Sunday, May 15, 2016

Volume 22.1


A Philosophical Review

Volume 22 • Number 1
Ó May, 2016

INSIDE THIS ISSUE:                                                                                                     

Matt Silliman
The Is-Ought Non-Problem 
David Kenneth Braden-Johnson
The Problem with the Is-Ought Non-Problem
Reply to Silliman                                                                   

Gerol Petruzella
Must Ought Imply Can?
Reply to Vranas 

Paul Nnodim
Utilitarian Dilemmas in the Literature of

Brett Belcastro
Reasoning and Intelligence                                                               

Matt Luz
On the Formulation of Successful Ad Baculum

J. Stanley Yake

The Is-Ought Non-Problem
Matt Silliman
With famous and characteristic irony, Hume lampoons a particular form of reasoning:

“In every system of morality, which I have hitherto met with, I have always remark’d, that the author proceeds for some time in the ordinary way of reasoning, and establishes the being of a God, or makes observations concerning human affairs; when of a sudden I am surpriz’d to find, that instead of the usual copulations of propositions, is, and is not, I meet with no proposition that is not connected with an ought, or an ought not. This change is imperceptible, but is, however, of the last consequence. For as this ought, or ought not, expresses some new relation or affirmation, ‘tis necessary that it shou’d be observ’d and explain’d; and at the same time that a reason shou’d be given, for what seems altogether inconceivable, how this new relation can be a deduction from others, which are entirely different from it.” (A Treatise of Human Nature, Bk. III, pt. 1, sect. 1)

I will offer a brace of arguments to suggest that, while formally problematic in certain highly artificial or abstract examples, such reasoning is quite natural and, interpreted with a modicum of charity, seldom if ever fallacious in itself. Taken together, I think these arguments give us good grounds for not worrying much about Hume’s problem.

(1) My first argument rests on the observation there is robust, though tacit, morally normative content, properly and naturally understood, embedded in most ordinary language of human consequence (including Hume’s “…observations concerning human affairs…” above). Take the following argument:

            1. Person X is person Y’s mother.
            2. Therefore, Person Y has obligations (of gratitude, respect, patience, etc.) toward person X.

Here a merely factual premise yields a strong inductive, and morally normative, conclusion without fallacy, since the term ‘mother’ describes a complex biological, affective, and normatively charged relationship out of which such obligations almost always flow. We could, if we chose, treat it as enthymemic, and render it deductively valid by supplying the missing premise:

            1’ All persons have obligations toward their mothers.

To do this risks unsoundness, however, for we might well identify at least a single counterexample (a particularly heinous mother, who actually tried and failed to murder person Y at birth, and has done everything else in her power to undermine person Y’s subsequent life). This singular example, however rare in real life, falsifies the categorical supplied premise. Better I think to treat the original argument as a very strong induction. It becomes cogent if we individuate the first premise (say that I am person Y and my mother is person X), without the need for supplied premises. Of someone who refuses to accept it, we might well say that she simply fails to understand the actual significance of the term ‘mother’ as we use it in normal human discourse.

(2) My second argument for deriving ought-statements from is-statements begins by observing a tacit normativity in the selection of premises from among an indefinite field of potentially salient facts. My reasoning looks like this:

1. There is an effectively unlimited number of facts (articulable states of affairs)
2. A nontrivially large subset of these facts is potentially salient in reasoning toward any substantive conclusion of human concern. (sub-conclusion)
3. A reasoner must bring perceptivity and judgment to bear on the process of selecting among potentially salient facts to employ as premises in an argument of human concern.
4. Perceptivity and judgment are irreducibly laced with normativity.
5. Therefore, all substantive reasoning of human concern has an irreducibly normative component.

One may well object here that I have here conflated epistemological with moral normativity, and thus that my argument turns on an equivocation. I plead not guilty, on the grounds that (at least) when a matter is consequential to humans, the epistemological process by which we determine the salience or relevance of facts is itself imbued with at least some moral normativity. Perhaps developing the Pythagorean Theorem was an epistemically pure process, but as soon as we employ it to design a building, which might (if our calculations are careless) collapse and injure someone, moral significance has entered the room.

I argue, then, that determination of salience is a normative process in the relevant sense, and irreducibly so. A full spelling-out of the criteria by which we select among possible facts plausibly yields robust (if tacit) premises expressing normativity – ought-statements that will not go away, and without which the argument itself is pointless. Of course, if any of the operative terms are of the type in my first example (e.g.: ‘mother’), this is fairly obvious, but I hope I have shown that it is subtly the case even in apparently purely factual examples.

Here is another example of the sort of reasoning Hume finds “altogether inconceivable” (though I am not certain that word means precisely what he thinks it does):

1) This rock is crumbly.
2) (therefore) We ought not to climb on this rock.

This argument might seem to Hume to require some tacit premises (to the effect that crumbly rock can be dangerous, that one ought not take unnecessary risks, etc.). But no one in a context where this would be a normal thing to say would fail to understand the reasoning on offer. There are of course many other potential premises which are evidently not germane – that the sun is 93 million miles away, for example – but quite a few others nearer to hand about which it is not so easy to say. The fact that my mother has an exaggerated fear of injuries from falling, in response perhaps to a childhood trauma, and thus that she will be excessively worried about my taking risks with this crumbly rock, may or may not be a relevant fact in my calculation of what I ought to do in this circumstance. How heavily should I weight the fact of my mother’s fears in this deliberation? The difficulty (though not, I think, impossibility) of answering this question illustrates just the sort of deliberative judgment that imbues almost all apparent pure-fact reasoning with irreducible normativity.

David Braden-Johnson raises a challenge to what I have said so far, on the grounds that Hume’s concern is only with deductive arguments, so my inductive treatment of the argument about motherhood would be beside the point. But Hume himself rejects induction altogether as an instance of reason. He views what we call induction as belief based on past experience which (since reason gives no guarantee that the future will resemble the past) is not sanctioned by reason at all. Perhaps this is a merely antiquarian point, since Hume is most probably mistaken about induction. It does, however, sharpen Hume’s position: on his account, no genuine (i.e.: deductive) reasoning process can get us from purely descriptive premises to prescriptive conclusions. I contend that this may be true in the abstract, but as there are no “purely descriptive” premises “concerning human affairs,” it is a moot and uninteresting point.

I intend my second argument above to show, not that Hume is mistaken about pure deduction, supposing there is such a thing, but that the question is moot in practice. By the definition of deduction, nothing can emerge in the conclusion that was not already contained in the premises, so unless there are at least tacitly normative elements in the premises, a normative conclusion will be invalid. Fair enough. I contend, however, that due to the irreducibly normative process of determining the salience of premises, selected from an indefinitely large pool of potentially salient facts, the requisite normative background already tacitly infects the premises of any interesting argument.

To illustrate by analogy, consider early-modern claims of scientific research to be free of extra-scientific values. Biologists, for example, presented themselves as simply studying the world of life that was before them, answering the questions that nature posed (“carving nature at its joints,” a carnivorous image from Plato’s Phaedrus much beloved by Roger Bacon). Professional training effectively determined, paradigmatically, what sorts of questions and research techniques would receive support and approval within the biology community. Such selectivity is at once inevitable and a principal reason for the success of the discipline as a science, lending it the specialization and focus to make progress. Equally inevitable, however, is the exclusion of legitimate biological questions that lie too far outside the paradigm shared by the community. Consider, for example, Harvard biologist Ruth Hubbard’s remarkable work on the historical/sociological/biological reasons for male-female size difference in modern humans – a question that never came up (it was taken, wrongly as we now know, as a biological given, not as a question in need of explanation) when all prominent professors of biology were men.

My point is not to criticize science – such systematic focus on some questions to the exclusion of others is both necessary and salutary for the process of scientific inquiry – though it is equally healthy and indispensable that it learns to listen to those affected or neglected by its exclusions. I only indicate by way of illustration the serious normative import of biology’s process of selecting methods and questions, and hence of its reasoning about them. Traditional biology’s deductions may have been valid, but they were only notionally free of humanly consequential normative content.

Matt Silliman teaches philosophy at MCLA

The Problem with the Is-Ought Non-Problem
Reply to Silliman
David Kenneth Braden-Johnson
Against (most interpretations of) Hume, Matt Silliman offers two arguments “for deriving ought-statements from is-statements,” or, in the common parlance of the debate, bridging the is-ought gap.  He claims that reasoning in this fashion is “quite natural” and “seldom if ever fallacious in itself.”  In the end, he advises us not “to worry too much about Hume’s problem.”

I think there is much to worry about here, not the least of which is the prospect of falling silent before those who would fallaciously derive ethical justifications from convention, God’s will, popularity, the “nature of things,” and the like.  But my more immediate worry is that Silliman fails to grasp the axiomatic status of “Hume’s problem,” and so unwittingly employs the very position he condemns.  I suggest, therefore, that both of Silliman’s arguments fail, and for the same reason: by insisting on the “tacit normativity” of premises designed to provide support for “any interesting argument,” each assumes (with Hume) the very unbridgeable gap he hopes to traverse or ignore.[1]

In addition, there is a permanent tension at the heart of Silliman’s analysis: it remains unclear whether he intends to offer counterexamples to Hume (that is, “non-fallaciously” bridge the is-ought gap), or merely to eliminate the gap from all or most human discourse.  At any rate, in what follows, I will argue that he fails to accomplish the former and ought not to pursue the latter.

The First Argument

Silliman’s first argument rests on the observation that the premises of “any interesting argument” contain “tacit, normative content” providing an inferential bridge to a “normative conclusion without fallacy.”  He provides this example:

1.  Person X is person Y’s mother.
2. Therefore, person Y has obligations toward person X.

Being a mother, he writes, is a “normatively charged relationship” which typically involves obligations.  The tacit normative content emerges as we parse the term “mother,” and can be made explicit in the following way:

1.  Person X is person Y’s mother.
2. All persons have obligations to their mothers.
3. Therefore, person Y has obligations to person X.

Recognizing the deductive and potentially unsound character of A1’ (given the remote possibility of a mother toward whom no one has obligations), Silliman suggests that we would be on stronger ground to interpret A1 inductively and with reference to a specific person, as in the following version:

1. Y is Matt’s mother.
2. Therefore, Matt has obligations to Y.

But this presents us with a puzzle.  A1’’ remains deductive in tone and (beyond those familiar with the particulars of Matt’s life), logically indistinguishable from A1.  The assumption is, of course, that Matt has an obligation-saturated relationship to his mother, supplying the “tacit normativity” of premise #1 made explicit in a new premise #2:

1. Y is Matt’s mother.
2. Matt has obligations to his mother.
3. Therefore, Matt has obligations to his mother.

But A1’’’ is both deductive and circular, effectively eliminating premise #1, the purported source of the argument’s “tacit normativity.”

Alternatively, rephrasing A1’’’ as an induction, we have:

1. Y is Matt’s mother.
2. Most persons have obligations to their mothers.
3. Therefore, Matt probably has obligations to his mother.

While A1’’’’ avoids overt circularity and deduction, the argument (like A1’’’ before it) fails to deliver the promised goods; that is, provide us with “an argument for deriving ought-statements from is-statements.” (Though I will not pursue the issue further here, the inference fails the test of ethical non-vacuity as well; see footnote 1).  It is, plainly enough, an argument that (fallibly, inductively) derives an ought statement from another ought statement.  Bridging is and ought is, apparently, impossible in the absence of an ought on both sides of the bridge, whether that structure is inductive or deductive.[2]  But now our metaphor has lost all sense, since the bridge was invoked in the first instance as a means of getting from exclusively is-statements to an ought-statement, while no bridge is required to infer ought-claims from a mix of claims containing, tacitly or explicitly, other ought-claims.

The Second Argument

Silliman’s second argument for bridging the gap is related to, but somewhat distinct from, the first in which the tacit normativity of the premise(s) eliminated the need for a bridge (and so validated Hume’s view).  Now Silliman locates the normativity not (merely) in the premises themselves, but in the very selection of premises.  The basic idea is that the “epistemological process by which we determine the salience or relevance of facts is itself imbued with at least some moral normativity.” He proffers on this basis that “all substantive reasoning of human concern has an irreducibly normative component.” 

Unfortunately, this claim is equivocal between (a) seeing the process of selecting premises as essentially normative (in any "interesting" context) and (b) seeing the logical functioning of premises in an argument as essentially normative.  At times Silliman seems to be suggesting both; yet in the context of “non-fallaciously deriving ought-statements from is-statements,” we rightly concern ourselves exclusively with the latter.[3]  Furthermore, the normativity of (a) is logically distinct from (b).  To see this is so, we need only reflect momentarily on the very example Silliman provides of the "epistemically pure" development of the (no doubt interesting and substantive) Pythagorean theorem. The typical normative epistemic imperatives that determine the salience or relevance of facts (be careful, be systematic, avoid contradiction, etc.) that govern argument construction and direct our inquiries have no bearing on the normativity of the argument or its components.  That is, no amount of selectivity, focus, bias, salience, or any other purportedly normative characteristic of the process of argument formation, has any determinative impact on the normativity of the argument.  Indeed, Silliman seems to admit as much as he claims that nomativity associated with the theorem enters the inferential scene only as we apply it, with the proper care of course, when “designing a building." That we generally care about the safety of homes is undeniable, as is the value we place on consistency and deductive certainty, when we claim that, for any right triangle, the square of the hypotenuse is equal to the sum of the squares of the other two sides.[4]  That we ought to be careful when arguing, that the contextual application of arguments in science and elsewhere can be for good or ill, or that we generally care about the content and construction and application of our arguments, are all equally obvious.  

Putting aside these equivocations, have we any reason to suppose that the normativity of argument construction produces or amounts to a normativity of content (reflected in at least one premise) that might serve to bridge the inferential gap between is and ought?  Silliman offers but one brief example:

1. This rock is crumbly.
2. Therefore, we ought not to climb on this rock.

He suggests that, to any clear-headed reader, the choice of premise #1 has been directed by some interested party, and so, despite all appearances to the contrary, is not a “pure fact.”  Its impurity bespeaks its (tacit) normativity.

Note that what Silliman claims is true not of the premise, but of the argument as a whole understood as a kind of cryptic, hypothetical imperative (“If you want to remain safe, you ought not to climb on unsafe rocks!”).  Let us agree that premise #1 either is or is not a “pure fact.”  If it is not -- if it tacitly contains some element of normativity -- then Hume’s is-ought gap remains untouched as we proceed validly to infer an ought from an ought.  Hume's gap is bridged only where the premise set remains exclusively empirical or “pure.”  I will take up below Silliman’s apparent suggestion that we (as in A1 and its progeny above) argue for premise #1’s contextual or holistic impurity in light of the intentions of its author.  Just now, let us assume that it remains a pure, empirical, non-normative fact.  Only in that way might we offer a counterexample to Hume.  The key, as Silliman suggests, is in the selection process.  Was this pure fact chosen for some interested reason?  No doubt it was, as a rational means to secure the intended end – a safe climb.  I reconstruct the fuller argument as follows:

1. This rock is crumbly.
2. Crumbly rocks are unsafe for climbing.
3. We want to be safe.
4. Therefore, we ought not to climb on this rock.

Premise #2 links crumbliness with lack of safety, and #3 introduces our desire to find a safe place to climb.  Note that all three premises remain purely descriptive; in particular, #1 remains a “pure fact,” despite its role in making the conclusion seem attractive.  So, while no tacit normativity infects the premise set, the explicit normativity of the conclusion seems to gather some kind of support from the purely empirical premises.  Have we, in A2’, finally discovered a viable is-ought bridge?  Not at all.  As it stands, the inference is clearly invalid, since none of the premises assert or (individually or collectively) entail “one ought not to climb on the rock.”  The normative force of the conclusion is parasitic, not on any kind of “tacit normativity” of the premises (there is none), but on the unstated rationality of choosing an effective means to one’s desired ends (something that is missing from, and the source of the invalidity of, A2’).  That is, this is not an instance of moral reasoning at all, but an example of prudential reasoning from a hypothetical desired end to an effective means of securing that end: desiring, or valuing X, and seeing Y as the best or only means of acquiring X, I “ought” to do Y.  If you want to be healthy, you should eat your vegetables.  If you want a safe climb, and if you ought to do whatever keeps you safe, then you should stay off those rocks: if both A and B then C. The conclusion has binding normative content only for those who accept the italicized antecedent of the conditional claim. And, once we do, we commit ourselves to an explicitly normative premise.

Once again, Silliman has not been able to forge a successful inferential bridge from is to ought.  A command or bit of prudential reasoning is not an argument.  A2, therefore, is either a hypothetical imperative (“you ought to do Y if you want X”) dressed up to look like an inference, or an inference with an explicitly normative premise.  I began with the suggestion that Silliman’s arguments vacillate between bridging and denying the importance of Hume’s gap, and that neither approach held out much promise.  I hope to have shown that A1 and A2 and their extensions either assume, along with Hume, an unbridgeable gap or substitute commands, explanations, contextual applications, or a notion of means-end rationality for arguments.  Given the axiological nature of the is-ought gap when applied to logical derivations, Hume’s problem is our problem, too.

Thanks to Nicole K. Braden-Johnson for her many helpful comments and suggestions.


Guevara, D. (2008). “Rebutting Formally Valid Counterexamples to the Humean “Is-Ought” Dictum.” Synthese, 164, 1.
Johnson, D. K. (1995). “The Ought-Is Question: Discovering   the Modern in Postmodernism.” Inquiry: Critical Thinking Across the Disciplines, 14, 4.
Peirce, C. (1998). The Essential Peirce: Selected Philosophical Writings, Volume 2. Peirce Edition Project. Bloomington: Indiana University Press.
Plutynsky, Anya. (2011). “Four Problems of Abduction: A Brief History.” HOPOS: The Journal of the International Society for the History of Philosophy of Science, 1.
Prior, A. N. (1960). “The Autonomy of Ethics.” Australasian   Journal of Philosophy, 38.

David Kenneth Braden-Johnson teaches philosophy at MCLA

Must Ought Imply Can?
A Response to Vranas
Gerol Petruzella

Is morality too demanding - "requiring us to do things that we literally cannot do, things that go beyond our abilities"? Peter Vranas finds this claim "hard to swallow" (Vranas 197) and so defends a version of the ought-implies-can principle, OIC, against several objections, including arguments based upon putative counterexamples, as well as conceptual arguments.

Vranas does excellent work in building definitional clarity into his treatment of the issue. He articulates OIC as synchronic and time-indexed, since its elements (obligations, abilities, opportunities) are, an analysis which seems right to me. He offers a clear deductive argument supporting OIC:

(P1) Obligations "correspond" to reasons for action: If an agent has an obligation to φ, then the agent has a reason to φ.
(P2) Reasons for action "correspond" to potential actions: If an agent has a reason to φ, then φ-ing is a potential action of the agent.
(P3) Potential actions "correspond" to ability plus opportunity: If φ-ing is a potential action of an agent, then the agent can φ.


(OIC) Obligations "correspond" to ability plus opportunity: If an agent (S at a given time t) has an (objective, pro tanto) obligation to φ, then the agent (at that time) can (i.e., has both the ability and the opportunity to) φ.

Vranas also does much good work in "clearing the field", deftly undercutting several purported grounds for attacking OIC - legal arguments (199), psychological arguments about obligations to feel (174), arguments based upon natural-language confusions of the sense of "ought" (178), exotic hypotheticals involving moral luck (179), and arguments which fail to realize that the grounds for rejecting OIC are also grounds for rejecting the ought-implies-logically-possible principle (187).

For all his rigor in defining and addressing the problem, however, Vranas does not apply the same standard of linguistic precision to a crucial relation in his core argument - "corresponding to". In his primary expression of the argument (171-172), he appears to define "correspondence" in terms of deductive entailment (as expressed in if-then statements). [Compare Plato's difficulty with metechein, to partake of.] I believe this leads Vranas into difficulties which his article leaves unaddressed, and which leave the door open for continued critique of OIC.

After identifying Vranas's argumentative flaw, I will offer an argument against OIC which relies on a simple, real-world counterexample, rather than exotic thought experiment scenarios involving brainwave monitors, morally confused Nazis, superhuman typists, or strangling canaries.

Vranas's Error

By implicitly defining "corresponding to" as logical entailment, Vranas necessarily commits himself to the deductive consequences of his premises as stated. Unfortunately, this commitment makes his accounts of potential, reasons for action, and ability too restrictive to account for certain real-world counterexamples.

Specifically, I want to take a closer look at Vranas' (P2). Vranas himself recognizes this premise as a likely locus of contention (173), but believes that he has mustered sufficient argumentative merit to defend it. I disagree.

First, let's express (P2) formally for purposes of clarity. The premise

(P2) Reasons for action "correspond" to potential actions: If an agent has a reason to φ, then φ-ing is a potential action of the agent

can be simplified to

If R, then P


All R are P.

Of course, then, the logically equivalent forms are equally true:

Contrapositive: All non-P are non-R.
Obverse: No R are non-P.

Some Moral Intuitions

Now, let's consider two intuitions regarding moral agency and human potential. A common, though unstated, intuition may be that

(G1) The scope of human action is no greater than the scope of human potential (or, All A are P).

This seems plausible. Surely, it would be nonsensical to claim that human acts could somehow extend beyond humans' own capacities.

We also presume that

(G2) The scope of moral responsibility extends to at least some consequences of human action (or, Some C are R).

That is, we can properly ascribe moral responsibility to not only the immediate acts which human agents perform, but also (in at least some cases) to the consequences of those acts, particularly when the agent acts intentionally and with foreknowledge of those consequences.

Now, what is interesting about (G1) is that it is false, or at least in need of clarification. Human agents are in many cases capable of engineering causal chains leveraging ordinary human capacities or potentials to cause "superhuman" effects. For example, I have the (ordinary human) capacity to break a small glass vial by dropping it to the pavement on a New York City sidewalk; depending upon the contents of that vial, my action (which, again, itself falls well within the scope of ordinary human potential) has the direct effect of killing billions of humans, an effect which far exceeds ordinary human potential, but for which nevertheless I am morally responsible.

(G2) prompts us to realize that it is not just that I am morally responsible (and blameworthy) for the act of dropping the vial; I am also morally responsible for the consequent deaths of billions, even though it is not within my human potential to kill billions. We live in a world where humans in the aggregate can act to cause tremendous, morally significant effects; but given our technologies, such possibilities arise not just in the aggregate, but for the individual agent. The push of a button can launch a nuclear firestorm; a computer keystroke can disable communications networks for entire nations; and so on. While human potential is limited, we have the creative capacity to place ourselves into causal relations with effects which are effectively unlimited. At the very least, any adequate account of moral responsibility must be prepared to account somehow for human moral responsibility for effects which surpass human potential actions, particularly when those actions are taken intentionally and with foreknowledge of the consequences.

Let's now compare these intuitions with Vranas's (P2) and its logical consequences. Again, (P2) states that "if an agent has a reason to φ, then φ-ing is a potential action of the agent," which we have expressed formally as "All R are P." If we unpack the contrapositive of this statement into natural language similar to Vranas's, we see that Vranas is necessarily also committed to the claim that

(P2-C) "If φ-ing is not a potential action of an agent, then the agent does not have a reason to φ."

How does this stack up with our earlier discussion? It seems that (P2-C) presumes some version of (G1), that is, that there is some necessary conceptual alignment between the scope of potential action and the scope of human choice-making.

Vranas's Responses to Agents Becoming Incapacitated

In 4.2.2 (179), Vranas addresses the scenario in which a man makes himself unable to fulfill a moral obligation to be in Boston at 9am to get married by boarding an airplane elsewhere at 8:30am. To maintain his claim that OIC holds true, Vranas claims that while it is "natural" to say that the man is blameworthy for failing to show up, strictly speaking, the man is not blameworthy for violating his obligation to show up, but rather for violating "the obligation to avoid doing anything that would make him fail to show up". This is a curiously convoluted and ad hoc explanation of the involved obligations.

In a similar vein, in 4.2.4 (182), Vranas bites the bullet and accepts that OIC does entail that people can sometimes get rid of unwanted obligations, with the "compensatory" understanding that, to make it more palatable, there are "residual" obligations that (necessarily?) arise. Such residual obligations are such as to "make up for" the failure to meet the earlier obligation, but within the newly-constrained capacities of the agent - being obliged to apologize for having not shown up for the wedding.

In general, Vranas's attempt in 4.2 to account for situations in which an agent becomes unable to fulfill an obligation falls short of completeness, as he acknowledges:

"Admittedly this does not establish that no cases exist in which the obligation persists after the inability sets in, but it does shift the burden of proof to those who would insist that such cases exist." (178)

I am happy to take up that burden. In the following section, I describe a realistic case in which a moral obligation persists after the agent's inability sets in. I argue, therefore, that since such a case does exist, it is an effective counterexample to OIC.

The Drunken Pilot Scenario

Aero Flight 311, also known as the Kvevlax disaster, was a crash in 1961 which resulted in the deaths of 25 people on board a domestic passenger flight in Finland (Wikipedia). Investigation found that the airplane had been airworthy, but autopsies revealed that the captain, Lars Hattinen, had a blood alcohol content of 0.20, while his copilot Paavo Halme had a blood alcohol content of 0.156. It is reasonable to ascribe the crash, and deaths, to the pilot's and co-pilot's intoxication.

Let us imagine a similar case - well within the realm of realistic possibility[5] - which allows us to sharpen the focus of certain relevant features to consider the moral implications of our current discussion.

Cathay Pacific Airlines schedules a 15-hour nonstop flight between San Francisco and Hong Kong. The vast majority of the flight path is over the open waters of the Pacific Ocean. The pilot and copilot begin the flight as fully qualified and competent moral agents, taking off from San Francisco with a full passenger manifest.

At time T0, the beginning of the flight, the pilot and copilot both have a moral obligation to fly the airplane safely, and they both have the ability to fulfill that obligation. However, at time T1, about 6 hours into the flight, both the pilot and the copilot decide to drink large quantities of alcoholic beverages, such that they are both physiologically impaired. The blood alcohol content of both individuals exceeds 0.15; the CDC describes the predictable effects of this level of BAC as "substantial impairment in vehicle control, attention to driving task, and in necessary visual and auditory information processing" ( Let us stipulate that the physiological effects of the alcohol on both pilot and copilot are the predictable effects.

By time T2, about 7 hours into the flight, we stipulate that the pilot and copilot have both become physiologically unable to fulfill the obligation to fly the airplane safely. The pilot remains conscious but deeply impaired; the copilot passes out. We further stipulate that no other individual on board the airplane has any greater ability to fly the airplane safely. We can bring this about in any number of ways: perhaps the passengers are all minors, or physically or mentally disabled; or perhaps in their drunken state the pilot and copilot have locked themselves in the cockpit and broken the lock mechanism. Feel free to fill in whatever details you like. The plane carries sufficient fuel to complete the flight under normal circumstances; however, there is not sufficient fuel to allow the pilot to stay in the air until he can sober up.

The Key Question

At time T2, does the pilot have a (current) moral obligation to fly the airplane safely?  Vranas would say no, since ex hypothesi he no longer has the ability to do so. He might say that he has some "residual" obligation instead. But in this situation, what possible residual obligation is available to perform? The ability to land the plane safely is the same ability as that necessary to fly the plane safely, and so the pilot cannot "fall back" to an obligation to perform an emergency landing. We have stipulated ex hypothesi that no other available moral agent can fly the plane any more safely, and so he cannot have the obligation to "hand over" the controls to a more capable individual. To simply step out of the cockpit and leave the plane on autopilot will, if anything, be more likely to cause harm, and so this course of action cannot be a residual moral obligation – unlike a drunk driver, the pilot cannot just “stop driving.” Perhaps Vranas would maintain that the only remaining moral obligation the pilot owes to the passengers is to offer a heartfelt apology.

Perhaps, following his strategy in dealing with the wedding scenario, Vranas might claim that while it is "natural" to say that the pilot is blameworthy for failing to fly safely, strictly speaking, he is not blameworthy for violating his (past) obligation to fly safely, but rather for violating "the obligation to avoid doing anything that would make him fail to fly safely". So perhaps, at T2, the pilot is morally culpable for having failed to avoid making himself incapable, but is not culpable for failing to fly safely. How realistic is this workaround?

Let’s assume that Vranas is correct, and that at time T2, the pilot does not have a current moral obligation to fly the airplane safely, because he does not have the ability to do so. We have stipulated that no other moral agent has the ability to fly the airplane safely. Therefore, at time T2, no one has a moral obligation to fly the airplane safely. This seems prima facie strange; but I would like to figure out why, exactly.

One source of strangeness is perhaps that the moral weight of obligations is mismatched between the original obligation (which is to preserve many people’s safety and keep them from dying) and any possible “compensatory” obligation which one might ascribe at T2 (which is, at most, to offer regret). Remember, the pilot chose knowingly to incapacitate himself from performing his obligation; no “compensatory” obligation available to him at T2 seems comparable in seriousness. He seems to have “traded down," releasing himself from a serious moral obligation in exchange for assuming a less onerous moral obligation.

A second source of strangeness is that the morally relevant situation is ongoing – not past, and not time-limited; the situation persists, with an odd emptiness where the relevant moral agency used to be. In most ordinary circumstances of moral agency, when agent A is responsible for act R, our standard assumption is that A’s agency is active so long as R is ongoing, and so long as R is in effect, A’s responsibility is also in effect. (If I am responsible for stealing from my grandmother’s bank account, so long as the theft is ongoing, my moral responsibility continues.) But even when a situation continues in the sudden absence of the original morally responsible agent, we presume both that the moral agent has ceased to exist, and the availability of some moral agent to pick up the mantle of moral responsibility: if a professor suffers a heart attack and cannot continue teaching a course, we presume that the college bears the responsibility to the students to get the course taught. The current scenario blocks both of these avenues. We are not accustomed to a morally significant circumstance persisting beyond the agency of the agent who bears moral responsibility for it.

Note to Readers

Thesis XII: A Philosophical Review is published biannually as an open forum promoting respectful philosophical exchanges among students, faculty, alumni, and the public.  Submissions reflect a diversity of disciplinary perspectives, philosophical approaches, and topics.  Those new to the discipline are especially encouraged to participate.

Address all correspondence to:

Dr. David K. Braden-Johnson, Editor
Thesis XII: A Philosophical Review
Department of Philosophy, IDS, and Modern Languages
Massachusetts College of Liberal Arts
North Adams, Massachusetts  01247 

Telephone: (413) 662-5448.

Associate Editor: Dr. Matthew R. Silliman

A third source of strangeness is that a persistent and ongoing process of huge moral consequence (life or death of individuals) can be morally ‘driverless’. The pilot brought about the morally significant consequence through his action; but he cannot take any action which can undo the consequence. In such a case, OIC absolves him of the moral responsibility, and at most, attempts to backfill our moral intuitions with talk of “compensatory” obligations to account for our intuitions of the pilot’s ongoing blameworthiness. But this still leaves the core, relevant moral consequence ‘driverless’.

Humans can take actions which have consequences which are irreversible to us. This is not a profound revelation; but we have perhaps not fully appreciated its implications. The key relevant implication: the consequences of our actions often escape the boundaries of our capabilities. And so if our moral responsibilities are capped by the boundaries of our capabilities, as OIC articulates, we must bite the bullet and acknowledge that, contrary to our G2 intuition above, our moral responsibilities do not, in fact, always extend to the consequences of our actions; and that we can “game” the system of moral responsibility itself, by causing situations for which we know we cannot properly be held morally responsible (hence blameworthy); situations in which we will bear some lesser, “compensatory” moral obligation, within our merely human capacities to fulfil.

Many thanks to my colleagues David Braden-Johnson, Paul Nnodim, and Matthew Silliman for their critical feedback in this ongoing conversation.


Vranas, Peter B. M. "I Ought, Therefore I Can." Philosophical Studies 136.2 (2007): 167-216. Web.
Wikipedia contributors. "Aero Flight 311." Wikipedia, The Free Encyclopedia. 25 Oct. 2015. Web. 5 May. 2016. 

Gerol Petruzella is the Assistant Director of Academic Technology and teaches philosophy at MCLA

Utilitarian Dilemmas in the Literature of Scapegoats
Paul Nnodim
The scapegoat problem
I’m about to take issue with those critics of classical utilitarianism who employ extreme scapegoat scenarios to demonstrate inherent weaknesses in the system. Since I’m one of those critics, it appears I might be, at least tangentially, arguing against myself. But how does one argue against oneself without intentionally weakening one’s proffered claims? Obviously, in many instances, we do not freely choose our beliefs, but are socially conditioned to accept them. Our personal beliefs can be predicated on prejudice, sweeping generalizations, oversimplification, false dilemmas, etc. Thus, it is hardly surprising to see many of us struggle to grasp the idea behind raising hypothetical objections to our considered judgements of what is true, plausible, or morally acceptable. But this is what reflective thinking is all about. However, my goal in this paper is not to celebrate the power of self-criticism or self-denial, but genuinely to take a closer look at the utilitarian enterprise and to see if there are compelling reasons for a rethink.

The term scapegoat possibly has its origin in the Hebrew word -עֲזָאזֵל (Azazel) found in the Pentateuch. The original meaning of Azazel is hard to figure out, but I’m willing to accept, if only momentarily, William Tyndale’s 1530 translation of the Hebrew Bible. There, he seems to break the term into ez (the goat) and azel (that escapes), hence, the [e]scapegoat: “And Aaron shall bring the goat upon which the Lord’s lot fell and offer him for a sin offering. But the goat on which the lot fell to escape, he shall set alive before the Lord to reconcile with and to let him go free into the wilderness.” [6] The scapegoat then was a goat upon whose head the priest symbolically placed the sins of the people before letting the animal escape into the wilderness. This ritual was part of the Hebrew ceremony of Yom Kippur or the Day of Atonement.[7]
Critics of classical utilitarianism often find unconventional allies among fiction and nonfiction “scapegoat” writers. As wide-ranging as these storytellers may be in style and scope, they share one thing in common: Their stories are disconcerting, gruesome, and to the joy of their philosophical appropriators, anti-utilitarian.

Classical utilitarianism

Classical utilitarianism is the principle of “the greatest happiness or good for the greatest number.” Thus, it is an aggregate-maximizing, hedonistic, and consequentialist ethics, since it teaches that the end of human conduct is happiness and the discriminating norms for distinguishing a right behavior from a wrong one are pleasure and pain. Jeremy Bentham (1748 – 1832) was perhaps the earliest scholar to articulate the principles of the doctrine in a more systematic and coherent manner. Utilitarianism as a form of ethics does capture the imagination of rational individuals. It only says we all ought to be happy by seeking and maximizing the pleasurable while minimizing the un-pleasurable or the painful. According to Rawls, “It is natural to think that rationality is maximizing something and in morals it must be the good. Indeed, it is tempting to suppose that it is self-evident that things should be arranged so as to lead to the most good.” [8] Critics of the system, however, argue that utilitarianism could permit gross inequalities if such acts produce the best consequences for the majority of the people. Some of these detractors cite fictional works, such as Ursula Le Guin’s “The Ones Who Walked Away from Omelas,” Edgar Allan Poe’s The Narrative of Arthur Gordon Pym of Nantucket or a real-life story, such as the Mignonette,[9] to illustrate the utilitarian jeopardy. Here are synopses of the stories.


In Le Guin’s story, for example, Omelas (Salem, Oregon written backwards) is an exhilarating, quixotic, and idyllic utopian city: a place distinguished by its grandeur, magnificent buildings, and fairytale-like summer festivities. At the first glance, everything seems to work perfectly well in this city. The inhabitants are happy-go-lucky, wealthy, and affable. However, a closer look at Omelas reveals an odious incongruity. In the cellar of one of its buildings lives a malnourished, neglected, and grief-stricken child. This child used to scream for help, but now only whines, her voice apparently muffled by fatigue and despair. The child sleeps on her excrement and the sores all over her body have become septic. The cellar where they keep the unfortunate child is as malodorous as a fetid pile of garbage. People know about the existence of the woebegone child, but no one dares to help her; otherwise, as they believe, their good fortunes go with the wind. Sometimes, a group of Omelas residents would visit the gruesome cellar to inform themselves about the child’s fate. Initially, they would feel some disgust. But as time passes by, they adopt evasive stances to dampen any feelings of antipathy or any roiling of conscience. If we release the child from the cellar, they reason, she is already too degraded to know any real joy. She has been afraid for too long and would not rid herself of fear. Other times, a few, sensible residents would visit the child and choose not to return home. They would walk through the main street of the city, pass the city gate, and head silently to some unknown place. “Each one goes alone, youth or girl, man or woman. …They leave Omelas, they walk ahead into the darkness, and they do not come back.”[10] Since classical utilitarianism seeks to maximize the aggregate utility (pleasure) of the largest number of people, it would seem morally permissible to abuse the rights of one innocent child, the scapegoat, for the greater happiness of the many.

A Tale of Two Richards

In Edgar Allan Poe’s 1838 novel, The Narrative of Arthur Gordon Pym of Nantucket, the crew of a ship called Grampus find themselves cast away and starving to death. In order to survive, they would have to kill one of their members, eat his flesh and drink his blood. They cast a lot and it falls on Richard Parker, a former mutineer. They quickly slaughter him and the rest of the crew feeds on his body until rescue comes.

In what appears to be one of the most bizarre coincidences in literature, 46 years later and more than three decades after the death of Poe, a cabin boy, also named Richard Parker, became a victim of castaway cannibalism. But this time, the story is real. In 1884, an English ship called the Mignonette sank in the South Atlantic, possibly en route to Australia. Its four Englishman crew survived treacherous days and nights at sea on a lifeboat. Within a couple of days, their meager rations ran out and they were starving to death. Now, the cabin boy, Richard Parker who was an orphan had fallen ill. The other crewmen conspired against poor Richard, murdered him, ate his flesh, and drank his blood. As they crew was anticipating whom next to kill for food, a passing boat found and rescued them. Upon their return to England, the men went on trial for murder. Two of them served minor sentences, while one even walked away free.[11] As Sandel notes, the strongest argument for the defense was a utilitarian one: “… given the dire circumstances, it was necessary to kill one person in order to save three. Had no one been killed and eaten, all four would likely have died. Parker weakened and ill, was the logical candidate, since he would soon have died anyway. … he had no dependents. His death deprived no one of support and left no grieving wife or children.”[12]


The central idea of these stories is founded on a certain supposition. If we were offered a world in which the majority is kept enduringly happy on the condition that they deny some innocent citizens their inalienable rights, would it not be an offensive conduct? Critics of classical utilitarianism draw on such extreme suppositions to chide the system. But wouldn’t it be clear to most rational and moral agents, including utilitarians themselves, that torturing an innocent child or eating a sick cabin boy to maximize the aggregate pleasure of other people is a bad idea? As a normative ethical principle, classical utilitarianism may not defend itself successfully against such critics, since their goal is to show in whatever way possible that the normative standpoint of utilitarianism is unsound. However, if we were to see classical utilitarianism as an applied ethical practice, then marginal cases would constitute weak analogies against it. While the main thrust of normative ethics is to explore, in general, what moral lives we ought to live, applied ethics concerns itself with the particular instantiations of those “oughts.” We, the critics, tend to ignore this distinction.

(1)  Stolnitz, Jerome. "On the Origins of "Aesthetic Disinterestedness"" The Journal of Aesthetics and Art Criticism 20.2 (1961): 131-43. JSTOR. Web. 12 September, 2013.
(2) Zangwill, Nick. "Aesthetic Judgment." Stanford Encyclopedia of Philosophy. Stanford University, 28 Feb. 2003. Web. 12 September, 2013.
(3) Slater, Barry H. "Internet Encyclopedia of Philosophy." Aesthetics. N.p., 25 July 2005. Web. 12 September, 2013.

Paul Nnodim teaches philosophy at MCLA

Reasoning and Intelligence
Brett Belcastro
Sentience and other measurements of sufficient intellectual complexity are the essential foundations for a criteria of moral status (1). This challenges ethicists to define the observable empirical conditions of sentience, the most important of which is a demonstrable intelligence. But the identification of intelligence relies on the distinction between entities which are capable of reasoning and those that are not, and if reasoning is synonymous with logic, then the distinction becomes impossible and the project of sentience fails before it can even begin.

Traditional definitions of logic often conflate it with reasoning and inference, as when Patrick Hurley defines logic as “the organized body of knowledge, or science, that evaluates arguments.” (2) If an inference represents the movement from premises to conclusion, then it appears possible to provide a comprehensive definition of argumentation without reference to reasoning at all.

Is it necessary that an intelligence is also a reasoning agent? A pocket calculator follows a system of logic, arranges electrical signals to model a set of beliefs, and makes inferences to produce the output of its equations; what is to stop an observer from saying that the calculator is properly intelligent, and on those grounds to consider it for moral status?

Logicians such as Terence Parsons describe “reasoning structures,” (3) spaces within which inferences occur. This structure follows the evolution of inferences from one to the next like the points that form a geometric line. For Parsons, reasoning is the series of relationships which inferences have to each other, and at all times a reasoning agent is located at some point on the line of inference. This is similar to Gilbert Harman's assertion that reasoning is “the process of modifying antecedent beliefs and intentions.” (4)  A pocket calculator, though capable of using inferences to produce output, is not capable of cognizing its inferences, only of repeating them: 2 + 2 = 4 + 2 = 6 + 2 = 8 …

An implication of Parsons' criteria is that a more sophisticated calculator could occupy the line of inference. The distinction is between a calculator which only accepts input and one which possesses a project, which recognizes its inferences as steps in a wider sequence towards the achievement of a valuable end. A computer which is capable of programming itself (5), and thereby demonstrating its agency and project, closes the gap between itself and entities with moral status. If there is no better reason to deny computers normative value than that they are inorganic, then the question of personhood obtains to digital constructs. Parsons' and Harman's definition of reasoning therefore has a direct implication on the conditions of sentience and opens the possibility of an ethics which must take an artificial intelligence seriously.


(1) Silliman, Matthew R. Sentience and Sensibility. Las Vegas: Parmenides, 2006.
(2) Hurley, Patrick J. A Concise Introduction to Logic. Belmont: Wadsworth/Thomson Learning, 2006.
(3) Parsons, Terence. “What Is an Argument?” The Journal of Philosophy 1996: 93.4 164-185.
(4) Harman, Gilbert. “Practical Reasoning.” The Review of  Metaphysics 1976: 29.3 431-463.
(5) Dreyfus, Hubert L. What Computers Still Can't Do. Cambridge: MIT Press, 1994.

Brett Belcastro is a student at MCLA

On the Formulation of Successful Ad Baculum Arguments
Matt Luz
Hurley explains that ad baculum arguments “occur whenever an arguer poses a conclusion to another person and tells that person implicitly or explicitly that some harm will come to them if he or she does not accept the conclusion.”[13] I agree with most of Hurley’s definition, except that the arguer does not state a conclusion. Instead of stating a conclusion, the arguer uses a threat to coerce the audience into engaging in the behavior suggested by the arguer.
The audiences’ response to the threat is based on various factors, which include: (i) the arguer’s use of insufficient evidence, (ii) the credibility of the threat, (iii) the proximity of danger, and (iv) the ability to easily avoid harm. The arguer’s goal through the use of ad baculum arguments is to convince the audience to engage in the behavior suggested by the arguer. I argue that these four components are necessary conditions for structuring a successful ad baculum argument.
Insufficient evidence provides a claim, which implements an emotional appeal, and is constructed by the arguer in order to manipulate the audiences’ behavior. The arguer uses insufficient evidence when they cannot provide a logical reason as to why the audience should comply with what they want, and so the arguer will use insufficient evidence to convince the audience not to behave as the arguer wants them to.[14] In other words, insufficient evidence is used to support poor reasoning. In ad baculum arguments, the insufficient evidence appeals to fear. If the arguer is to say, “If you tell the boss I missed work yesterday, I will lose my job. Furthermore, my family will go hungry,” it would qualify as an appeal to pity.[15] This is another form of insufficient evidence because it involves appealing to emotions. Instead of appealing to fear, it appeals to guilt.[16]

If ad baculum arguments are to be successful, they must appeal emotionally to the audience by delivering a threat. For threats to be effective, they must identify with psychological discomfort and an unwanted outcome against the interest of the audience. The threat presents the audience with a choice, to either engage or not engage in the behavior the arguer wants them to. This decision to comply with the arguer’s request is based on the credibility of the threat. Force is a necessary component in delivering an effective and appealing threat; its proper use makes some threats stronger than others. It also shows to be more compelling when the insufficient evidence provided in the case is strong. This provides the audience with a stronger inclination toward the likelihood of the unwanted consequence actually occurring.

The American Psychological Association defines fear as “a rational reaction to an objectively identified external danger that may induce a person to flee or attack in self-defense.”[17] Based on this definition, I would like to address the concept of danger with the same concepts in this definition, for one fears the possibility of danger. Danger necessarily involves perceived outcomes based on rational behavioral-reactions, proximity to the audience, and the degree to which one would engage in self-defensive strategies to avoid it.

The presence of danger poses a threat to self-sustainment and/or some part to existence, or existence as a whole. Danger is the threat operating on the threat and they are dependent on perception and emotional reactions. Pfau references Aristotle’s claim that “emotions are permeated by reason” in his essay.[18]  If one experiences an event where danger is proximal, chances are that they would not wish to experience that same event again, for when one faces the possibility of danger, one perceives it as a threat and experiences psychological discomfort. When talking about danger and its influence on the audience, it is important to identify what I am talking about. I refer to this object which presents danger as the fear-object. Fear-objects introduce danger to the audience. These fear-object are anything (events, persons, etc.) which the audience perceives to be dangerous. When the audience is confronted with danger, one tailors their behaviors in order to avoid presupposed consequences which the danger underscores.

Harm is the permanent loss of some aspect of one’s existence, whether it is bodily functioning or the continuation of a positive experience significant to one’s existence. From this, harm appears to always, in every case, allude to an unwanted consequence. Harm is something I believe one will tailor any actions towards eliminating the thought of, and in doing so is behaving rationally. “Witte’s research indicated that fear appeals can be effective, but only under two conditions: (1) the threat must be credible, so that the respondent takes it as a real danger to him or her, and (2) the action recommended to deter the threat must be perceived by the respondent as feasible and easy to carry out.”[19]

When the audience sees that the fear is both relevant to them and also harm is easy to avoid, they are more likely to comply with the arguer. In this moment, the arguer is either victorious or fails in convincing the audience of their argument. If the ad baculum argument is successful, the audience will comply with the arguer.

Matt Luz is a student at MCLA

J. Stanley Yake
The hills pass by in radiant dress.
The Indian Summer sun
Bejewels the softened contours of these friends.
Their scattered oranges
Set the darkened greens to rest.
Their yellow greens lay open doors;
They beckon as we past.

We stop a while for track repairs.
A moving freight
For minutes draws a horizontal line
And sets afloat those colored invitations.
The gentle shoreline soon returns
And nature’s softness quickens as we move.

But West Point squares the shoreline meant for trees,
For gentle lines and natural grace.
It’s flat, wide walls and perfect angles
Drive the concrete down and up
And back and forth on every side.
No graceful rooting here;
No going where the goings good
For self-fulfillment.
No luck personal growth
From soil enriched by fruits of former years,
By moving toward the sun, or toward the water,
Or toward the other good we need to save our lives.
No. Here the concrete buttresses against attack
Stake out the boundaries of the world:
The place the flag’s unfurled.
These are not the Palisades,
The natural comfort of the river guest.
These are not the stolen parks
That structure evening trysts
For timid wanderers.

This is the call to arms!
The definition of the enemy
Just outside the walls.
This is protection from the man-made!
But who or what protects from the protection?
These golden, gorgeous hills

Invite the strangers in,
Provide some warmth to passers by
And sustenance to rooted trees.
What warmth could come from definitions?
What strength when social roots are sheared
By concrete slabs,
When flags exclude and push away?

The Indian Summer sun
Bejewels our social roots
With colored warmth and synergistic life.
It covers o’er the reckless growth of shoots,
Protects from definition’s strife.

I’ll take these shaded, green and golden hills,
These Palisades, this flowing river’s open doors.
I’ll take the invitation every time.
I’ll take the splendor in the grass,
The Wordsworth joy at nature’s call.
I’ll move with strangers through the pass,
And live with them outside the wall.

[1] I take the issue to be that we cannot derive a non-vacuous, action-directing normative conclusion from any set of purely descriptive (that is, non-moral, non-normative) claims.  The qualification “non-vacuous” follows from the many successful efforts, beginning with Prior (1960) validly to provide clever (yet apparently vacuous, non-instructive, or non-action-guiding counterexamples to Hume (see Guevara, 2008).  For example:
1. Either the moon is made of cheese or you ought to respect your mother.
2. The moon is not made of cheese.
3. Therefore, you ought to respect your mother.
The argument is formally valid (Either A or B/not-A//B), and appears to derive an ought-statement from a set of is-statements; yet the argument offers no compelling basis for accepting the normative conclusion beyond the insistent formal mechanisms of logic.  Furthermore, we may freely substitute any (repugnant or particularistic) ought-statement for B without loss of validity:
1. Either the moon is made of cheese or you ought to insult your mother.
2. The moon is not made of cheese.
3. Therefore, you ought to insult your mother.
[2] In an earlier paper (1995) which Silliman references, I argue that the is-ought gap is clearest when attempting to get from is to ought in deductive contexts (a process I called “get-D”), while suggesting that one might “get-ND” (“get in some non-deductive sense”) from one to the other (in an explanation, for instance).  I never mention induction in that earlier work, but assume that abduction (which only some take to be a species of inductive inference) may at times appear to bridge the gap, where, for instance, actually having obligations to one’s mother best explains some other observed phenomenon.  But in these cases, the bridging occurs not on the basis of some tacit normativity in the premises, but in the process of clarifying an otherwise unexplained observation. Peirce first described abduction (later dubbed “retroduction”) as involving “some very curious circumstance, which would be explained by the supposition that it was the case of a certain general rule, and thereupon adopt that supposition” (1998, p. 189).  Here is his attempted formalization of abduction:
1. The surprising fact, C, is observed.
2. But, if H were true, C would be a matter of course,
3. hence. There is reason to suspect that H is true. (p. 231). 
Most commentators agree that, in affirming the consequent, abduction resists successful formalization.  I’ll leave it to the reader to solve these and other riddles of abduction and to apply the results to Silliman’s examples. (See Plutynsky for a list of worries attending the formalization and justification of abduction.)
[3] Silliman at least momentarily appears to agree: “By the definition of deduction, nothing can emerge in the conclusion that was not already contained in the premises, so unless there are at least tacitly normative elements in the premises, a normative conclusion will be invalid.”  Serving as a clear example of the “permanent tension” I note above, he prefaces this Humean insight with this contradictory assertion: “on [Hume’s] account, no genuine (i.e., deductive) reasoning process can get us from purely descriptive premises to prescriptive conclusions. I argue to the contrary, even in the case of deduction, at least in real-world reasoning….” (emphases mine). Notice that the final proviso regarding “real world reasoning,” in underscoring either the tacit normativity of premises or the processes of argument construction which, in his words, “imbues” reasoning with irreducible normativity, does not violate but actually presupposes Hume’s position on the is-ought gap.
[4] And not, as the newly embrained yet still incompetent scarecrow in The Wizard of Oz would have it: "The sum of the square roots of any two sides of an isosceles triangle is equal to the square root of the remaining side."
[5] See, e.g. the case of David Hans Arntson, charged with flying two Alaska Airlines flights under the influence of alcohol in 2014 (; or the as-yet unidentified American Airlines pilot stopped on the tarmac from flying under the influence of alcohol this year after failing a Breathalyzer test (
[6] Leviticus, Ch.16, Tyndale’s Old Testament, 1992, 172. Yale University Press; Modern Spelling ed edition.
[8] John Rawls, A Theory of Justice, Oxford: Oxford University Press 1971, 24-25.
[9] Both stories also can be found in Michael Sandel’s Justice: What’s The Right Thing  To Do? New York: Farrar, Straus and Giroux, 2009, 31-35, 40-41. See also –The Illustrated London News, September 20, 1884
[10] Ursula K. LeGuin, The Wind’s Twelve Quarters, New York: Perennial, 2004, 283-284.
[11] See Michael  Sandel 2009, 31-33, as well as Edgar Allan Poe, The Narrative of Arthur Gordon Pym of Nantucket Wiley and Putnam (1838), and the following link:
[12] Michael Sandel’s Justice: What’s The Right Thing  To Do? New York: Farrar, Straus and Giroux, 2009, 32.
[13] Hurley, P. J. Concise Introduction to Logic. 9th Edition. Cengage Learning. Print. 113.
[14] Childress, J. F. “Appeals to Conscience.” Ethics 89.4 (1979).
[15] See note 1: 114.
[16] See note 1: 114.
[17] “American Psychological Association – Glossary of Psychological Terms.” APA. APA, 2002. Web. 27 March 2016:
[18] Pfau, M. W. “Who’s Afraid of Fear Appeals? Contingency, Courage, and Deliberation in Rhetorical Theory and Practice.” Philosophy & Rhetoric 40.2 (2007): 216-237.
[19] Walton, D. N. “Practical Reasoning and the Structure of Fear Appeal Arguments.” Philosophy & Rhetoric 29.4 (1996): 304.