“Obligation”, “Ought” and “Can” “Plikt”, ”borde” och “kan”

(1)

“Obligation”, “Ought” and “Can”

“Plikt”, ”borde” och “kan”

— Jakob Sjölander

Department of historical, philosophical and religious studies Magister Thesis, spring 2020

Version: 1.4

Magister och magisteruppsats i filosofi Supervisor: N/A

Examiner: Sofia Jeppsson

(2)

I fully subscribe to the judgment of those writers who maintain that of all the differences between man and the lower animals, the moral sense or conscience is by far

the most important. This sense, as Macintosh remarks, “has a rightful supremacy over every other principle of human action”; it is summed up in that short but imperious

word ought, so full of high significance.

Charles Darwin, The Descent of Man

(3)

Abstract

This paper criticises the famous “”ought” implies “can”” on the grounds that its main claim - that there can be no unfulfillable obligations - is false. The first part of the paper investigates the use, history and previous literature on the topic, as well as the proper understanding of the principle. The second part presents the main argument, directed at the interpretation of “”ought” implies “can”” as a conceptual truth. It is argued that it must be possible to split the meaning of the term “ought” into component parts, parts that must each be quite capable of functioning on their own. The result is that we cannot conclude that one part of the term “ought” (such as whatever is its moral content) must disappear just because another part of it (such as “can”) does so. Finally, in the third part, we tackle some common arguments and intuitions in favour of the principle, and provides a few concluding words and a summary.

(4)

Abstract

Den här uppsatsen kritiserar den berömda “”bör” implicerar “kan””-principen, (här tolkad som “konceptuell implikation), på så vis att principens viktigaste punkt är falsk, nämligen idén att det inte kan finnas några ouppnåbara plikter. Uppsatsens första del undersöker användningen, historien och litteraturen bakom ämnet, samt formuleringen av principen. Den andra delen presenterar huvudargumentet, riktat mot ””bör”

implicerar ”kan”” som en konceptuell sanning. Argumentet fungerar på så vis att det måste vara möjligt att dela upp meningen i termen ”borde/bör” i komponenter, komponenter som måste vara kapabla att fungera även på egen hand. Resultatet blir att vi inte kan dra slutsatsen att en del av termen måste försvinna (såsom dess moraliska innehåll) bara för att en annan del av termen (såsom ”kan”) gör det. Slutligen, i den tredje delen, så behandlar vi några vanliga argument och intuition som förs fram till principens försvar, och summerar texten.

(5)

1.0 Introduction

Does "ought" imply "can"? The idea that "ought" does imply "can" is known as the

""ought"-implies-"can"”-principle, or OIC for short. The principle, traditionally attributed to Immanuel Kant, has a long and distinguished history. OIC is important to many parts of ethics, their applications and the challenges they face. Key examples include the possibility or impossibility of moral dilemmas, deontic logic, blameworthiness, supposed limits on what morality can demand of us¹, free will, responsibility, and moral motivation et cetera².

There does, however, remain a serious question concerning the principle, namely: is it actually true? Does "ought" really imply "can"? Can we always do what we ought?

Ought we not sometimes do things we cannot do? Though often just taken for granted, the OIC-principle has been challenged by such philosophers as Sinnott-Armstrong³, Robert Stern⁴, and Alex King⁵. This paper will join the ranks of these sceptics.

Written with the purpose of investigating the OIC-principle, this paper will eventually argue that while "ought" may or may not imply "can", it cannot do so in a way that makes it interesting. That is, in the way that allows us to draw certain ethical conclusions. The most common understanding of the “”ought” implies “can””-principle interprets it as a

“conceptual implication”, or that the word “ought” would necessitate a “can” just as the word “bachelor” necessitates “unmarried”. The argument (see sections 2.0 – 2.6) against this view works by showing that, for this conceptual implication interpretation of OIC to be true, it must be possible to subdivide the term “ought” in such a way that it can be separated into a moral part (“X”) and a “can”-part. But when such a division is made, it becomes clear that those two are separate concepts capable of standing on their own. The unification of “X” and “can” in the word “ought” is an arbitrary one, and it gives us no reason to believe that the two component parts cannot exist without the other. This defeats the purpose of the OIC-principle, which is to strictly link some moral term to

“can”. Thus, the principle (or at least the “conceptual implication”-version) should be rejected.

An important point to make is, however, that this is not quite the same as arguing that the OIC principle is false. There are many irrelevant, unhelpful and trivial things that are still true. The OIC principle, we will argue, is not so much untrue as impotent. Though it

(7)

is not something we will discuss in this paper, it is doubtful whether a principle can be said to be false. Either way, we will show that the OIC-principle cannot do what we want and need it to do in order for it to be interesting. An important distinction here is between the “”ought” deactivating” and the “”Ought” dispelling” version of the principle, discussed in section 1.2.

With this introduction we have already got stated on Part 1, which consists of four sections. The next section, 1.1, will provide background and an introduction to the principle. Section 1.2 will elaborate on the discussion of the previous paragraph, and investigate what we want from the OIC-principle, and the conditions under which it should be rejected. Section 1.3 will begin the main work of understanding OIC and its constituent terms, the meaning of "ought", "implies" and "can". Most importantly, we will here also present the most important version of OIC that we will be dealing with, Alex King’s “Standard OIC”.

Part two is where the main argument of this paper is presented. It consists of 7 sections, 2.0 – 2.6, with each section dedicated to a step in the argument. It is argued that if the OIC-principle is to work, the word “ought” will have to be understood as a combination of a moral term (probably but not necessarily “pro tanto obligation”) and the term “can”. It is shown that since both parts of the term can function on their own, it makes little sense to assume the kind of necessary connection and co-dependence that the OIC-principle does. With that, the OIC-principle cannot fulfil the desiderata previously detailed in section 1.2, namely that the absence of a “can” dispels all moral obligations.

In Part three we start rounding off. It is the shortest of the three parts, and contains two sections apart from its introduction. Section 3.1 attempts to deal with the OIC- principle’s main support, intuition. We investigate why the principle is accepted, and a few common arguments in its support are dealt with. It is suggested that an excessive respect for terminology has mislead philosophers into accepting a kind of castle in the air.

1.1 “Ought” implies “can” - origin, history and basic structure

We are here interested in modern, or maybe timeless, versions of the OIC-principle rather than in any specific historical or once-fashionable version of it. Nevertheless, a

(8)

brief overview of the concept's history and significance is likely to be useful.

Peter B. M. Vranas writes in his introduction to the field:

“The "Ought implies Can" (OIC) thesis establishes a link between obligations and abilities. It is associated with Kant, but the Kantian attribution is debated. Its main interpretation goes along the latin motto "ad impossibilia nemo tenetur" [Nobody is

held to the impossible] and allows you to discharge an obligation when you lack the possibility to do what is commanded. Others interpret it as saying that, given the fact that you have the relevant "Can", there is no way not to do what you Ought to. OIC has

been used as an (unquestionable) principle in many fields. from moral and legal philosophy (moral dilemmas, alternate possibilities) to deontic logic and epistemology (doxastic voluntarism). Since the 1960's it has undergone multiple criticisms from many

areas.”⁶

As Vranas points out, though typically attributed to Immanuel Kant, the sentiment on which the OIC-principle is based can be traced far back in time. Pelagius, A famous heretic of late antiquity, writes: "Nothing impossible has been commanded by the God of justice and majesty"⁷. Saint Thomas of Aquino, perhaps unconsciously, echoes the heretic: "every sin is voluntary"⁸. These ideas are not very strange, since the supporters of the principle have always claimed a great deal of intuitive appeal. More modern supporters include Sidgwick⁹, Singer, Nagel¹⁰, Parfit¹¹, Tännsjö¹², Furrow¹³ and Vranas¹⁴, from whom the above quotation came.

Though his significance to the modern debate on OIC, and on this paper, is slight, even a brief summary of OIC would be incomplete without at least a mention of Immanuel Kant. A few decades ago, the principle was often referred to as "Kant's Law" or “Kant’s Dictum”¹⁵. Despite this famous attribution, and the fact that versions of the phrase "ought implies can" appears numerous times in several of Kant's works¹⁶, it remains unclear what he thought of it. There is little reason to think that Kant even considered the principle to be a "principle" or a "law", since he never refers to it as such. Kant uses it in his reasoning and argumentation, but does so quite informally. According to Robert Stern's overview in his 2004 paper "Does "Ought" imply "Can"? And did Kant think it did?", Kant probably meant that "ought" implies "can" in a fairly weak sense, clearly different from what we

(9)

will be discussing in this paper. Though this paper’s discussion of Kant is now over, for anyone interested in his view of the principle, I would recommend Stern’s article.¹⁷

Since Kant’s days, the principle has been called upon by many famous philosophers, and it appears that respect for the principle has increased greatly. This despite, or perhaps due, to the fact that the principle remained comparatively uninvestigated. Michael Stocker’s 1971 “Ought” and “Can”¹⁸ and Sinnott-Armstrong's 1984 ""Ought"

conversationally implies "can""¹⁹ were important early sceptics. Starting with those, a small but (I believe) growing group of dissenters have attempted to undermine the principle. Other significant criticisms can be found in Graham 2011²⁰, Stern 2004²¹, King 2019²² and Henne et al 2019²³.

Today, OIC remains important across ethics. The two most important subfields where the principle sees the most frequent use may perhaps be in the discussion surrounding the possibility of moral dilemmas in deontic logic, and the discussion surrounding what can be morally demand of people. The influence on the latter application is quite obvious - if we only ought to do what we can, then it cannot be said that we ought to do what we cannot. The importance of OIC on deontic logic is, however, less obvious.

Deontic logic signifies several attempts to create coherent systems of logic to formalise ethical concepts and how they relate to each other²⁴, or, in the words of Alex King,

“…logical systems that govern necessity and possibility or deontic terms like ought, must, and can”²⁵. As for any system of logic, this requires the establishment of certain axioms that are not allowed to lead to contradictory conclusions For some systems of deontic logic such as Standard Deontic Logic, one example of (a possible) such contradiction would be moral dilemmas, where we would be morally obliged to fulfil two incompatible duties. In other words, we would be given an impossible task. This has led some supporters of deontic logic to conclude that such dilemmas are actually impossible, and they have appealed to the OIC-principle to support this. In other words, if OIC is false, it is possible that at least some parts of or versions of deontic logic collapses²⁶.

But moral dilemmas and deontic logic are mere examples of the principle’s importance, and we will not concern ourselves specifically with them. Sufficient to say, it is an important principle with many and important applications. The great importance and force of OIC comes from two key inferences you can make by appeal to it. They are:

“if you “ought” to do something, you can”, and “if you can’t do something, it is not true

(10)

that you ought”. In other words, OIC provides one possible way of proving that we “can”

do something, and, perhaps more importantly, to disprove that we “ought” to do something. This makes the principle interesting wherever a philosopher wants to prove that something can be done, which happens often enough, or when a philosopher wants to prove that it is not the case that something ought to be done, which is even more common.

Despite the critics and the criticism of OIC mentioned here, the principle is all too often taken for granted and is rarely explained or specified in any detail. In the words of Bart Streumer, himself an OIC supporter:

“Many philosophers believe that the following claim is true: (O) It cannot be the case that a person ought to perform an action if this person cannot perform this action.

According to other philosophers, however, this claim is false. And most philosophers who think that this claim is true do not give arguments for its truth. Instead, these philosophers often merely assume that (O) is true, or merely argue that (O) has not been

shown to be false.”²⁷

The sentiment is echoed in other places in the literature:

“…to my knowledge no systematic and comprehensive defense of the principle against its opponents has been so far undertaken…”²⁸

Or:

“Many philosophers—including Kant and Smith (1933), Sidgwick (1884), Moore (1922), Parfit (1984), Zimmerman (1996), and Vranas (2007)—accept OIC. Several

of these philosophers simply assume that this principle is true or hold that it is an analytic or conceptual entailment.”²⁹

Naturally, it is a bit ironic that the above three quotes all come from papers that does deal in depth with the OIC-principle, and no matter whether they support or reject it gives it a proper investigation. But while it is clear that the literature on OIC has been slowly

(11)

increasing in the last fifty years, it is clear that it is still often taken for granted by philosophers who have not taken a direct interest in the debate. Here is an example from Sue Donaldson’s and William Kymlicka’s 2011 book Zoopolis:

“But circumstances change. Ought implies can, but what we can do changes over time, and so, therefore, does the “ought”. Today, most of us are no longer in the circumstances that would justify imprisoning and killing animals for food, labour or clothing. We have no need to engage in the tragic necessity of harming animals in order

to meet our needs.”³⁰

The quote is not followed or preceded by an explanation of exactly what is meant with

“ought” implies “can”.

As another example of someone who “simply assume that this principle is true or hold that it is an analytic or conceptual entailment”, we can take Sidgwick in his The Methods of Ethics:

“I cannot conceive that I “ought” to do anything which at the same time I judge that I cannot do”³¹

Or Ishtiyaque Haji, who writes in his Deontic Morality and Control:

“One cannot perform an act that is morally right, wrong, or obligatory if one does not have appropriate control over that act; deontic acts presuppose control. This is, perhaps,

most evident in the case of moral obligation. It is difficult to accept the verdict that Leno, though pinned to his seat, ought to have saved the child, or that the paraplegic

ought to have walked across the lawn when “ought” denotes moral obligation or requirement. […] Again, it is paradoxical to suppose that the mother’s not spanking the

child was morally right when she literally lacked the ability to spank her child. These enigmas are not merely pragmatic or do not simply have to do with substantive moral

claims about what is fair but are conceptual; the verdict about the paraplegic, for instance, seems conceptually inconsistent. […] This is the famous “ought” implies “can”

principle and its corollary that “ought not” implies “can avoid”…³²

(12)

The faith in the principle seems somewhat excessive. This is not necessarily because the principle is false, but rather because even if it is true there is a considerable amount to be said for and against it. This will hopefully be made clear in the two following sections where we will attempt to specify exactly what the principles actually says. True or false, the OIC-principle is NOT obvious or above criticism. Though we will in this paper only discuss a single method to attack the principle, there are many other ways to do so that are worthy of attention. See Stern³³, King³⁴, Henne³⁵, Korsgaard³⁶, Graham³⁷ and Sinnott- Armstrong³⁸ for examples.

1.2 What do we want from the “”ought” implies “can””-principle?

A necessary question when investigating the ""ought" implies "can""-principle is: what are we really looking for? The question is less obvious than one might think. This is not only because of the many versions of the principle, but a deeper problem that arises from the very fact that we will proceed to investigate the terminology the principle consists of.

This means that the very meaning of the principle may shift quite considerably, perhaps beyond recognition. Some of these changes are guaranteed to make the principle false, and others are guaranteed to make it true. Especially the fact that some versions (or possible versions) of the principle are guaranteed to be true is problematic. This may seem strange, but consider, for instance, if we define "ought" as "7+5", "implies" as "=" and

"can" as "12". That would make the principle true. But it would not make it very helpful for dealing with ethical problems, and it would be a bizarre change in the meaning of the terms.

The most common but still occasionally endorsed version that sails closest to this charge of being trivially true is the idea that "ought" conversationally implies "can"³⁹. Basically, that "ought" would imply "can" in a very informal sense, as in "when people say that a person “ought” to do something, they usually mean that they think that particular person also “can” do it". The result of this is, in the words of Sinnott-Armstrong, that: “If 'ought' conversationally implies 'can', and the agent cannot do the act, then it might be true that the agent ought to do the act”⁴⁰. In other words, that OIC is just a matter of

(13)

custom or habit, and there is no necessary relationship between "ought" and "can", then it might very well be true that someone “ought” to do something, while at the same time it is true that this someone “can’t”. In other words, we would be unable to logically go from “no can” to “no obligation”.

As can clearly be seen, this “conversational implicature” version of the principle is a lot weaker than most alternative formulations. Indeed, it is so weak that it is usually considered a rebuttal and rejection of the principle⁴¹. If OIC is so very weak, the argument goes, it is simply too different to be considered a true OIC anymore. In other words, those who consider OIC-as-a-conversational-implicature to be a rejection of OIC holds that we accept the principle as such it is too different from what we originally believed it to be. It has changed beyond recognition, and can no longer be said to be the same principle. This applies even if one should accept that “ought” does conversationally imply “can”.

With this we are presented with a new problem: where exactly does this line between acceptable and unacceptable change go? How far can the OIC-principle change and still remain true to its core? Even Standard OIC (see the next section) comes with quite substantial changes. For instance, "ought" is almost invariably understood as synonymous with having some kind of "obligation".

In the end, it is clear that it is a very unclear matter. The question of how much OIC can change and still remain OIC is not so much for us to discover as for us to decide. We already know that there are those who accept OIC as a mere conversational implicature, while most philosophers consider that to be a rejection. At the same time, we cannot say that those who both accept OIC as a conversational implicature and thereby consider it to be true are wrong. There is a wide range of possible nuances and shades of OIC that we can accept or reject.

This does, however, not mean that "anything goes", which would present us with a kind of sorites-problem. We can argue how many grains of sand you can remove from the pile before it no longer can be called a "pile", and we can argue how many hairs a man must lose before he turns “bald”, but in the end we must all accept that there is some limit to these. Sooner or later we must put our foot down, lest we end up with an absurd consequence, like the one we discussed a few paragraphs ago, where "ought" means "7+5",

"implies" means "=" and "can" means "12".

The understanding of the OIC-principle as conversational implicature is typically

(14)

classified as part of a larger group of “pragmatic” understandings of the principle.

Consider the following more colloquial phrasings of it:

1. "It is unfair to demand something of someone who cannot deliver it".

2. "It is bad advice to advice someone to do something he cannot do".

3. "Sunk costs should not influence future decisions".

4. "Don't cry over spilt milk".

5. "Accept you mistakes and move on".

6. "The past is the past, look to the future".

7. "We shouldn't blame people for not doing things they cannot do".

8. "If a rule or law is excessively demanding, we should try to modify it".

All of these could be seen as down-to-earth versions of the ""ought" implies "can""- principle. They are also, overall, quite sensible suggestions. At least provisionally, I would agree with all of them.

There is, however, a very important distinction that must be made between general folksy good advice, such as the eight points above and (as we will see in the next section) the unquestionable philosophical principle OIC is typically presented as. We cannot go from one to the other, in either direction. There is plenty of good advice that should not be axioms in ethical systems, such as "cough into your elbow", and plenty of axioms of ethical systems that would be catastrophic if people started following them uncritically, such as the utilitarian "the only thing that matters is pleasure".

The main reason for this situation being problematic is that OIC is typically interpreted not as general good advice, but something to help us gain knowledge and make deductions. The purpose and idea behind the principle is as something that allows us to pass either from the fact that we "ought" to do something to the fact that we "can"

do something, or, as something that allows us to pass from the fact that we "can't" do something to the fact that we "not ought" something. The OIC-principle is about, or supposedly about, how we can gain this knowledge, and not merely to advise us on what the best available strategy would be.

One could say that this distinction is between the ethics of the philosophers and the day-to-day actions of the man on the street. There is nothing that necessitates that these

(15)

two realms of human action will have much in common with one another. We philosophers are not widely known for giving very useful advice to the common man, and the moral ideals that guide the common man are rarely accepted uncritically by philosophers. What happens if OIC is not literally true but only pragmatically so is that it moves from one realm to the other. Basically, OIC would then cease to be philosophy as concerned with knowledge of morality, its terminology and its mechanics, and instead become some very general good advice. This corresponds fairly closely to Kant’s distinction between “anthropology” and “practical reason/morality”, or between the empirical part of actions and the purely rational part⁴². If OIC moves from one category from one category to the other, it ceases to be part of the philosophical study of ethics, and becomes merely “a good tip” for action.

To understand why a merely pragmatic version of OIC is problematic we have to consider what we want the principle to do for us. This quickly leads us to a distinction between two versions of what a “can’t” or “cannot” does to an “ought”. In the first sense, the lack of a “can” means that the “ought” disappear, ceases to exist like a puff of smoke.

In the second sense, the lack of a “can” only means that the “ought” is temporarily deactivated or disabled, but still remains. In other words, a distinction between the

“”ought”-deactivating “can’t”” and the “”ought”-dispelling “can’t””.

The “”ought”-deactivating “can’t”” is the pragmatic/anthropologic/informal/everyday version of OIC, illustrated by the eight-pointed list above. It does not deny that an “ought”

or “obligation” does remain even if we cannot fulfil it, only that the fact that we cannot fulfil it means that we should disregard it in our everyday decision making. This is a strategic OIC, rather than a descriptive OIC. It advises us on how to act, rather than about how the world actually is. It is the difference between “if you can’t fulfil an obligation, don’t bother with it and pretend as if you don’t have it”, and “if you can’t fulfil an obligation, you don’t have an obligation”.

There are two main problems with an exclusively pragmatic OIC. The first is not that it is false but rather that it is trivial. Of course we should let what we can and can’t do affect our decision making! Who would doubt that? When picking a course of action, it is perfectly natural that we consider both our moral goals and our abilities and opportunities. It is certainly excellent if OIC gives useful advice when picking a course of action, but if that were all it did it would become trivial. No sensible philosopher denies

(16)

that we should, as far as possible, chose means that actually allows us to reach our goals, whatever they may be.

The second problem for pragmatic OIC is the fact that it cannot help us with the two key inferences OIC is taken to be able to help us with:

Smith ought, therefore, Smith can.

Smith cannot, therefore, it is not true that Smith ought.

In other words, OIC must allow us to go from “ought” to “can”, and from “not-can” to

“not-ought”. If it cannot do this, we should consider it to be such a different or weak principle that it has changed beyond the permitted. Even a “true” version of the principle but which cannot handle these inferences should be rejected as too weak. Since the pragmatic OIC is just that, pragmatic, it simply advises us on what to do. It is quite compatible with some sort of moral obligation remaining even in the absence of “can”.

Consider this example from Derek Parfit’s Reasons and Persons, where he uses OIC in order to dispel (as opposed to temporarily deactivate) the theory that we should always act in our own self-interest:

“It may now be said that, as described by Kate, S [Self-interest Theory] lacks one of the essential features of any theory. It may be objected: “No theory can demand what is

impossible. Since Kate cannot always avoid doing what S claims to be irrational, she cannot always do what S claims that she ought to do. We should therefore reject S. As

before, ought implies can.”⁴³

What Parfit here does is an example of the second of the two key inferences: “Smith (or “Kate” in Parfit’s example) cannot, therefore, it is not true that Smith ought”. This is a very clear example of the “”ought”-dispelling “can’t””, where it is not only true that Kate should temporarily modify her course of action because of her inability, but that the entire obligation placed on her by the Self-interest Theory disappears. Parfit has (to his own satisfaction) shown not only that we should temporarily ignore Self-interest Theory or that we fail in our obligation to obey it, but that the entire theory fails and goes away:

there cannot be an “ought” if there is no “can”. This is a conclusion that can never be

(17)

drawn from a merely pragmatic version of the principle. A pragmatic version could at best recommend us to temporarily disregard Self-interest Theory.

This gives us our desideratum for the OIC principle, and for this investigation of it.

The OIC-principle must be able to handle the two inferences, to go from “ought” to “can”

and from “not can” to “not ought”, and it must be able to do so by dispelling rather than merely deactivating obligations. If the principle cannot help us do so, we will consider it to have failed, and if it has failed, it should be rejected. This, that it cannot fulfil this desideratum, is what we will attempt to demonstrate in Part 2.

1.3 What does ""ought" implies "can"" mean?

A significant problem when it comes to studying the ""ought" implies "can""-principle is that there are a large number of versions of it about, many of which are quite different from each other. Since it is not possible and, quite frankly, not very desirable, to deal with all of these minor versions, what we will do is to focus on the most widely accepted and well defended one. Though not always explicitly spelled out, most philosophers who accept OIC, such as Hajj, Parfit, Vranas and Streumer seems to accept it or something very close to it. In order to achieve precision we will largely follow what Alex King in her 2019 book "What We Ought and What We Can" calls “Standard OIC”, or the “Standard Version”⁴⁴. She formalises it as follows:

“An agent's being morally obligated to perform some action analytically entails that that agent is able and has the opportunity to perform it”⁴⁵.

To this we could make a quick comparison to Bart Streumer:

“(O) It cannot be the case that a person ought to perform an action if this person cannot perform this action.”⁴⁶

Or Peter Vranas:

(18)

“(OIC) By virtue of conceptual necessity, if an agent at a given time has an objective, pro tanto obligation to do something, then the agent at that time has both the ability the opportunity to do the thing.”⁴⁷

Or Paul Henne:

“(OIC) For any agent, A, and act, C, if A ought to C, then A can C.”⁴⁸

To understand the “standard version of OIC”, we will - according to Alex King - have to answer six questions regarding OIC :

“(1) what ‘ought’ means, (2) what ‘implies’ means, (3) what ‘can’ means,

(4) what the subjects of the ‘ought’ and ‘can’ are, (5) what the objects of the ‘ought’ and ‘can’ are, and (6) how to temporally index these different elements.”⁴⁹

The first three questions are most relevant to this paper, and we will deal with them here. Hopefully, this brief discussion will serve the double purpose of explaining the principle, and also to make clear that the OIC principle is not as simple and clear-cut as is sometimes assumed.

1. What does "ought" mean? There are two main contenders: the first one being the “ought as obligation”, or “pro tanto ought”. The contender is the “all-things- considered ought”, also known as "ought simpliciter". Though we will in this paper accept

“pro tanto obligation” rather than “ought simpliciter” as the proper “ought”, most of the time the distinction does not matter.

To be "obliged" to do something is to be bound, to have a duty, to be committed. If you are obliged to perform a certain act you cannot refrain from doing so without failing that obligation, which is bad thing to do. In the words of Peter Vranas (who’s formalisation of the principle probably inspired Alex King):

(19)

“I understand the claim that an agent ought to do something as the claim that the agent has an objective, pro tanto obligation to do the thing. The obligation is objective in the sense of corresponding to what is in fact the case about the agent’s situation, not necessarily – as a subjective obligation does – to what the agent is epistemically justified

in believing to be the case. The obligation is pro tanto (i.e., prima facie) in the sense of corresponding to some reason or combination of reasons for action that apply to the agent in her situation, not necessarily – as an all-things-considered obligation does – to

the combination of all such reasons.⁵⁰

One could perhaps say that to claim that someone “ought” to do something is to state that there is a factor, a reason, a motivation, or perhaps a “weight on the scale of morality”

pushing them to do whatever it is they “ought” to do. The difference between pro tanto obligation and all-things-considered ought is then that the all-things-considered ought would consider all weights on the moral scale together, while the pro tanto ought would concern each one of them separately.

For example, if I were considering whether to give money to a beggar, the equation (with completely arbitrary numbers) might look like this: Being helpful (+10), supporting idleness (-3), losing money (-1), supporting shady lifestyle (-20), acting with good intentions (+3), giving in to sentimentality (-2). This would all add up to a total of -13. In other words, I ought not to give money to the beggar. From the all-things-considered view of ought, the “ought” would then be the total sum, -13. According to the pro tanto obligation version of “ought”, each and every individual factor (+10, -3, -1, -20, +3, -2) would qualify as an “ought”, positive or negative.

It is important to note that one thing that the “pro tanto obligation” does not consider is whether we can do something. Or to be more precise, it is not something we can just assume. After all, it is this paper’s purpose to find out in what way “ought”, “pro tanto obligation” and “can” are related to one another. The OIC supporter would naturally be forced to maintain that “pro tanto obligation” implies “can”, at least if “you have a pro tanto obligation” is supposed to be interchangeable with “you ought”. For now, at least we will have to keep it an open question whether “pro tanto obligation” implies “can” – hopefully, we will have an answer at the end of the paper.

(20)

Yet another important thing to note is that “ought” and “obligation”/“pro tanto obligation”, strictly speaking, are not synonyms. In the words of Vranas, just quoted above: “I understand the claim that an agent ought to do something as the claim that the agent has an objective, pro tanto obligation to do the thing”⁵¹. In other words, if someone tells you that “you ought to be kind”, that means that whoever uttered the sentence claims that you have an obligation to be kind. “You ought” does not mean “you obligation”, rather, it means “you have an obligation”. The reason that “ought” and “obligation” cannot be exact synonyms is perhaps best illustrated by pointing out that they belong to distinct word categories. “Obligation” is a noun, whilst “ought” is a verb in imperative form.

For all intents and purposes, this slight distinction will not matter much for our discussion, and with very few exceptions “ought” and “obligation” can be used interchangeably. That someone has an “obligation” means that they “ought” to do something, and that they “ought” to do something means that they are under an obligation.

2. What does "imply" mean? This is probably the question most critical to this paper, and its interpretation will have a very significant influence on the main argument presented in Part 2.

The "imply" in the principle is in Standard OIC taken to mean “conceptual implication”, "analytical implication" "analytical entailment", “conceptual entailment” or

“definitional truth”. All these come down to the same thing, namely that it is an analytical truth that "ought" implies "can", or that the word "ought" somehow also includes the meaning of the word "can". Basically, we can go from “ought” to “can” in the same way we can go from “bachelor” to “unmarried”, because the second term is just a “part” of the first term⁵². In other words, ““ought” implies “can”” could also be interpreted as “”ought”

includes the meaning of “can””. Just as “bachelor” is defined as “unmarried man”, the definition of “ought” supposedly includes “can”, perhaps alongside a few other things.

Because of the importance of the meaning of “implication” to the main argument, we will return to this question in section 2.1.

In the words of Peter Vranas:

“I understand the claim that one proposition implies another as the claim that, by

(21)

virtue of conceptual necessity, the latter proposition is true if the former is. (I use

“proposition” and “claim” interchangeably.) So I understand implication as conceptual entailment, such as the entailment from – the proposition expressed by – “Smith is my

uncle” to “Smith is male”.”⁵³

There are a few other interpretations of “imply” available, though we will not deal with them in this paper. Most are weaker than conceptual implication. One could once again mention the “conversational implication” of Sinnott-Armstrong, where the implication simply means that we usually also mean “can” when we say “ought”⁵⁴. Another is the

“presupposition view”, where “ought” “presupposes” “can” in the same manner as someone who says “my cat is on the mat” can be presupposed to own a cat⁵⁵.

3. What does "can" mean? According to Alex King, “Of all the components of OIC,

‘can’ has received the most attention. ‘Can’ is a wildly flexible word, and it is used in all sorts of ways. Most of these have, at some point, been invoked in discussions of OIC”⁵⁶. Though the exact meaning of "can" remains controversial both inside and outside the debate concerning the OIC-principle, we are in this context typically meant to

understand "can" as a combination of "opportunity" and "ability". Anyone who is not chained to the floor may have the opportunity to perform a somersault, but few possess the ability. On the other hand, even if someone does possess the ability to perform a somersault, that will not help if he or she is tied to the floor. To properly say that we

"can" do something, we must therefore possess both opportunity and ability.

In the words of (once again) Peter Vranas:

“I understand the claim that an agent can do something as the claim that the agent has both the ability and the opportunity to do the thing. The agent has the ability to do the thing in the sense of having the requisite skills, physical capacities, and knowledge – even if psychologically she is “unable” to do it (e.g., she “cannot” get herself to stick her arm into a cesspool to retrieve her wallet…”⁵⁷

The exact meaning of "can" plays an important role in the scope of the OIC-principle, since it determines how strongly the principle is to be interpreted⁵⁸. If we accept a very

(22)

wide interpretation of "can", such as "you "can" do something as long as doing it involves no logical contradiction", that would mean that we would be able to do a very large number of things indeed if we "ought" to do them. On the other hand, if we had a fairly narrow interpretation of "can", such as "you "can" do something if you have more than 50% chance of succeeding if you try", that would mean that the possible interpretations would be far fewer. The number of logically possible things is, of course, infinitely higher than the number of humanly doable things.

A danger with accepting a wider definition of "can" is that we by doing so would run the risk of making it trivial. After all, there are plenty of logically possible things that are, for all intents and purposes, impossible. We might, for instance, logically run faster than the speed of light, but that is still physically impossible. If all the principle does is to graciously grant us such a permission, it would be applicable in far fewer circumstances, making it rather toothless. It might, under such circumstances, still be the case that we

“ought” to do things that are for all intents and purposes impossible. On the other hand, if we demanded that the principle gave us a decent chance of success we might end up in a situation where we "ought" to do only quite easy things. On the one extreme we would have an OIC-principle that demands the doing of the nigh impossible, and on the other we would have an OIC-principle that excuses virtually anything⁵⁹. As is often the case, a well-functioning “can” must land somewhere between these two extremes.

Let us thus repeat the principle, in Alex King’s Standard Version, hopefully much clarified:

"An agent's being morally obligated to perform some action analytically entails that that agent is able and has the opportunity to perform it."

Naturally, this definition is not set in stone. The exact meaning of the principle and its parts will remain a topic of interest and discussion throughout this paper, especially in sections 2.2 and 2.3. After all, the point of this paper is to challenge the above definition, and argue that it cannot do what we want it to do.

(23)

2.0 The Main Argument

We will now get started on the second part of this paper. In the first section of the previous part we explained and investigated the OIC-principle, its history, literature and use. In the second section we moved on to detailing a key desideratum that will now grow important as we criticize the principle, namely that the OIC-principle must cause an obligation to disappear when it cannot be fulfilled. In the third section of the first part we attempted to understand and specify exactly what the principle means and claims.

The argument of this second part attempts to demonstrate that the OIC principle cannot, if understood as “ought” conceptually implying “can”, fulfil the desideratum of showing that inability and obligation cannot coexist. The reason the principle cannot do so is because there, if “ought” conceptually implies “can”, always will be possible to make a clear distinction between the part of the term “ought” that concerns morality, and the part of the term “ought” that concerns “can”. In other words, if “can” is removed, so may

“ought”, but not necessarily the part of “ought” that concerns morality, which may remain even with “can” - and thereby “ought” - gone. “Ought” may require “can”, (whether it does is a matter of how we chose to define it), but even if we get rid of “ought”, the moral part may remain. And since it is that part that we are interested in getting rid of when using OIC, OIC (at least as a conceptual implication) fails.

Section 2.1 follows G. E. Moore in elaborating the notions of “simple” and “complex”

terms. This is important for the succeeding steps of the argument.

Section 2.2 applies the analysis of complex and simple terms to “ought”. In order for

“ought” to conceptually imply “can”, “ought” must be a complex term. “You ought” or “you have an obligation” can then be broken up into “”X”+”can””, where “X” is the part of

“ought” that remains after “can” has been removed. Though not the only one, the most plausible interpretation of “X” is suggested to be “pro tanto obligation”. For that to be possible, we argue that despite what most OIC supporters think, there is a distinction between “you ought” and “you have an obligation”.

Section 2.3 applies Ockham’s razor to the interpretation of “ought”. It is argued that simpler terms are preferable and give greater clarity. We should therefore seriously consider to stop using the term “ought” and instead replace it with its component terms.

Section 2.4 is perhaps the most important part, and argues that if “ought” is

(24)

understood as the previous sections have argued that it must, then it does not fulfil the desideratum presented in section 1.2.

Section 2.5 clarifies a few points surrounding the view of “ought” presented here, especially the difference between the OIC-principle being false and the fact that it should be rejected. It is explained that the OIC-principle may very well be literally true simply by intentionally defining “ought” as implying “can”, but doing so does not mean that there can be no obligation without the presence of a “can”.

Section 2.6 summarizes the argument.

2.1 Simple and Complex Terms

What we are about to do is to discuss the structure of the term “ought”. As we made clear in section 1.3, the “”ought” implies “can””-principle is typically understood as a conceptual truth. In other words, that the term “ought” must convey the meaning “can”.

Whenever we say “you ought” we therefore, if the principle is true, also say “you can”. This means that “ought” must be something called a “complex term”, or, a term with many parts. The purpose of this section is to illustrate exactly what is meant by this.

A term is “simple” if it consists of only a single part, which makes it indivisible and unanalysable. If a term is simple, no part of its meaning can be removed without rendering it nonsense. Complex terms, on the other hand, are the opposite of this. A complex term consists of more than a single part, and can be divided yet retain at least some meaning, though not its original one. This usage follows G. E. Moore's use of the

"simple term" or "simple notion"-term in his Principia Ethica⁶⁰. In his own words (with

“good” and “yellow” as examples):

“…”good” is a simple notion, just as “yellow” is a simple notion, that, just as you cannot, by any manner of means, explain to anyone who does not already know it, what

yellow is, so you cannot explain what good is. Definitions of the kind that I was asking for, definition which describe the real nature of the object or notion denoted by a word,

and which do not merely tell us what the word is used to mean, are only possible when the object or notion in question is something complex. You can give a definition of a

(25)

horse, because a horse has many different properties and qualities, all of which you can enumerate. But when you have enumerated them all, when you have reduced a horse to his simplest terms, you can no longer define those terms. They are simply something which you think of or perceive, and to anyone who cannot think of or perceive them, you

can never, by any definition, make their nature known.”⁶¹

Let us take a few examples. "Bachelor" is not a simple term, (and must therefore, by Moore’s dichotomy, be a complex term), since it contains at least two parts: anyone who is a "bachelor" is: 1. A Man. and 2. Unmarried. The word "bachelor" can therefore be analysed and broken down into its components, "man" and "unmarried". Examples of simple terms could be, to use G. E. Moore’s examples, the terms "good" and "yellow"⁶². Neither of those term can be broken down into smaller concepts in the manner that

"bachelor" can.

The easiest test to determine whether a term is simple or complex is to ask whether the term can be described. If I want to describe to someone what a complex term like

"bachelor" is, I simply tell them something along the line of "a bachelor is an unmarried man". But how would I go about trying to explain a word if it were simple? What words would I use to describe the concept of "yellow"? I could of course point to something that is yellow, or provide synonyms, but neither of those options is quite enough. I could perhaps say "the part of the colour spectrum with a dominant wavelength between 570 and 590 nanometres", but I would still only have been successful in providing examples or explaining which objects we call "yellow", but not the concept itself. There is, after all, a difference between the concept of "yellow" and the objects that possess the property. A blind man would not understand the concept of “yellow” by being told about “colour spectrums” or “dominant wavelengths”.

The key to the simple terms is their indivisibility. They are for that reason sometimes called "atomic" terms, in the ancient Greek sense of signifying something indivisible.

Simple terms cannot be broken up without turning them into nonsense. All complex concepts, on the other hand, consist of other concepts. Some of these concepts may in turn - if they are not simple - consist of still other concepts. Complex concepts are constructed out of other concepts, and for that reason they may also be reduced to them.

Let us take a few examples:

(26)

A "horse" is a...:

Quadrupedal Mammal

of the genus equus

A "bachelor" is a...:

Man

Unmarried

A "vixen" is a...:

Fox Female

The complex terms, being opposed to the simple terms, are distinguished by the fact that they, in the just used formulations, contain more than one word in the description below the term. Naturally, there is no limit to how complex a given term can be, and many of the words we put below the name may very well be other complex term. The term "fox", for example, that we included under the term "vixen", could in turn be described as a 1.

Quadrupedal. 2. Mammal. 3. Of the genus vulpes. A simple term, on the other hand, would if we attempted to describe it as we have done above could only contain itself, or a synonym:

Yellow:

Yellow

Good:

Good

The point that we are trying to make is, naturally, that the “ought” of the OIC-principle is a complex term. This is because Standard OIC claims that “ought” “conceptually entails” “can”. A simple term is, as we have just discussed, a term that does not contain

(27)

other terms aside from itself. But "can" is a term. So, if "ought" is to include "can", then it means that "ought" would include at least two terms, "can" and whatever is left of the

"ought" once the "can" has been removed. A simple term cannot, as a matter of definition of what a simple term is, include more than one term. Thus, if the OIC-principle is to be true as a matter of “ought” conceptually implying “can”, then “ought” cannot be a simple term.

2.2 “Pro Tanto Obligation” + “Can” = “Ought”?

But if “ought” is a complex term, what are its parts? First, it is clear that if “ought” is complex, it must (from the fact that it is complex) contain at least two but possibly more parts. And if “ought” is to conceptually imply “can” as we discussed in 1.3, one of those parts must be “can”. The other part (or parts) of “ought” must convey some moral meaning. After all, “ought” cannot be perfectly synonymous with “can”, it must contain something more than that, something that gives its relevance to ethics. Therefore, at least for now, we can say that “ought” = “”X”+”can””, where “X” is the part of “ought” that is not “can” but rather something else, at least partly something with a moral meaning.

Interestingly though perhaps not very importantly, it is probably also the case that the term “can” is not simple either. This provides a good example of how complex terms can include other complex terms. In the case of “can”, this is because, as we made clear in 1.3, that it signifies a combination of “ability” and “opportunity”. “Can” might otherwise have been a candidate for being a simple term. Perhaps a deeper analysis would reveal it to be so, perhaps by challenging whether “ability” and “opportunity” may be defined without reference to “can”. This is, however, not a debate we will get into here. This is because it will be largely irrelevant to the argument against OIC that we are building up to. The argument works as long as “ought” is a complex term. Aside from that, it does not matter exactly how complex it is and how numerous its parts are.

Moving on: what, then, is “X”? First, I would like to make perfectly clear that it does not really matter. The main argument of this text works just as well even if “X” remains unknown. This is because we are arguing against the OIC-principle, the idea that there is a necessary connection between the two parts of “ought”, “X” and “can”. As long as we can

(28)

show that no such connection is necessary, our argument holds even if “X” is never defined. More about this in 2.4.

Even so, a brief investigation into “X” is still worth doing because there are some basic characteristics about “X” that should be made clear, and because there is also be some value in knowing what “X” might be.

We already know a few things about “X”: “X” is normally part of “ought”. “X” is either simple or complex. “X” is (ex hypothesi) not the same as “can” and does not include it.

“X” conveys some form of meaning pertaining to morality.

Perhaps slightly more controversially, “X” must probably also be a fully functional and intelligible concept, capable of functioning on its own even when separated from “can”.

This is because the concept in question was apparently part of “ought” without causing any problems, and is unlikely to have grown incoherent just because it has been separated from “can”. If one would deny this, one would have to explain how “X” alone is to be incoherent and dysfunctional, without also spreading this incoherence and dysfunction when united with “can” in the concept of “ought”. (Unless one also considers “ought” to be incoherent, in which case the entire debate is void).

The coherence of “X” matters partly because of what it suggests of its identity, of what term hides behind it. If “X” is a perfectly coherent term, it is probably already known to us. After all, if OIC is true we are already familiar with “X” when combined with “can”. It does not seem very likely that a spectacularly different or hitherto unknown ethical concept would appear out of “ought” just because we remove “can”. Whatever “X” is, it must be able to conceal itself within the confines of “ought”.

I believe that our prime suspect for being “X” should be “obligation”. More specifically,

“pro tanto obligation”. At first glance, this may seem like an absurd conclusion. After all,

“ought” is often explicitly defined as someone having a “pro tant obligation”. We spent much of section 1.3 to establish that fact, and there quote several philosophers who use the two terms interchangeably. So how can “X” be “pro tanto obligation” if “X” is a

“canless-ought” and if “pro tanto obligation is the “”can”-including “ought””? If “X” is a part of “ought”, and “pro tanto obligation” is the whole of “ought”, one would think that

“X” would be less than “pro tanto obligation”, and therefore not identical with it.

All this is true. If “you ought” signifies that “you have a pro tanto obligation”, and “you ought” implies “you can”, then “you have a pro tanto obligation” must also imply “you

(29)

can”. The solution to this problem is very simple: I believe that those who use the two terms “ought” and “pro tanto obligation” (or “you have a pro tanto obligation”) are mistaken. Or more precisely, they are mistaken in believing that there is a necessary connection between “pro tanto obligation” and “can”.

At first sight, this might seem to be quite radical. Surely – philosophers would know the meaning of the terms they use? This is indeed strange. It is, however, a necessary assumption one must make if one is to discuss the OIC-principle at all, no matter which side of the debate one picks. Obviously, if I am to deny the OIC-principle (which is the purpose of this paper), I must maintain that some philosophers do not understand the meaning of the term, specifically those who accept OIC. Likewise, if one wishes to defend OIC, one must presumably also be aware that there are philosophers who deny it and who, if we assume that OIC is correct, do not know the meaning of the term they use. That philosophers cannot be so confused about their terminology as to have missed that

“ought” is ““pro tanto obligation” + “can”” is therefore no defence, since the entire debate concerning OIC requires us assume that they in fact can be so confused.

But why think that “X” is “pro tanto obligation”, rather than something else? First, whatever “X” is, it is unlikely to be very strange. After all, it must be something so close to

“ought” that it can be included in it. “Pro tanto obligation” fulfils this criterion, since it is, as the OIC-supporter shows, possible to believe that it is identical with “ought”.

Furthermore, a “pro tanto obligation” without “can” is already accepted by most of the philosophers who today reject “ought”. That one rejects OIC and believes that “ought”

does not imply “can” does not mean that one thinks that there is no such thing as

“obligation” – it merely means that one thinks it possible that to have an “obligation”

without having a “can”. Therefore, that “X” is “pro tanto obligation” is more or less the default position if OIC is false.

Second, there is the matter of intuition. Or to be more precise, its absence. The

““ought” implies “can””-principle is typically defended at least in part as a matter of intuition. That “obligation” implies “can”, on the other hand, is rarely defended so. It appears, to my mind, that the idea that “obligation” implies “can” is an indirect carryover from the original notion that “ought” implies “can”. In other words, the OIC supporter reasons as follows: ““I have an intuition that “ought” implies “can”. Therefore, “ought”

implies can”. “Ought” is the same (or nearly the same) as “obligation”. Therefore,

(30)

“obligation” must also imply “can””. The problem with that form of reasoning is however that it could be turned on its head. Rather than going from ““ought” intuitively implies

“can”” and ““ought” signifies “obligation”” to ““obligation” implies “can””, we might instead go from ““obligation” does not intuitively imply “can”” and ““obligation” signifies

“ought”” to a contradiction. We could on those grounds start to suspect that maybe

“obligation” and “ought” are not as closely related as we think.

Naturally, arguments from intuition are treacherous. The OIC-supporter could to the reasoning of the last paragraph simply reply the he or she really, really does have an intuition that “obligation” implies “can”, every bit as strong as that “ought” implies “can”.

I myself have a somewhat far-going dislike of arguments from intuition, and would not like to make one myself. All I want to do here is to criticise the intuition-based arguments of the OIC supporters. This is especially true, since it is doubtful how widely shared these intuitions are⁶³.

In sum: “X” is (probably) “pro tanto obligation” because:

1. “X” should be coherent.

2. “X” should be a familiar concept, rather than completely new.

3. “X” should be so closely associated with “ought” to explain the confusion.

4. “X” should convey some moral meaning.

5. We need an explanation of what “obligations” would be if it is not tied to “can”.

6. There is a lack of better candidates than “pro tanto obligation” for being “X”.

Ultimately and importantly, it does not matter to the main argument of this paper whether “X” is “pro tanto obligation” or something else. “Pro tanto obligation” is the main suspect because it is the default position when rejecting OIC – a “canless ought”. Keeping this in mind, we will still proceed throughout the rest of the paper by using “Pro Tanto Obligation” instead of “X”. If the reader is suspicious of this, he or she can mentally exchange it for “X”. It will not affect the validity of the main argument, which only states that we cannot conclude that we have no reason to dismiss the possibility of an “X” just because we have no “can”, since it is possible to separate the terms. It would not be fatal for the argument if “X” turned out to be something other than “pro tanto obligation”.

(31)

2.3 Simpler terms preferable

The question this gives to us is: if “ought” is constituted out of simpler terms, why not use those terms directly? It may naturally be harder to say that “Smith has a pro tanto obligation to wave (alternatively “Smith has an “X” to wave), and Smith can wave” than to say “Smith ought to wave”, but if OIC is a conceptual truth, both would be equally correct. And though it is a mouthful, the unabbreviated formulation of the sentence more clearly conveys the intended meaning than does “ought”.

All other things equal, simplicity is to be preferred over complexity. This suggests that we should abandon “ought” entirely and replace it with the term’s component parts.

Or at the very least clarify that “ought” is merely an abbreviation of the more correct

““pro tanto obligation” + “can””.

This is all in accordance with Ockham’s famous razor. Grammatically, it is of course more complicated to write that “Smith has a pro tanto obligation to wave and Smith can wave” than to simply write “Smith ought to wave”, but if we consider the apparently complex meaning of “ought” then it is clear that we risk missing a lot of its content. There always comes a point when our attempts to simplify merely lead to confusion, and this appears to be one of those.

The whole debate regarding exactly what is meant with “ought” can in this way be avoided. This for the simple reason that we would no longer be using the term. There might of course be arguments surrounding the exact meaning of its component terms, but even that means that we will have reached a deeper level of analysis. We have, so to speak, opened the box that is “ought” and are now discussing its content. Perhaps the simplest thing to do now would be to get rid of the box entirely, and just keep its content out in the daylight, plain for all to see?

Either way, today philosophers still use “ought”, and we must deal with it.

2.4 “X” + “Can” Does Not Fulfil the Desideratum

(32)

Why can’t the OIC principle fulfil the desideratum detailed in section 1.2 if it is broken up into “”X” + “can””? We wanted OIC to help us deal with inferences of the following type:

Smith “ought”, therefore, Smith “can”.

Smith “can’t”, therefore, it is not true that Smith “ought”.

Can OIC as a conceptual implication help us make these inferences? And in the

“dispelling” rather than “deactivating” sense? In one way, yes, OIC can do it. In another, more important way, OIC cannot do it. As a matter of logic, if we define the concept of

“ought” as including the concept of “can”, we may go from “can’t” to “not-ought”, no problem. This works in the same way as we can go from “dog” to “mammal”, if “mammal”

were to be included in the concept of “dog”. But the problem starts if we try to apply this knowledge to the outside world. What happens then is that we are forced to ask ourselves if we apply the terms correctly.

The desideratum was the question of whether an “X” or “pro tanto obligation” can remain even when we cannot fulfil it, whether a “can’t” dispels, rather than, as pragmatic OIC would have it, merely deactivates the moral content of “ought”. If a version of OIC stronger than the pragmatic one is correct, it must do both. But a “pro tanto obligation”

is, if the reasoning in the previous sections has been correct, something different from

“ought”. “Ought” is a combination of some moral meaning with a “can” added to it. The problem that arises is that if we disprove “can”, then we have also disproved “ought”, since

“can” is part of “ought”. But that is not the same thing as us having disproved the moral part of the term. It is still very possible that we may have, say, a pro tanto obligation, despite not having a “can”. And that possibility is what we are interested in, and what the OIC-principle must disprove if it is to be accepted.

Basically, it does not matter whether “ought” implies “can”. We can make it do so merely by intentionally defining “ought” as implying “can”. That would make OIC trivially true, a mere matter of how we humans decide to define “ought”. What matters is whether the moral content remains even after we cannot fulfil it. And that is something we cannot decide with a mere definition. Therefore, the question of whether “ought” implies “can” is an uninteresting matter. What really matters is whether we can have an obligation

“Obligation”, “Ought” and “Can” “Plikt”, ”borde” och “kan”