Samuel Freeman's Review of Parfit

In the latest issue of The New York Review of Books there is a review of On What Matters by Samuel Freeman. Since I have previously covered reviews of the work by Simon Blackburn and Philip Kitcher it seems only right to provide a general summary and response also to this one by Freeman.


Freeman's review opens with comments aimed at the general reader concerning the philosophical attempt to explain the purposes of common sense moral rules and to resolve moral dilemmas. Freeman refers to Kant almost immediately afterwards mentioning both the unconditional character of moral demands on Kant's conception and the Formula of Humanity. The argument of the first part of the Groundwork is implicitly referred to when Freeman mentions the view that the categorical imperative "justifies" our common sense duties to one another and also provides a "more fine-grained method of reasoning" concerning what we ought to do. Following Parfit's own startling conjunction of names in the "Preface" to On What Matters Freeman moves from Kant to Sidgwick and the latter's argument in The Methods of Ethics that it is the principle of utility that truly "justifies" common sense morality. The move from Kant to Sidgwick also allows Freeman, however, to refer to the Rawlsian challenge to the dominance that the utilitarian tradition had, until recently, in the area of social theory.


The introduction of the account of utilitarianism into Freeman's narrative also allows him to mention that the method of ethical appraisal promoted by this school is really a specific form of the general notion of consequentialism and that many consequentialists today have abandoned a simple commitment to hedonism and do not necessarily view utility as the best way of assessing consequences. After mentioning the shift within moral theory from utilitarianism to consequentialism Freeman fills out his picture of the general state of contemporary moral reflection by mentioning Thomas Scanlon's notion of "contractualism", a much more recent theory than Kantian and consequentialist ones. Scanlon is presented by Freeman as someone who modified the Rawslian view so that it was applied not to the theory of justice but instead to personal duties or duties towards others. Not only is this the basis of contractualism according to Freeman but this view is related by him to the Kantian one in terms of recognising the equal status of persons as integral to moral reasoning and, in making this emphasis so significant, as ensuring that these accounts stand together against consequentialism.


As Freeman mentions in a footnote, the three theories of Kantianism, consequentialism and contractualism are the centre of Parfit's book and are related together by him as the key traditions that have to be brought together, a point that ensures that virtue theory is simply left without consideration by Parfit. Parfit intends overall to articulate what he terms a "Triple Theory" that will combine optimific considerations with Kantian and contractualist ones thus producing what Parfit calls a "Kantian rule consequentialism". This argument for a "Triple Theory" of normative ethics is combined by Parfit with a general account of moral reasons that is aimed at showing their objectivity with the arguments in favour of this claim being the centre of the first and the sixth parts of the book. The latter claim is specifically presented by Parfit in response to subjectivist views of reasons and in riposte to a claimed nihilism with regard to values said to characterise much contemporary theorising about morals.


Parfit's attack on subjectivism is rightly presented by Freeman as mainly consisting in responses to the "Humean" theory of reasons that understands reasons primarily through the prism of desires. In response to such a claim Parfit insists on an account of objective reasons based on a view of intentional objectivity but the nature of the specific theory Parfit here elaborates and the problems with it are not considered by Freeman who simply presents it as laudable that Parfit attacks relativistic subjectivism without discussing whether such an attack is, as stated, either too blunt or unconvincing. This is a peculiar weakness in Freeman's review and is likely connected to a dialectical strategy of refuting the "Triple Theory" but upholding the significance of Parfit's book in terms of the defence of "objectivity" in morals. One of the reasons why Freeman may have taken this tack is due to the prevalence of forms of the "Humean" theory in contemporary economics and the need to challenge such a model as applied there.


Freeman looks at Parfit's treatment of the Formula of Humanity and chides Parfit for considering it in separation from Kant's general moral theory. However there are further significant problems both in terms of how Freeman views Parfit's discussion of the Formula of Humanity and with how Freeman understands the place of the Formula of Humanity in Kant's own theory. With regard to Parfit's treatment Freeman stresses the way that the discussion of the "mere means" requirement appears to be favoured by Parfit above other considerations though this point is here not well made given that Parfit considers another point to flow from the Formula of Humanity than this. Parfit also stresses a notion of "rational consent" that he articulates as the basis of the first part of the Formula and, whilst Freeman may not find this a persuasive reading of the Formula it is remiss of him to simply present Parfit as viewing the Formula only through the "mere means" requirement. It is true that Parfit does not view the Formula in terms of a "respect requirement" but Parfit does provide arguments for not viewing the "respect requirement" as really providing further normative guidance and this requires to be discussed and answered.


Freeman indicates some awareness of the "respect requirement" but moves rather quickly from referring to it to an account of some themes from Kant's philosophy of right including the "innate right to freedom" and the notion of independence from being constrained by another's choice. Since these themes belong not in Kant's general moral theory but to the account of right it is not at all clear why they are referred to by Freeman and it is not remiss of Parfit to have failed to engage with them when formulating his notion of a Kantian consequentialism. It is correct to argue that the Formula of Humanity appears to include a constraint on action in terms of describing something one should not act against but Freeman fails to bring this point out and generic reference to the notion of "individual rights" is insufficient to make clear the specific problem with Parfit's view of the Formula of Humanity.


Parfit is, in any case, as Freeman recognises, more concerned with Kant's discussion of universal law than with the Formula of Humanity. The discussion of universal law is interpreted by Parfit in accord with a principle taken from Thomas Scanlon so that it becomes understood as prescribing the requirement that everyone ought to follow the principles whose universal acceptance everyone could rationally will. This notion of a "Kantian Contractualism" is one that requires viewing the notion of universal law in terms that are not derived from Kant himself but Freeman emphasises more the application of the principles that follow for Parfit. Freeman terms Parfit's application "peculiar" and part of what is thus peculiar is the way the combination of Kant with Scanlon is meant to underpin rule consequentialism. Parfit requires us to see the reference to what makes things go best as grounded on some kind of formal rule process built into an understanding of rational willing. This has the result, as Freeman puts it, that "morality is then a kind of efficiency in promoting universal good". 


Once we have a concentration of this optimific sort ethics clearly becomes axiological and this is part of what Freeman rightly regards as controversial in Parfit's view. Kant and Scanlon are generally taken to have provided moral views that are not consequentialist (although there are other varieties of "Kantian consequentialism" such as the one articulated by David Cummiskey). In making the move of bringing Kant and Scanlon into alliance with consequentialism Parfit follows the method inaugurated by Sidgwick who made a similar move the centre of The Methods of Ethics. Freeman suggests that this move on Parfit's part will have particular effect for Kantians since he claims that: "Parfit's consequentialist interpretation of the categorical imperative will stimulate philosophers for years to come".


Despite making the suggestion that Parfit's reading of Kant is one that is likely to be philosophically significant Freeman rejects quite central aspects of Parfit's methodology. One of the elements of Parfit's work that provoked Simon Blackburn's ire was the insistent use of Trolley problems in the work and Freeman joins with Blackburn in finding recourse to them problematic. One of the reasons Freeman gives for rejecting the use of Trolley problems is that such problems can be varied with resultant differences of appraisal. Freeman also cites Susan Wolf's reply printed in the second volume of On What Matters in which she claims that there is "no single principle" underlying our moral intuitions in Trolley type cases. In citing Wolf (and also Allen Wood) in opposition to the use of these cases, however, Freeman seems simply content to state their objections to Parfit's procedure without considering the extensive replies Parfit includes to interlocutors, something which certainly seems rather odd. Freeman also seems to rest content with claiming that if Trolley problems really are of little use in moral philosophy that this in itself undermines Parfit's 'Triple Theory' thereby implying that there are no substantive arguments given by Parfit with regard to the formulation of the 'Triple Theory' that are separable from the use of Trolley Problems and this contention, as put, seems a rather strong reading of the recourse to Trolley Problems.


Freeman's overall evaluation is strongly negative since he argues that Parfit does not address the questions as to why we should view morality axiologically or provide us with a view of what the ultimate good is that we are to optimise. However, the first point is stronger than Freeman himself ultimately puts it since he recognises a strain in Parfit's work that derives from Sidgwick and is based on the notion of a general impersonality. It may be, as Freeman says, a kind of "refined philosophical sensibility" that prefers this concentration to one on personal affairs of the sort provided by Scanlon but it is the basis of a ground for axiological conceptions of morality. In so being it is, as Freeman rightly stresses, somewhat out of key with Kantian and Scanlonian views of morality and this cuts against the sense that the basis of the three views Parfit treats can easily be reconciled. However in considering why Parfit takes it that there is an ultimate ground of unity between the views Freeman retreats again to the assertion that Parfit's argument substantively depends on an inductive generalisation from the consideration of Trolley problems and this argument is one I think over-states the importance of the Trolley problems Parfit considers.  


It is a separate problem and a better point to argue that Parfit is never specific about the general good that is to be optimised though it seems to me that part of the basis of this general good is to be found in the view of objective reasons that Parfit defends and which Freeman fails to challenge. Freeman concludes his review with a statement to the effect that Sidgwick's syncretic project has been brought to a point by Parfit that ensures it has a greater range than even Sidgwick would have thought possible. Whilst this might be true the grounds of the difficulty with the 'Triple Theory' cannot simply reside within normative ethics alone but must also be grounded on the account of the "objectivity" of ethics that Parfit articulates, an account that Parfit himself presents as importantly in tension with the kind of Rawlsian view that Freeman himself defends and to which therefore Freeman should have taken time to respond in his review.

Ethics and Teleology

In the previous two postings I have looked at the notions of deontology and the alleged priority of the "right" over the "good" and whilst the former notion has proved trickier than the latter there would appear to have been a certain message emerging from consideration of these notions with regard to the nature of Kant's ethics. However, when turning to the views alleged to be held by those who think of ethics in a teleological way, we will discover some areas of substantial confusion.


Samuel Freeman, as noted in the earlier posting on deontology, views deontological theories as involving pluralism with regard to value whilst, by contrast, he presents teleological views of ethics as having a monistic conception of value. This position of Freeman's is, however, not sustained when we look back in the history of ethics. One of the first works to present a broad characterisation of teleological conceptions of ethics and, in the process, to distinguish such a view from the deontological conception, is by C. D. Broad. Broad's classic text Five Types of Ethical Theory agrees with contemporary usage in that it opens by treating teleological conceptions of ethics as primarily concerned with consequences as certain types of consequences are viewed as intrinsically good or bad. However, whilst this equation of teleological conceptions of ethics with consequentialism might seem to confirm certain contemporary views of teleological ethics there are further complications in Broad's account which converge rather less with contemporary usage. So Broad is, for example, unconvinced that there are any ethical theories are either purely deontological or purely teleological and he presents these conceptions as instead ideal limits of theory rather than characterising any ethical view as such. Secondly, and an important corrective to the claim of Freeman, Broad is quite clear that there can be pluralist forms of teleological conceptions of ethics in which several distinguishable views of the good can be held in tandem. So there is, for Broad, nothing intrinsic to teleological views that requires them to be classed as monistic.


William Frankena, as mentioned in the earlier posting on deontology, presents teleological theories of ethics in a different way to Broad. On Frankena's view, the distinctive point about teleological theories is that they assess the moral qualities of actions and persons in relation to certain non-moral values which are held to be important. So, for Frankena, what is important is the claim concerning value that is held by teleologists though, like Freeman, he does tend to view teleological theories as monistic in nature. 


If we move from these general conceptions of teleological accounts of ethics to attempts to view Kant's ethics in terms that are either resistant to deontology or embracing of teleology we will note that the problems already inherent in the accounts of teleological ethics considered will multiply. The two most prominent dissenters from the view that Kant should be considered to embrace a kind of (rule)-deontology are Barbara Herman and Paul Guyer. The reasons why Herman and Guyer resist the deontological characterisation are different but they lead both to views that require a revision of the simple conception that Kant asserts a priority for the right over the good. Barbara Herman focuses on the opening section of Kant's Groundwork where the "good will" is taken to have absolute value and she uses this to resist attributing a deontological conception of ethics to Kant where she means by "deontology" a thesis concerning value (a claimed independence of moral considerations from value). Herman claims that Kant has a view of the good "both as the formal final end and as the ultimate internal condition of rational agency". 


The problem with deontological conceptions of ethics, according to Herman, is that they do not provide us with an understanding of how it is that moral rules have any kind of claim upon us. To account for this is to give what she calls "a grounding conception of value" and we need this in order to make intelligible moral requirements. So the key for Herman is to provide an account of Kant's ethics that enables us to see it as giving us a motivational view of practical reason that provides a conception of principles that allows for a notion of value. The interesting point about these claims is that they aim less at providing a strictly teleological conception of ethics than one that does not rest so evidently as the previous posting might have led one to think on an assumed priority of the right over the good.


Paul Guyer brings together teleology with a view about the relation between the right and the good as he views teleological ethics as consisting in an assertion of an antecedent good prior to principles of right being established but whilst the notion of the good tends to be assimilated with some notion of consequences it is not so understood by Guyer for whom the intrinsic value of freedom could be grasped as such a "good". Since, though, somewhat unsurprisingly, Guyer views Kantian freedom as autonomy (or freedom governed by law) it would follow that such freedom incorporates a sense of duty, though, on his view, it also includes the value that is so important to Herman. On these grounds, however, Guyer asserts a surprising preference for the argument of the Groundwork over that of the Critique of Practical Reason and dismisses Kant's claim for the "fact of reason" in the latter as indicative of a view that abandons finding reasons for why we would adopt the moral law (hence effectively finding in the Second Critique the view that Herman rejects). Like Herman, Guyer views the good will as something that is antecedently valuable and indicative on Kant's part of an understanding of a substantive conception of the good with the second part of the Groundwork amplifying this into the notion of rational being as an end-in-itself.


Guyer's view hence includes elements not present in Herman's and allows for at least one point not congruent with Herman's. This is that Guyer allows that the intrinsic value of freedom is not itself demonstrable, something that seems to ensure that it does not meet Herman's demand for a rational construction of Kantian ethics.


What we have noted in this posting is that the view of what teleological views of ethics consists in is less settled than many think.  Further, at least some prominent contemporary Kantians are less than happy with the effect of characterising Kant's ethics in a way that does not make room for considerations that have often been thought to belong only with teleological views. In the next posting we will consider consequentialism.

The Priority of the Right Over the Good

Following the discussion in the last posting of the sense(s) that can be given to the characterisation of a view as deontological I will, in this posting, look at the ways the assertion of the priority of the right over the good has been understood. The first thing that is obvious about the assertion of the priority of the right over the good is that it asserts a lexical ordering. According to Samuel Freeman, the point of asserting the priority of the right over the good is to make a claim concerning "the desires and interests moral agents can take into account".


On this conception of the claim for the priority of the right it is a principle that limits what we should include in our practical reasoning. This point is also made explicit by John Rawls in A Theory of Justice which states that the principles of right "put limits on which satisfactions have value; they impose restrictions on what are reasonable conceptions of one's good" (TJ 27 rev/31 original). So to assert the priority of the right is to have some standards by which we can determine what things would be "good" for us though here the point seems to be a more limited one than Freeman has suggested as in this citation it is clear that the limit is placed on satisfactions. However, although the formulation from Rawls is in this respect more limited than that from Freeman, it is, in another sense, more inclusive. After all, on the Rawls conception given, the assertion of the priority of the right does involve a thesis concerning value (whilst, on at least some construals, deontology, by contrast, seems not to be concerned with value).


Similarly, in Political Liberalism, Rawls suggests that the priority of the right "characterizes the structure and content" of his view such that it indicates what can count as a good reason. If it so determines what counts as good reasons then in a sense the assertion of the priority of the right indicates what kind of place things that might be deemed to be "good" can have in the view in question. The connection of this claim to Kant involves an appeal to the view Kant presents in the Critique of Practical Reason as a "paradox" of method where he claims that the concepts of good and evil "must be defined after and by means of the law" (Ak. 5: 63). This is more fully set out when Kant effectively articulates something like a "Copernican turn" in ethics by discerning the error of philosophers when previously discussing the supreme principle of morals:


"they sought an object of the will in order to make it into the material and the foundation of a law (which would then not be the directly determining ground of the will, but would be the determining ground of the will indirectly, only by means of that object referred to the feeling of pleasure or displeasure); instead, they should have first looked for a law that a priori and directly determined the will, and only then determined the object conformable to it." (Ak. 5: 64)


This second passage makes clearer the type of change involved. Kant is indicating that we should not begin by finding something we take to have "value" and then base our notion of moral law upon it but rather articulate the moral law that is a priori and then comprehend what can be said to have value in terms of what fits with it. If this is the kind of consideration that is at work in the notion of the assertion of the priority of the right over the good then it will be necessary next to turning to what it is that is involved in viewing an ethical position as teleological.

Deontology

It's a commonplace in discussions of Kant's moral philosophy to characterise it as "deontological" where the term "deontological" is meant to indicate that Kant places a priority of the right over the good. However, I have long been troubled by the assumption that in terming Kant's moral philosophy "deontological" that there is a basic and simple rationale for describing him as not allowing teleological considerations into his ethics when it has seemed to me that there are a lot of pretty good reasons for seeing his ethics as being teleological. In fact the suggestion to the effect that Kant's ethics are teleological was a prime motive when I wrote my book Kant's Practical Philosophy: From Critique to Doctrine. However, when this work was reviewed in Mind by Oliver Sensen the objection to this argument followed the lines I have suggested, namely, it pitted against the book's line concerning teleology the view that Kant asserts the priority of the right over the good and is hence committed to a form of deontological ethics. Sensen's review in Mind was very fair (unlike some that appear there) and took a considerable amount of time to make a lot of my argument explicit so I am far from complaining about it. But in making this stock response to the work he reiterated a distinction which I had been trying to unsettle. It has recently struck me that, in responding to the view that Sensen was orthodoxly presenting, it is necessary to attempt a very nuanced set of distinctions and I'm going to use a few postings to attempt to draw them out. After doing so, I will then return to the question of the view that I was attributing to Kant in my book when I presented an argument for understanding his ethics as teleological.


So: I think the following distinctions need to be made. First of all, the difference between "deontology" and the priority of the right over the good. Secondly, the difference between a teleological view of ethics and a consequentialist one. Thirdly, the prospects for constructing a picture of Kant's ethics that views it as teleological whilst also allowing for the point that Kant does articulate a reason for viewing the right as prior to the good. In order to arrive at the picture suggested here though it will first be necessary, for these three points to be convincingly made, to articulate what, in any event, the real meaning is of such expressions as "deontology", "teleology", "the priority of the right over the good" and "consequentialism", each of which are more challenging than is often thought.


In this post I will begin with a focus on "deontology" looking at what is specific to it. The problem with doing with this is that those who specifically try to focus on it in order to discuss the nature of it, by contradistinction to any other view, hold very conflicting and contrary views as to what it consists in. So, Barbara Herman, for example, in a piece articulating a reason why Kantians should bid farewell to deontology, views it as primarily a thesis about value where deontology is understood to be asserting that, in some sense, ethics is "independent" of value. By contrast, Samuel Freeman articulates what is distinctive to deontology by asserting that on such a view there is not a single rational good or, put positively, that there is a plurality of rational goods. So, on one of these views deontology is a thesis concerning value, whilst on the other, it is a pluralist theory of what is good.


These quite different views of the nature of deontology indicate a reason to look more carefully at how it has been characterised in previous works, works that may well be at the back of contemporary definitions but whose implications may have been forgotten or only remotely remembered, hence indicating a basis for the divergent conceptions of deontology. So, if we look back at William Frankena's 1963 volume Ethics we will find ways in which the divergent characterisations of Herman and Freeman can be brought together. In relation to Herman's conception, to begin with, we should note that Frankena describes the deontological view as concerned with "certain features of the act itself other than the value it brings into existence" and in making this point Frankena suggests that taking account of the "value" in moral considerations is to bring in reference to something that is, in some way, distinct from the moral nature of the act, something, that is, that has non-moral force. Since the deontologist is committed to understanding acts in this way it would appear that there is an independence from "value" thus understood in their considerations.


However, Frankena makes a further set of claims concerning deontology that appear to make more sense of the view of it proclaimed by Freeman. So, on further examination of deontology, Frankena claims that it can be distinguished into two separate forms, one that is termed act-deontology and another that he labels rule-deontology. Starting with act-deontology we get quickly moved towards a view of ethics very similar to situation ethics though this is slightly complicated by Frankena's suggestions of a distinction between moderate and extreme views of act-deontology. On the moderate view of act-deontology, Frankena claims, general rules can be built up on the basis of particular cases but a general rule cannot ever supersede a particular judgment concerning what should be done. By contrast, extreme act-deontologists assert "we can and must see in or somehow decide separately in each particular situation what is the right or obligatory thing to do" in a way that is independent of general rules which gives us a form of particularism that effectively converges with situation ethics. Either form of act-deontology clearly amounts to a suspicion of general principles with attention focusing on specifics of situations and does imply, as Freeman suggested, a form of pluralism.


Frankena subsequently distinguishes both forms of act-deontology from the view that he calls rule-deontology. The latter does embrace some sense of rules where these rules establish the nature of what is right or wrong and do so independently of considerations of production of the type of non-moral "value" that Frankena had earlier mentioned. So it would appear that there are grounds for accepting that both Herman and Freeman have captured part of the point of a deontological view though both are also partially misleading. In the next posting I'll set out how, by contrast, the question of the claim of the priority of the right over the good should be understood.

Publicity and Permissive Laws

At the end of yesterday's posting I suggested that there might be some kind of argument at the close of Kant's treatment of the second definitive article of Perpetual Peace that was worth consideration and that would, if so considered, perhaps alter the view I have previously expressed concerning the division of the preliminary articles in Perpetual Peace. Now, without perhaps entirely delivering on this promise, I want to give an account of publicity that might lie at the back of the idea that there are permissive laws to do certain things that a full condition of right would not authorize.

In the course of one of the many splendid chapters of his book Justice and the Social Contract Samuel Freeman makes the following remark: "Publicity assumes that principles of justice will provide a shared public basis for political discussion, criticism, and justification. Principles that cannot gain knowing, widespread support from reasonable and rational members of a well-ordered society are to be passed over in favour of those that do." (91) I am now going to re-interpret this remark in relation to Kant's admission at the close of his discussion of the second definitive article that since nations reject the hypothesis of the world republic (despite its correctness as a thesis) then they should adopt instead the surrogate of the pacific league.

In making this claim Kant is, I want to suggest, formulating a kind of permissive law. When he formulated the notion of the permissive law earlier in Perpetual Peace he indicated that such laws permitted to continue in the present something that would be outlawed in the future. If we take the future here to refer to the condition under which right was fully and completely realized then the continuance of such separate states would not then be permitted but, under present conditions, it is allowed to remain. The ground for its being allowed to remain is only that the separate nations will not allow the abolition of themselves. Since they will not the surrogate is adopted instead.

Now if we re-formulate Freeman's account of public reason we can understand it to be saying that if a principle cannot reach agreement between all parties to discussion in a public forum then a different principle needs to be adopted instead with the intent of serving the same end as the first (rejected) principle. In this way the public opinion of those in the discussion has been respected whilst right has not been neglected. This is, I think, the consideration by which Kant proposes the pacific league. What brings this closer to Freeman's account than would otherwise be the case is that the pacific league is formed by republics uniting together. Since republics are moral persons (unlike despotisms) then their public agreement is akin to that between individual members of a state of nature.

Reverting to the division of the preliminary articles: the argument considered in the posting of two days ago assessed it in relation to principles of publicity. The point here was that principles 2-4 are all ones that can be assessed publicly and it is the fact that they can that enables an understanding of what is at issue in delaying their implementation since measures that lead towards full recognition of them can be commonly understood. By contrast, principles 1 and 6 are ones that cannot be assessed by public methods which is why they have to immediately cease. This argument works also by considering public opinion as a rightful force operative within the moral personalities of the international set of states. However, there remains a problem with understanding the fifth article (concerning non-intervention) as one that requires immediate implementation. It is unlike 1 and 6 in being available to public measurement. It is also not congruent with the first definitive article since one of the grounds for this article was that republics are more inclined to peace than despotisms. Since this is so, the existence of despotic constitutions is itself not merely something not in accord with right but also something that tends to constantly produce conditions of war. Given these points there are reasons for treating despotisms as little better than states torn by civil war and with regard to the latter there is permission to intervene. So shouldn't there also be a ground to intervene to overthrow despotic regimes? I will return to the topic of intervention in future postings.