is up to her (i.e. she believes A will not happen if it is not herself who brings about A). Broome’s (2013) provides an excellent discussion on further restrictions concerning the ‘enkratic’ requirement of rationality.
Kolodny (2007) supposes that (i) (or a cognate formulation) represents a necessary entailment.
Ralph Wedgwood makes an analogous proposals for beliefs about an option’s goodness that ensure that a choice is rational. He writes (2003: 203) that ‘… a choice is rational just in case the agent believes that the option chosen is (in the relevant way) a good thing to do. But this would not be a very plausible thing to say: if the agent’s belief that the option chosen is a good thing to do is a grossly irrational belief, then surely the choice will be equally irrational. So it would be more plausible to say this: a choice is rational just in case it is rational for the agent to believe that the option chosen is (in the relevant sense) a good thing to do’.
As I have instead argued in Sect. 5, the application of a rational requirement depends on one’s rational capacity, where capacity can be roughly understood as one’s understanding of and ability to establish coherent relations among one’s attitudes. A more precise characterisation of coherence is provided in Sect. 8.
Here I assume the correctness of the ‘principle of explosion’, as it applies in classical logic, i.e. one can validly derive any statement from a conjunction of contradicting propositions (p and not-p).
Parfit (2011: 112–113) hints at this solution by saying that ‘[w]hen our beliefs are inconsistent, some of our desires or acts may be rational relative to some of our beliefs, but irrational to others’.
Of course, this solution works only if we decompose ‘conjunctive beliefs’. Suppose you have a single belief that [p and not-p]. Then we would need to decompose this into the atomic propositions that constitute the content of your conjunctive belief and see what their normative consequences would be. That is, we would need to see what you have most reason to do if p were true, and what you would most reason to do if not-p were true.
Broome (2007d: 365) makes a similar point.
For additional criticism of Parfit’s counterfactual account of rationality, see Section 6 in Broome (2007b).
By saying that S has a non-maximal degree of coherence, I wish to express that S has degree of coherence that is not maximal.
By saying that ‘S not-Xs can explain why S is non-maximally coherent’ I simply wish to express the notion that it is possible for the fact that S not-Xs to explain why S is non-maximally coherent.
For example, being rationally required to believe that the cat is on the mat if you believe that [the cat is on the mat and it is snowing] will presuppose different rational capacities than being subject to a requirement that requires your degrees of believing that it snows, it rains, and it is sunny to add up to one if you believe that [necessarily, either it snows or it rains or it is sunny] and you believe that [it snows, it rains, and it is sunny are mutually contrary propositions].
Compare this with the non-defeasibility of a valid argument. If p validly implies q, so will any conjunction containing p.
Requiring the necessary explanation in the preferred analysis to be efficient brings another advantage. It disallows an implausible aggregation principle for rational requirements. By ‘aggregation’ I mean this: if, at w, rationality requires of S that S As and, at w, rationality requires of S that S Bs, then, at w, rationality requires of S that [S As and S Bs].
Here is why this principle is implausible. Suppose there is a situation in which: (i) rationality requires you to believe A, (ii) rationality requires you to believe not-A, yet also (iii) rationality requires you to not believe [A and not-A]. Considering ‘the preface paradox’ (Makinson 1965) such a situation seems quite possible. Now assume that (iv) rationality requires you to not believe [A and not-A] only if rationality requires you not [to believe A and to believe not-A]. That is to say, rationality requires you to not believe a contradiction only if rationality also requires you not to have contradictory beliefs. From (iii) and (iv) we can derive that (v) rationality requires you not [to believe A and to believe not-A], which in turn implies (iv), that it is not the case that rationality requires you [to believe A and to believe not-A]. Consequently, we cannot aggregate (i) and (ii) to ‘Rationality requires you [to believe A and to believe not-A]’.
For an excellent discussion of attitudinal consistency and unity, and their relation to rational requirements, see Reisner (2013).
Reisner (2013) discusses this point in detail.
See Velleman (1992).
The use of ‘logically excluded’ is pragmatically motivated here. It ensures that the threshold for objective incoherence is as high as possible.
In other words: you believe that p will not be the case unless you bring about that p.
Cf. Broome (2005: 1–2, n. 5).
I say ‘normally’ because it could be the case that it is logically excluded that: (i) your intention that that p; (ii) your belief that [p only if q]; and (iii) your belief that [q only if you intend that q] meet their success conditions jointly. This is precisely so if, necessarily, p implies not-q.
There is an important restriction one needs to apply to this type of incoherence when it comes to second-order attitudes. Suppose you believe that you know that p. Suppose further that one of the success conditions of this belief is truth. Consequently, as long as it is not the case that you know p, it is logically excluded that your belief that you know p will meet all its success conditions. But, arguably, you are not necesserily incoherent if you believe that you know something without knowing it—e.g. suppose you believe you have excellent evidence that p, though this is not the case. This shows that this view needs to be restricted when considering the incoherence of second-order attitudes.
On the distinction between ‘objective’ and ‘subjective’ rationality, see Kolodny (2005).
For a detailed account as to how mental dispositions can be sensitive to a particular property of a mental event, see Wedgwood (2006).
By ‘makes’ I do not mean that S’s not-Xing ‘causes’ but ‘constitutes the fact’ that S is objectively incoherent.
Cf. Section 5 of Broome (2007b).
References
Blumberg, A. E. (1976). Logic: A first course. New York: Alfred E. Knopf.
Bratman, M. (2009). Intention, belief, practical, theoretical. In S. Robertson (Ed.), Spheres of reason (pp. 29–62). Oxford: Oxford University Press.
Broome, J. (2005). Have we reasons to do as rationality requires? A comment on Raz. Journal of Ethics and Social Philosophy, 1, 1–8.
Broome, J. (2007a). Requirements. In T. Rønnow-Rasmussen, B. Petersson, J. Josefsson & D. Egonsson (Eds.), Homage à Wlodek: Philosophical papers dedicated to Wlodek Rabinowicz (pp. 1–41). www.fil.lu.se/hommageawlode.
Broome, J. (2007b). Does rationality consist in responding correctly to reasons? Journal of Moral Philosophy, 4, 349–374.
Broome, J. (2007c). Is rationality normative. Disputatio, 11, 153–171.
Broome, J. (2007d). Wide or narrow scope? Mind, 116, 359–370.
Broome, J. (2013). Rationality through reasoning. Chichestery, UK: Wiley Blackwell.
Cherniak, C. (1981). Minimal rationality. Mind, 90, 161–183.
Cherniak, C. (1986). Minimal rationality. Cambridge, MA: MIT Press.
Davidson, D. (2001). Essays on actions and events (2nd ed.). Oxford: Oxford University Press.
Engel, P. (2004). Truth and the aim of belief. In D. Gillies (Ed.), Laws and models in science (pp. 79–99). London: King’s College Publications.
Evnine, S. (2001). The universality of logic: On the connection between rationality and logical ability. Mind, 110, 336–367.
Fink, J. (1976). Is the enkratic requirement of rationality normative? (unpublished manuscript).
Fink, J. (2010). Assymetry, scope, and rational consistency. Croatian Journal of Philosophy, 29, 109–130.
Fink, J. (2012). The function of normative process requirements. Dialectica, 66, 115–136.
Gauthier, D. (1986). Morals by agreement. Oxford: Oxford University Press.
Hintikka, J., & Bachman, J. (1991). What if …? Toward excellence in reasoning. Mountain View: Mayfield.
Kant, I. (1788/1997). Critique of practical reason (M. Gregor, Ed., M. Gregor, Trans.). Cambridge: Cambridge University Press.
Kolodny, N. (2005). Why be rational? Mind, 114, 509–563.
Kolodny, N. (2007). State or process requirements? Mind, 116, 371–385.
Korsgaard, C. (1985). Kant’s formula of universal law. Pacific Philosophical Quarterly, 66, 24–47.
Lewis, D. (1973). Counterfactuals. Oxford: Blackwell.
Makinson, D. C. (1965). The paradox of the preface. Analysis, 25, 205–207.
Parfit, D. (2001). Rationality and reasons. In D. Egonnson, J. Josefsson, B. Petersson, & T. Rønnow-Rasmussen (Eds.), Exploring practical philosophy: From action to values (pp. 17–39). Aldershot: Ashgate.
Parfit, D. (2011). On what matters. Oxford: Oxford University Press.
Price, A. W. (2008). Contextuality in practical reason. Oxford: Oxford University Press.
Raz, J. (2005). The myth of instrumental reason. Journal of Ethics and Social Philosophy, 1, 1–28.
Reisner, A. (2009). Unifying the requirements of rationality. Philosophical Explorations, 12, 243–260.
Reisner, A. (2013). Is the enkratic principle a requirement of rationality? Organon F, 20, 437–463.
Scanlon, T. M. (2007). Structural irrationality. In G. Brennan, R. Goodin, & M. Smith (Eds.), Common minds (pp. 84–103). Oxford: Oxford University Press.
Schroeder, M. (2004). The scope of instrumental reason. Philosophical Perspectives, 18, 337–364.
Smith, M. (1994). The moral problem. Oxford: Basil Blackwell.
Velleman, J. D. (1992). The guise of the good. Noûs, 26, 3–26.
Velleman, J. D. (2000). The possibility of practical reason. Oxford: Oxford University Press.
Wedgwood, R. (2003). Choosing rationally and choosing correctly. In S. Stroud & C. Tappolet (Eds.), Weakness of will and practical irrationality (pp. 201–229). Oxford: Oxford University Press.
Wedgwood, R. (2004). The aim of belief. Philosophical Perspectives, 16, 267–297.
Wedgwood, R. (2006). The normative force of reasoning. Noûs, 40, 660–686.
Acknowledgments
I am greatly indebted to John Broome, Krister Bykvist, Alexandra Couto, Olav Gjelsvik, Micha Gläser, Herlinde Pauer-Studer, Christian Piller, Andrew Reisner, Mathias Sagdahl, Anne Schwenkenbecher, Martin Vacek, Ralph Wedgwood, numerous anonymous reviewers, and audiences at Graz and Vienna for very detailed and helpful comments on earlier drafts of this paper. During the paper’s long gestation, I received financial support from the Austrian Academy of Sciences, the Research Council of Norway, and the European Research Council (Advanced Grant ‘Distortions of Normativity’), and the National Scholarship Programme of the Slovak Republic. I thank those institutions for their very generous support.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Fink, J. A Constitutive Account of ‘Rationality Requires’. Erkenn 79, 909–941 (2014). https://doi.org/10.1007/s10670-013-9591-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10670-013-9591-8