University of Southern California Dissertations and Theses

The causal account of liability to self-defensive harm reconsidered

(USC Thesis Other)

The causal account of liability to self-defensive harm reconsidered

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE CAUSAL ACCOUNT OF LIABILITY TO
SELF-DEFENSIVE HARM RECONSIDERED
by
Reeves Davis Cameron
A Thesis Presented to the
FACULTY OF THE USC SCHOOL OF PHILOSOPHY
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF ARTS
PHILOSOPHY AND LAW
August 2023
Copyright 2023 Reeves Davis Cameron

ii
TABLE OF CONTENTS

Abstract……………………………………………………………………………iii

Introduction………………………………………………………………………...1

Section I: The Causal Account……………………………………………………..4

Section II: The Metaphysics of Causation………………………………………….6
II.a: Causation Generally…………………………………………………….6
II.b: Counterfactual Causation………………………………………………7
II.c: Other Concerns in Causation…………………………………………...9

Section III: Causation and Rights Violations……………………………………...12
III.a: Introduction to Rights………………………………………………...12
III.b: Claim Rights and Why We Have Them……………………………...13
III.c: Duties…………………………………………………………………14
III.d: Rights, Duties, and Causation………………………………………..15
III.e: Killing………………………………………………………………...19

Section IV: The Account in Principle and Intuition……………………………….22
IV.a: Bystanders…………………………………………………………….22
IV.b: Innocent Threats and Aggressors……………………………………..24

Section V: Remaining Questions………………………………………………….28
V.a: Preemption…………………………………………………………….28
V.b: Indirect Threats………………………………………………………..29
V.c: Problems from Agency………………………………………………...30

Conclusion………………………………………………………………………...32

Bibliography………………………………………………………………………33

iii
Abstract:
There are many hotly-debated questions within the philosophy of self-defense. These range
from those about which self-defensive actions are proportional or necessary, how our theories of
self-defense apply to a wartime context, and even which sorts of killing may be more or less
morally wrong than others. Further, there is also the question (perhaps the most important in the
discipline) of when and why a person may be liable, at bottom, to self-defensive harm. It is my
answer to this question that I wish to put forth in this thesis, and it will proceed roughly as follows.
After first establishing the counterfactual metaphysics of causation that will be used to determine
what makes a person count has having caused something (and more specifically someone's death),
I proceed give a version of Hohfeldian claim rights that allows for ascribing rights violations to
individuals who stand to harm another. Next, I discuss what the interaction between a person's
causal role in events and their duties to other people is that can ultimately lead us to say that they
are liable to self-defensive harm, defend this conception of liability against our intuitions, and try
to address some questions that may still remain. After all of this is done, I will hopefully have
established an account of liability to self-defensive harm based on causation that is more fully
fleshed-out and palatable than those furnished in the past.

1
Introduction
One of the most crucial and uncontested ideas in moral philosophy is the fact that each of
us has a right against being killed or, more broadly, subject to any form of violence by another
human. Just as commonly acknowledged – except, of course, by deeply committed pacifists – is
our intuition that there are circumstances in which this right can be lost. While there are various
reasons why this may be the case, much of the work of identifying which circumstances these are,
and explaining them, is done within the philosophical field of self-defense.
While there are many areas of inquiry in this field, such as how much defensive force is
permissible or excusable (proportionality), or when there are constraints upon it (such as those of
necessity), the keystone of any theory of self-defense which seeks to completely address the
concept are concerns of liability. That is to say, when exactly someone is liable to defensive force,
or when they make it permissible to do harm, even lethal in some situations, to their person. The
various popular accounts of liability to defensive harm, keeping in line with the field as a whole,
seek to explain when and how it becomes permissible to harm another in self-defense while
maximally accommodating our pre-theoretical intuitions about the subject. While, at least as it
seems by reading the literature of the subject, some of the cases in which the question of self-
defense arises are more cut and dry – such as the fact that one may not kill a bystander or that they
can certainly kill someone with intact agency culpably trying to kill them – there exists an
extensive grey area in which it is simply not clear whether one has made themselves liable to
defensive harm or not.
In seeking to explain this grey area various theories have been put forward, such as the idea
of agent-relative prerogatives, the culpability account of liability, the moral responsibility account
of liability, the causal account of liability, and the moral status account of liability, to name some
of the more significant ones.
1
But, I believe – even though it has fallen somewhat out of style
recently – that with some defense and modification the causal account of liability can best match
most if not all of our intuitions about liability, and is certainly defensible in principle. So, it is that
account that will be the main subject of this thesis.

1
For examples of these alternative views, see Quong, Jonathan, (2020). The Morality of Defensive Force, New
York: Oxford University Press., and Frowe, Helen (2014). Defensive Killing. Oxford, UK: Oxford University Press.,
and Otsuka, Michael (2016). The moral responsibility account of liability to defensive killing. In Christian Coons &
Michael Weber (eds.), The Ethics of Self-Defense. Oxford, U.K.: Oxford University Press., and Ferzan, Kimberly
Kessler (2012). “Culpable Aggression: The Basis for Moral Liability to Defensive Killing”, Ohio State Journal of
Criminal Law, 9(2): 669–697.

2
One of the most influential philosophical works in the development of the causal account
is Judith Jarvis Thomson’s 1991 article “Self-Defense”. In this article, she proceeds using a
foundational understanding of rights – a largely Hohfeldian formulation of them – and proceeds
to go through ways in which someone may violate our rights that could render it morally
permissible (not merely excusable) to kill them in self-defense. In doing this, she creates cases of
ways in which actions may render someone liable to defensive harm given justifications that are
roughly similar to what we now consider to be the culpability and moral responsibility accounts
of liability to defensive harm. These versions of an explanation for why someone may be liable to
self-defensive harm she rebuts with her “Irrelevance of Intention to Permissibility” and
“Irrelevance of Fault to Permissibility” theses respectively. Her methodology, however, proceeds
largely by an analysis of her beginning principle, being most similar to an account which says
permissibility to defensive force hinges on culpability. Then, she tests this principle against
intuitions, the strength of which she uses to whittle that starting principle down to one that merely
relies upon a causal role of an agent to render it permissible to defensively harm them.
As I’ve surely made clear in this thesis up to this point, my formulation of why a version
of the causal account is ultimately the most sensible lens through which to view liability to
defensive force will rely heavily on its intuitive appeal very similarly to Thomson’s work and much
of the philosophical work in the literature.
2
But, I feel that this has the potential to overlook the
theoretical underpinnings of the causal account – favoring too much the intuitive appeal that it has
for me (and I hope for the reader as well by the end of this thesis), and not developing and
defending enough the principle that underlies it. So, I will proceed with trying to reach, at bottom,
a strong theoretical backing for a causal account of liability to defensive harm beginning with
section three.
Here I will give a brief sketch of how the thesis will proceed in entirety from this point on.
In section two directly proceeding this one I will go about giving a brief, but I hope comprehensive,
explanation of what exactly the causal account of liability to defensive harm I am seeking to defend
here and what its key feature are. Then, in section three, I will set out the metaphysical notion of

2
This move is made across the literature outside of just those works which promote a causal account, but for another
example of a causal account of liability besides Thomson’s, see Kamm, F.M., 1992, Creation and Abortion: A Study
in Moral and Legal Philosophy, New York: Oxford University Press.

3
causation that I will rely upon when setting out my account. After that, in section four, I will go
over the notion of rights that I take to be the most complete and defensible one, and will be using
in this thesis. This will include discussion about what exactly rights are, what their violation means
and the permissions that such a violation gives to those who seek to defend their rights, why mere
causation can constitute moral action sufficient to say that a rights violation has occurred, and what
constitutes killing – the rights violation that will be discussed most often in this thesis. In section
five, I will go on to show where our intuitions line up with my causal account of liability to
defensive harm, hopefully clearing up some areas where the account has previously been contested
and showing why it is ultimately the most desirable view of what creates liability to defensive
harm and makes killing permissible at times. Then, in section six I will address some of the
common objections to the causal account of liability generally, and some points of tension between
our various intuitions and the proscriptions of the causal account as I have laid it out. I will seek
to answer these objections and at least address some of the weak points of the theory I lay out here,
and that should close out this thesis.

4
Section I: The Causal Account
Now, I will explain what exactly the causal account of liability to defensive harm is that I
will actually be defending in this thesis. The three key components that make up this account are
an understanding of rights and violations of rights (along with the duties, permissions, and
liabilities they create), a unified account of causation through which one can actually be said to
cause a rights violation, and the interaction between these two concepts.
As for rights, while I will go into a deeper discussion of them shortly, the causal account
that I am defending takes them to be such that we naturally have rights and that those rights create
duties in other people to do or not to do certain things. When another person is credibly threatening
to violate these rights, meaning that they are already in the midst of an action that will do so or are
clearly and imminently about to begin such an action, they fail to uphold a duty, and as such it
becomes permissible for you to take action to enforce that duty. In other words, they become liable
to actions that are – in the scope of this thesis – necessary for you to enforce that duty. In most of
the cases that will be discussed here and are discussed elsewhere in the philosophical discipline of
self-defense, the duty they fail to uphold is not to kill you, and the permission that you gain from
that failure is to kill them.
This conception of rights is, however, common to many theories of liability to defensive
harm. What makes the causal account that I am defending unique is that it relies upon mere
causation as the thing which can do the moral work of violating a right, rather than requiring some
more stringent standard on what does and does not count, such as culpability. The specific flavor
of causation that I will utilize is one that is rooted in David Lewis’ development from
counterfactual, to then causal dependence, and while this will be elaborated on even sooner than
my conception of rights, I will say for now that this basically means that if one event would not
have occurred were it not for another, then the former event can be said to play a causal role the
latter. To get even more specific, for us to say that one person, A, has violated the right to not be
harmed of another, B, we must be able to say that the causal role that A plays in B’s harm is one
that is relatively insensitive to changes in circumstance. This conception of insensitivity is one that
plays a large role in my explanation of causation, and I will explain it in the section directly after
this one. Backed by those conceptions of rights and of causation, the causal account of liability to
defensive harm that I will ultimately defend is this: that if one person plays an insensitive causal

5
role in the harming of another, through their actions, movements, or presence,
3
that person is liable
to defensive harm which may even be lethal if the particular harm that they have played a causal
role in is death.

3
This summation of the ways in which one can cause (by their actions, movements, or presence), which I have
found particularly useful, comes from Helen Frowe in chapter 1 of Frowe, Helen (2014). Defensive Killing. Oxford,
UK: Oxford University Press.

6
Section II: The Metaphysics of Causation
II.a: Causation Generally
Key to any causal account of liability is an account of causation itself. And, I take it to be
the case that any account of causation must answer for two key questions in order to adequately
support the idea that a person’s having a causal role in a threat to another generates in the first
person liability to defensive harm: what exactly counts as causation (the thing which would allow
us to say that one person has a causal role in certain events), and why causation should be a basis
for moral judgements the likes of which we would make in saying that it is permissible to
defensively harm someone. In general, the form of causation that I will be looking at in this thesis
is token (or actual) causation, where specific relata are said to have a cause and effect relationship
with each other in specific instances. I will also largely be relying on the counterfactual theory of
causation primarily promoted by David Lewis, with a conception of the relata in question in a
causal relationship as being events.
4

As far as the relation between the cause and effect in the theory of causation that I will be
relying upon goes, one that is counterfactual is relatively simple and elegant, and, in my mind, is
quite aligned with our intuitions about what exactly causation is. Before going into the matter of
the relation between cause and effect, I will go ahead and clarify what exact sorts of causes and
effects, or the relata, that I will be discussing are. Here and in the rest of this thesis, causes and
effects will be viewed as events, rather than facts, propositions, or anything else. Events are the
sort of relata that I believe can be most readily applied to a discussion about self-defense. And
while there is merit in seeking to address the relationship between causes and effects of other
forms, as has been frequently done in the literature, that is not something that I will seek to do
here. To explain this decision further, as it is one that has consequence in any discussion of
causation in regards to situations of self-defense, I will say that I think events are the kind of
classification that can cover the whole spectrum of relata in the scenarios that will be discussed.
When compared to propositions and facts, I think that this is clear. Any proposition that would be
useful in a discussion of relata will naturally be one that pertains to the existence or occurrence of
facts, events, actions, or other possible relata, so I find it reasonable to cut them out and simply

4
For more on Lewis’ counterfactual theory of causation, see Lewis, David (1973). Causation. Journal of Philosophy
70 (17):556-567. and Lewis, David (1986). Causation. In Tim Crane & Katalin Farkas (eds.), Philosophical Papers
Ii. Oxford University Press. pp. 159-213.

7
focus on the more basic relata that they would be making claims about. For facts, it is my view
that they also would only be facts about some more basic relata and are thus less desirable than
something like, say, actions or events. In the literature about causation, facts as relata have been
primarily put forward to address the problems that could arise from absences, which are generally
taken not to count as events, but I do not see them as being properly relevant to our discussion
about causation, rights, and self-defense. There is still another possible relata, being actions, that
could be seen as more appropriate to use in my discussion here. When it comes to self-defense
specifically, we would first thing of one person’s pushing a boulder off of a cliff at another, for
example, as an action by that person, and to say that it is an event could sound like a mislabeling
or overcomplication. However, the actual creation of a threat is only one half of any self-defense
scenario, as someone being struck by that previously mentioned boulder could not be seen as an
action at all. So, what I will do is call that instance of being struck as an event, and the action,
movement, or presence of whatever individual causes or acts as the threat another event, in order
to have matching relata while being as clear as possible. Beyond these, any possible relata said to
have or not have a causal relationship here will be events, and I think it is appropriate that this be
so.
Further, I do not think that a choice of grain as to what exactly events are will be helpful to
the discussion here either. While we may go into whether an event is constituted by something as
fine as a single impact or minute motion, or as broad as an entire happening taking place over a
large area, I think that such concerns could only serve to muddle our conception of causation as it
relates to scenarios of possible self-defense. And, as such, causes and effects in this thesis should
be understood as events in the most common-sense way possible.
The discussion herein will unsurprisingly be focused on token, and not type, causation –
meaning that I will not be concerned with generalities about what sorts of things constitute causes
of rights violations the sort that would merit a response of self-defensive action. Rather, I will be
addressing specific events that do, in specific situations, raise the question of whether a person has
been a cause of another’s harm or, as the cases will be more frequently in the examples in this
thesis, death.

II.b: Counterfactual Causation

8
So, it is between specific events, as they are commonly understood, that I will say there is
causation by virtue of counterfactual dependence. This notion of causation, as previously
mentioned, comes largely from the account developed by David Lewis in his 1973 article
“Causation”, and his chapter by the same name in the 1986 book Philosophical Papers Volume II.
In these works, he lays out a theory of causation that first establishes what it is for two events to
be counterfactually dependent, then causally dependent, and then what exactly it is for there to be
causation between them. In seeking to explain counterfactual dependence in a simple way, I will
first say that Lewis states that “a counterfactual is nonvacuously true iff it takes less of a departure
from actuality to make the consequent true along with the antecedent than it does to make the
antecedent true without the consequent,” (Lewis, 1973, p. 561). And if a counterfactual, or the
statement that if one proposition (this proposition in the case of events specifically being that a
given event has occurred) is true then another proposition would be true, is true itself, then the
events which that statement is about can be said to be counterfactually dependent. As a side note,
what Lewis means for a counterfactual to be vacuously true, as opposed to nonvacuously, in terms
of events is when that counterfactual is simply true by virtue of the fact that there are no possible
worlds in which the event the occurrence of which is the subject of the antecedent proposition of
the counterfactual does occur.
Next, Lewis goes on to state that “[an event] e depends causally on [an event] c iff, if c had
not been, e never had existed,” and further that because causal dependence implies causation
(although a lack of causal dependence does not rule out causation), “one event is the cause of
another iff there exists a causal chain leading from the first to the second,” (Lewis, 1973, p. 563).
This largely completes Lewis’ account, but there are still a few minutiae to clarify, some with the
potential to cause my account problems, down the line. The first couple of these, and ones that are
fortunately not problematic, are those of effects and epiphenomena, in other words when a
counterfactual analysis makes it difficult to differentiate causes from effects in terms of
counterfactual dependency, and when there is a causal byproduct that comes as a result of a given
causal chain the occurrence of which can make exact causal relationships difficult to identify.
These issues are not pertinent to the discussion in this thesis, but I believe that Lewis addresses
them sufficiently enough to say that the account of the metaphysics of causation that I adopt here
is complete in regards to its ability to handle those sorts of questions.

9
II.c: Other Concerns in Causation
Next, there are the more troublesome questions of preemption and normality concerns (the
latter seeming to me to nag at any theory of causation that seeks to answer for our intuitions about
the subject and is not needlessly complex). Preemption is the problem that occurs when there is a
causal overdetermination such that if one cause were absent or failed to create an effect, another
cause would then create that same effect. One clear example of preemption, pertinent to discussion
about self-defense, is a case such as Two Hitmen: John is walking home from work, when he is
approached by who hitmen, Bill and Will, each hired by different clients and each previously
unaware of each other’s existence. Bill is preparing to shoot John, but when he sees that Will is
already about to do so, he simply refrains and watches, ready to fire his shot if needed.
5
Here, if
one hitman did not kill John, the other would have, so it can be said that John’s being shot is not
counterfactually dependent on either of the hitmen individually shooting given that if one did not,
the other would. In such cases, it cannot be said that the effect is counterfactually dependent on
either potential cause, whether the preemptory one or the preempted, because its occurrence is not
dependent on either one of them. As we will later see, cases of preemption could present a real
problem for my causal account of liability to defensive harm, given that we seek for it to be
grounded in a complete metaphysical account of causation. However, here and going forward I
think it will suffice to say that cases of overdetermination in self-defense scenarios may be
considered cases where although causation does exist, by virtue of a causal chain connecting two
events in question, that causation exists without causal dependence.
As for normality concerns, they are broadly those which ask which events, out of the
myriad small factors upon which an effect can causally depend, are we actually willing to say is a
“cause” of a given effect in a meaningful way. Some attempts to solve for this problem simply rely
upon conceptions of what is and isn’t out of the ordinary to determine what is an isn’t a significant
cause, with a common example being that due to oxygen’s presence being so ordinary on Earth,
we would never say that it is what caused a fire regardless of a fire’s counterfactual dependence
on the presence of oxygen.
6
Rather, we would say that it was the lighting of a match, or something
of the sort, that was the significant “cause”. This sort of normality analysis is obviously quite

5
This case, and much of my concern with preemption in general, comes from Victor Tadros’ remarks in chapter 8 of
Tadros, Victor (2011). The ends of harm: the moral foundations of criminal law. New York: Oxford University
Press.
6
This example is found in the entry “The Metaphysics of Causation” in the Stanford Encyclopedia of Philosophy

10
unstable ground upon which to base any theory of when a person is liable to defensive harm,
particularly due to subjectivity concerns – to reference the fire example again, an alien species
from a planet without oxygen in its atmosphere might in fact say that oxygen is the most important
cause of a fire happening on earth. But, while Lewis himself ignores that specific subject rather
dismissively, he fortunately provides an analysis of causation specifically in the case of killing
(which can be made to apply to simple harming as well), which answers for most if not all of the
normality concerns we may face when trying to ascribe causation to the event of someone’s death
or injury. There are some grey areas, though, where the idea of normality in our ascription of
causation becomes useful, and I will address those in section five of this thesis.
This analysis is provided in what is simply a very late post-script of Lewis’ original paper,
in a section which he calls “Insensitive Causation”. As the name of the section suggests, it is that
sort of cause specifically, which is not sensitive to minor changes that may impact the causal chain
that carries on from it, that is sufficient to be the right kind of “cause” for someone’s death. As
Lewis puts it, “killing must be a special kind of causing to die,” and I think this captures our
intuition about what exactly killing is well (Lewis, 1986, p. 185). So, in figuring out what exactly
could make causing someone to die in strictly counterfactual terms not actually count as killing,
he rules out the simple factors of that causation being a part of a causal chain or even one that is
temporally long or complicated, having a low chance to result in death, being unforeseeable, or
passing through further human actions. This is not to say that none of these may be present in what
we consider killings, or that none of them combined and in some specific manner may contribute
to what he eventually lands on as making something not count as a killing, but simply that none of
these are things the presence of which would tempt us to say that something is not a killing. What
he eventually lands upon is that causation of the right sort to constitute killing is causation that has
a low sensitivity to circumstances, such that even if some changes were to occur between the cause
and its effect (being death) the cause would nevertheless persist in bringing that effect about. In
other words, while some action such as selling a gun to an individual can have a causal role in a
person’s eventual death with that gun, that effect actually occurring is very sensitive to the
circumstances that obtain between the selling and the killing. Pointing a gun at someone’s head
and firing it, however, is a potential cause that would require extraordinary circumstances not to
have the resulting effect of death, and it is this insensitivity circumstances that allows us to call
that causal role an act of killing (Lewis, 1986, p. 186). I find this to be an elegant way of answering

11
for normality concerns as well as others in situations of killing and harming and will adopt it for
the purposes of this thesis.

12
Section III: Causation and Rights Violations
III.a: Introduction to Rights
Now that I have covered the sort of causation that we are discussing here, and as such what
exact kind of causal role one must have in a threat to another person for them to count as potentially
causing harm to or killing that person, I will go into the rights-based moral framework that I will
say ultimately justifies self-defensive harming in reaction to such a threat. In doing so, I will cover
what the exact conception of rights I will be using here is, and why a violation of those rights
makes self-defensive harming, and in some cases killing, justified.
The system of rights, most importantly that right that we all have to life, used by Thomson
and other defenders of the causal account, is widely accepted and utilized in the philosophy of self-
defense, but I will seek to explain it here within the framework of causation as something that can
do moral work for the purpose of establishing the completeness of my theory. What will be of
prime importance is that, as a basis for the further developments of my theory of liability, a rights-
based approach maintains that it is morally permissible (once again more than merely excusable)
to kill someone if they pose a threat that will cost you your life. The threshold for what sort of
rights violation is needed for exercising some kind of defensive force less than killing your attacker
is up for debate, and a hotly contested one at that, but as this thesis and much of the literature that
it draws off of focuses the majority of its attention towards the act of killing in self-defense, I will
not discuss rights violations in degrees here. Such a discussion is certainly valuable, but is more
one of proportionality than liability, and can be worked out after basic liability is established. And,
in any case, regardless of the fact that killing is simply another degree of harm, it does have a
powerful intuitive force behind it and a theory of liability to defensive harm which could be held
back merely by our intuitions about killing surely would not have a firmly generalizable principle
behind it. But, to go on from that aside, once it is established that a certain, impending violation
of your right to life can be said to be attributed to someone, the question primarily is: what is such
an impending violation of our right to life that it gives us moral permission to kill?
As I briefly mentioned earlier, the sketch of rights that I will broadly be using is that which
is used by Judith Jarvis Thomson in her discussions about rights and liability to defensive harm,
and one that she, in turn, takes primarily from Wesley Hohfeld’s conception setting out of a system
of rights in an article of his from 1913 (Thomson, 1991; Thomson, 1990; Hohfeld ,1913). This
Hohfeldian system of rights is now very widely used as a way in which we can understand certain

13
sorts of rights and their correlatives in others said to be interacting with rights-holders, and while
it was originally intended as a discussion of rights in the strictly legal sense, I, like Thomson, take
it to apply just as easily and appropriately to questions of morality. This account of rights covers
four categories of what we would refer to as rights in a common-sense way, being claims,
privileges, powers, and immunities. It is the first of these categories, claims, that both Thomson
and Hohfeld describe as being rights in the strictest sense – or the most basic, stringent form of
rights that people can have (Thomson, 1990, p. 39). Such rights are rights that we specifically have
in regard to the actions of others directed towards us or the things that we have interest in and are
attached to, and I believe that it is this sort of right that is most frequently and clear the subject of
discussion when it comes to the subject of self-defense. Claims, as well as the other (perhaps seen
as lesser by Hohfeld) rights, have correlatives: for claims, privileges, powers, and immunities,
these correlatives are duties, no-claims, liabilities, and disabilities respectively (Hohfeld, 1913).
So, if one person can be said to be possessed of one of those sorts of rights, it could also be said
that any other person engaged in a moral interaction with the first is in some way possessed of the
correlative of that right.

III.b: Claim Rights and Why We Have Them
So, it is clearly a basic claim-right that is the sort of right we are talking about when we
reference the right not to be harmed, and as such we will be talking about that claim, and its
correlative notion of a duty not to harm someone, in this thesis about self-defense. As an aside, I
am absolutely sure that there are areas in the literature of self-defense in which privileges, powers,
and immunities – other sorts of rights referenced by Hohfeld – may come into play. That will not,
however, be the subject of this paper, and the discussion of when rights such as those may arise in
the context of self-defense is another discussion entirely. Back to claims and duties, Thomson sums
them up with such simplicity that I will simply quote her here. In regards to claims, she states that
they are of the form that “X has a claim against Y that p”, with “p” being any sentence one plausibly
believes could be the substance of a claim, and for the correlative duty of such a claim she describes
it as “Y is under a duty toward X, namely the duty that Y discharges if and only if p,” (Thomson,
1990, p. 41). When discussing what exactly that “p” could be that serves as the actual substance
of a claim-right, I see it as evident that something the sort of “X not be harmed by Y” is perhaps
the version of it least able to be contested. As will be discussed, we as humans are instantiated in

14
bodies that make up a key component of who we are, and that body not being reduced in any way,
and therefore our very selves not being reduced, is one of the most crucial claims that we may have
against anyone else. This being said, I think that it is appropriate to move on to why we may be
said to have any claims at all, and how those claims may be violated.
In terms of why we have certain claims, such as that others not inflict bodily harm upon
us, it is my view that they are ultimately at the core of our conception of morality. While claims
that invoke positive duties such as claims that others do have a directed duty to do certain things
towards us, or that there are claims of society as a whole that impose general duties upon people,
are a little trickier and harder to defend, those that create negative duties in others are essential to
our existence as moral agents. Further, while even many negative duties could be up for debate
and may be seen as going too far in limiting the actions of other people, that claim – and its
correlative duty – that we may not be harmed is perhaps our most basic one. If we were not able
to have a firm moral claim upon the ability to simply exist in our bodies without those bodies being
diminished, then we cannot possibly have a basic notion of self-interest and self-respect as moral
agents or even patient. So, in our very existence and in seeking to continue that existence, we all
intrinsically make the claim upon others that they have a duty not to cause us bodily harm. Given
that making this claim – and further assuming that it constitutes a constraint that others ought to
respect – is something that we do in every single moment in which we exist, it would be absurd
for any one person to say that this claim is not a valid one. This may be the best we can do in
determining what exactly claim-rights and their correlative duties are at base: they are, as Thomson
says, simply a constraint on behavior (Thomson, 1990, p. 201). And, when it comes to the claim
that we not be harmed by others, this seems to be a constraint the validity of which nobody can
reasonably deny.

III.c: Duties
As I have mentioned, each claim-right establishes a correlative duty which actually gives
form to that constraint on behavior that constitutes the claim. In the case of a claim to not be
harmed, the duty imposes on others is, as simply as one would expect, that they not harm you. So,
the next question is how exactly we go from one person having a claim not to be harmed and
another having a duty not to harm them to the former person being morally justified in using self-
defensive harm against another. The answer to this question lies in what actions we may take to

15
actually enforce the duties that other have towards us. It is clear that the constraint imposed upon
a person by their having a duty means nothing if that duty cannot be enforced; if this were the case,
then what use would the concept of a duty be in the first place?
So, when someone is in the course of not fulfilling the duty that they have not to harm you,
we can absolutely say that have the ability to move to enforce that duty in a manner that you would
otherwise not have to if person who is going to harm were not going to violate it. I will be staying
away from a discussion about necessity in this thesis, but I will say that I generally take us to be
bound to enforce others’ duties towards us in a way that minimally infringes upon any claim-rights
that they themselves have. So, if one’s enforcement of a duty need not entail the self-defensive
harming of those upon whom that duty lies, then they are not liable to such harming, and it is not
a morally permissible action. The question of proportionality may also be raised here and, while it
is one that I will also not spend time addressing in this thesis, I take it to be a reasonable rule that
the duties towards a person that one may ignore in seeking to enforce those duties that that person
has towards them are constrained such that they may at maximum be of equivalent moral weight.
For example: you may not kill another person in seeking to enforce the duty that that person has
not to punch your arm. However, if self-defensive harming, and in the most severe cases killing,
is the only option available to one seeking to enforce the duties of an equivalent moral weight that
another has towards them, then we may say that it is permissible for that person to self-defensively
harm the other, and that the other person is indeed liable to such harm.
It would seem appropriate here to briefly address the issue of enforceability. I do not wish
to completely dive into the question of what makes some duties, at bottom, enforceable and others
not. However, it does seem that there is a clear divide in our intuitions about which duties are and
are not. Generally, those which pertain to one’s body and other property, such as the duties not to
harm someone, kill them, or steal their property are enforceable, while lesser duties such as those
not to degrade someone or speak ill about them are not. As mentioned, those duties which pertain
to situations of self-defense are enforceable at least to the level of proportionality which I have
described, and I think that this is likely so because of the unique nature of our interest in the
tangible things that matter to us in this world and can be impacted physically by others.

III.d: Rights, Duties, and Causation

16
Now that we have established what exactly causation is, as well as what rights violations
of the sort we are discussing are generally, the next question that must be answered is why
something like mere causation, and not moral responsibility or culpability, can and should be seen
as a basis for moral judgments. Consider those three common justifications for assigning moral
blame to somebody for an event (those mentioned – culpability, moral responsibility, and
causation). It seems clear that causation is at the center of each of them, and that which actually
differentiates them from mere causation is of considerably little moral substance without causation
giving it the main thrust. Here, I am obviously seeking to lay out a logical, rigorous, and well-
reasoned justification for why mere causation should be thing upon which we rest any ascription
of liability to self-defensive harm, but I do think it a useful demonstration to think about the way
that we think of blame and of moral wronging through common sense. When we speak of liability
to self-defensive harm and why it is not permissible to inflict on certain people (for example a
bystander who one could use to shield themselves with an arrow), we don’t think that those
people’s lack of liability comes from a lack of culpability or a lack of moral responsibility for the
imminent harm that another is facing. Instead, I believe that at least most of us would think that it
is plainly because they have not causally contributed in any way to the situation that stands to harm
someone else.
I think that this serves as a strong reminder as to what is at the core of our assignation of
liability to self-defensive harm, and that the causal account of liability to self-defensive harm
which I am advocating for and which is suggested by the common view need not be viewed as
overly draconian or set aside in search for some more stringent threshold for liability. It certainly
is a view which, if accepted, entails the permissibility of self-defensive harm more than any other,
and it certainly does allow us to say that people can be liable for the harm that they cause even if
the circumstances that lead to their causing that harm is entirely beyond what they can exercise
meaningful agency over. But, I think that such disregard to moral responsibility and to culpability,
while severe in its consequences, is correct in that it is a moral position that I am taking in response
to severe circumstances. When it comes to matters of life and death, there is no room to give the
benefit of the doubt and to judge their actions, movements, and presence by anything other than
the fact they are as they are. And, I do not think that this is a view that we are merely forced into,
but that it is proper that this is so. Adult humans at least simply are at least minimal moral agents,
to and by which the things done carry moral weight. And, as humans, we are instantiated in bodies

17
that are because of our inextricability from them an essential part of who we are. As such, from
the moment that a person becomes a moral agent the mere actions, movements of their bodies carry
moral weight. Without our bodies representing our presence and stake in this world, the rights
against harm of which I speak in this thesis would be incoherent in the first place, so it seems quite
reasonable to think that those bodies, and their interactions in the physical world, simply are the
things upon which our moral judgments in this area stem from.
I would like to make it clear here that I am not, or am at least not seeking to, imply that by
being what I have called a “minimal moral agent”, people are at all times possessed of some sort
of minimal moral responsibility – something which would essentially render my account a broader
version of the moral responsibility account. The idea that people are moral agents has its place, in
my account, not in determining what moral judgments can be made about what actions at which
time, but rather in determining who it is that any moral judgments can be made about at bottom.
So, I would hope that I am understood when I say that because of a person’s status as a moral
agent, we are able to make moral judgments about their actions, movement, or presence – those
judgments themselves being based in the causal role that such actions, movement, or presence may
play in affecting another person. We can make such judgments about people who are only minimal
moral agents having no meaningful exercise over that agency at a given time – such as a man who
has been pushed off a balcony towards another person and who my account will say is liable to
self-defensive harm – but the things about which we make moral judgments about must still be
moral agents to some level. As such, we do not make moral judgments about a boulder even though
it may play a significant causal role in a person’s death. To go further, I would think that we should
not be able to make firm moral judgments based on causation about mere moral patients, such as
dogs and children, who we do think have a moral stake in the world but have never themselves
been moral agents and so at no time have some minimal moral agency. So, as it stands not, the
causal account of liability to self-defensive harm would hold that we cannot make moral judgments
about stones, dogs, infants, or young children. We can make moral judgments about adults, though
– judgments that would hold that innocent threats, innocent aggressors, and villainous aggressors
are liable to self-defensive harm, while bystanders are not. There does remain, however, a grey
area between a child and an adult that presents a problem I will address later.
I think that from what I have established, we have a perfectly good foundation for seeing
the causal role played as the basis of our moral judgments about the consequences of one’s actions,

18
movements, or presence. However, it would still do to look at what is probably the nearest step
“up” (in terms of more stringency in ascribing liability to self-defensive harm) from causation. As
far as I have seen, this would be an account of liability to defensive harm centered around a
conception of moral responsibility, and I do not think that the reasons why one would look to a
person’s moral responsibility for a situation which will bring harm to another are compelling
enough to ignore causation, nor the result of such an account more intuitively appealing.
Looking at moral responsibility, this threshold is supposed to be differentiated from
causation because a moral actor can at least be said to be responsible for their actions. In this way,
it is said that they are made answerable to morality and its demands (even if those demands include
becoming liable to harm) because by being a morally responsible for one’s actions it is then
appropriate to include those actions in some moral system. This conception of that what can serve
as the basis for moral judgments does certainly have an intuitive appeal to some degree, and
provides an answer for why we feel that a baby, or even a dog, should not be held responsible for
things they may cause in the same way that a regular, adult moral agent would (though, mind you,
I do feel that I have a sufficient answer for this problem in my conception of a causal account of
liability to self-defensive harm). There is, however, little substance in the thought that moral
responsibility for given actions is what allows us to include those actions in our moral calculus.
What I mean by that is this: when actually trying to derive why our most basic notions about
morality and moral work only apply to those specific actions over which one can be said to have
some degree of moral responsibility, some of the most foundational papers establishing the account
seem to rely heavily on the well-established principle that one ought not harm an innocent
bystander in their efforts to save themselves from harm. From there, they use the idea that there
are in fact no morally relevant properties which differentiate an innocent bystander from an
innocent threat or even an innocent attacker (categories which we will dive more into later), and
therefore derive that although we may intuitively find killing an innocent attacker or innocent
threat acceptable this is clearly a mistake on our part.
7
So, when proponents of the moral
responsibility account then seek to determine what it actually is that differentiates someone like an
innocent threat or attacker from someone who engages in the actions that constitute a threat with

7
See McMahan, Jeff (1994). Self-defense and the problem of the innocent attacker. Ethics 104 (2):252-290., and
Otsuka, Michael (1994). Killing the Innocent in Self-Defense. Philosophy and Public Affairs 23 (1):74-94. for
examples of this attitude.

19
sound mind, they fall upon moral responsibility, and not the higher threshold of culpability, as what
must be the differentiating “morally relevant property”. It does not seem clear to me that a moral
relevance to our actions, movements, and presence simply relies upon some fleeting concept of
moral responsibility for our actions, it is merely something that we are, and this is at least clear in
our idea that actions towards us at all times, whether we are engaged in a way that could be said
to be morally responsible at the moment, have some sort of moral value. As such, I see no reason
why the causal role it plays cannot be a morally relevant property of one’s actions, movements, or
presence, and hence reject the dilemma posited by McMahan and Otsuka between seeing it as
impermissible to kill an innocent attacker or permissible to kill an innocent bystander.
The position championed by the casual account, that playing a causal role in an impending
violation of one’s right to life renders a person liable to defensive harm, seems as I have said to be
the lowest generally accepted threshold for liability to obtain. While it would be interesting to
consider what even lower thresholds may be possibly drawn upon as making a person who crosses
it liable to lethal defensive force, I do not think that such an account would be either theoretically
defensible or intuitively desirable, so I will abstain. Going up the ladder towards a higher threshold
for liability to defensive force, I think that I have already established why at least the next least
stringent account of liability to self-defensive harm is not preferable to the casual account in
principle, and will show later on how it is not more intuitively appealing either. So, I take what we
have already discussed about rights violations and causation to be sufficient to move forward in
developing more of how causation and rights violations interact.

III.e: Killing
Here, I do not mean to turn towards the more fine-grained distinctions like those considered
in supporting a local or global view of causation or considering causal location and contribution.
Instead, in trying to establish the basic principle of a casual account for liability to defensive force,
I will focus on a very specific set of circumstances: where one is going to play an insensitive causal
role in your death unless you kill them in self-defense first. To be clear, this can be a person with
a causal contribution that is not sensitive to circumstances in a causal chain that will invariably
result in your death. This can be a role that sets into motion some merely physical chain of events
which leads to your death, or convinces another person to kill you, or is merely the last thing
(which will actually kill you) in a whole causal chain. That last sort of causation will clearly be

20
one of the most contentious in this thesis, as it has been in the literature, and I will discuss it further
on. However, as Thomson points out if another object was the last in a causal chain of things
leading to someone’s death – for example, a falling rock – we would certainly say that the thing
“killed” that person.
8
This is the sort of causation leading to a rights violation that I wish to say
makes someone liable to defensive harm.
An initial reaction to this may be that not every version of what I have said would constitute
someone playing a causal role in your death is the sort “killing” than we’d accuse a murderer of.
Ultimately, this is true, but that need not be the case. One version of the word seems to impute
some sort of agency upon the killer, and the other only refers to a position in a causal chain. Further,
the use of the word is garbled even when it is merely causation that’s being discussed, for if one
moral agent is located reasonably proximately to a person’s death on a causal chain then we would
say that it was they who did the killing. If a villain were to drop a boulder off of a building onto
someone’s head, some may still say that a boulder “killed” that person, or that a person was “killed
by” a boulder, but I hope that you will agree with my intuition here that the formulation of things
that most accurately captures what occurred would be to say that the villain “killed the person with
a boulder”. However, I think that this is solely because it is the villain who, as a person, has the
most clear and significant role in the casual chain that leads to death. It is still proper to say that
someone is killed by a falling object or by an unfortunate meal or by any inanimate object that
may cause death. And while this is different from the intentional sort of killing we’d usually ascribe
to murderers, it is killing nonetheless. The reason why it feels different, as I have pointed out
earlier, is the fact that some inanimate object, like many things that may kill you, is not a moral
being. Therefore the word killing, while still valid to use given the causal role those things may
have, has no moral significance when ascribed to something that is not inherently moral.
The relationship between our use of the word killing and causation unfortunately becomes
even more muddled when we consider how far back a sole person could be in a causal chain and
still count as exercising their moral agency over the entirety of that chain, or who we could say is
the killer in a lethal causal chain that includes more than one person. Still, even with all of this
confusion there must be some common strand running through all of our uses of the word “kill”.
If saying that one thing kills another imputes upon that first thing some kind of property such that

8
For discussion of this, see Thomson, Judith Jarvis (1991). Self-defense. Philosophy and Public Affairs 20 (4):283-
310., page 288.

21
the motion of the first is responsible for the death of the second, could we potentially say that a
moral agent’s actions somehow imbue all later objects in the causal chain with their own motion
and therefore carries their ownership of that motion through the causal chain? I think that this a
plausible answer, but it still runs against the issue of multiple agents in a chain together. If a villain
pushed one man off of a building onto another, would we say that the it is the villain killed the
man? Or that the villain killed the one man with the other? Or that the one man was killed by the
other man? The interaction between moral agents and causation becomes most confusing here.
What seems to best cut out all of the possible confusion when it comes to the word killing is still,
in my opinion, the use of the terminology of “insensitivity” as used by Lewis, where he looks at a
person’s causal role as constituting killing if it is insensitive to a change in the variety of
circumstances that can some to affect any causal chain (Lewis, 1986, p. 185). This is the
formulation of killing that I will use, so in my causal account of liability to self-defensive harm
when person A has played an insensitive causal role in a causal chain that will lead to person B’s
death, therefore violating their duty not to harm person B, person B may enforce person A’s duty
by killing them in self-defense.

22
Section IV: The Account in Principle and Intuition:
IV.a: Bystanders
I believe that by this point, I have firmly established what exact metaphysical conception
of causation as a whole (and of harming and killing in particular) I will be operating from, the
framework of rights and more specifically the claim-right against being harmed that I say is being
violated when one is harmed, and what exactly the interaction between causation and rights is that
can make someone liable to self-defensive harm. Now, an appropriate next step would be to test
the casual account as I have set it out against our intuitions about more specific cases, where I will
pay particular attention to points of tension between the causal account and some of our more
strongly held opinions in cases of self-defense. In doing this, I will also seek to show how the
causal account can overcome the objections that have led philosophers like McMahan and Otsuka
to claim that it does not draw a sufficiently solid line between harming bystanders and others, and
therefore must not be the correct thing on which to base liability to self-defensive harm.
9

I believe that out of all of our intuitions regarding self-defense, the most central is that one
cannot kill an innocent bystander in order to protect themselves from harm. And, it should be quick
work to show that the causal account of liability to self-defensive harm avoids saying that such a
thing would be permissible. It is by definition impossible that a bystander play any sort of casual
role in harm that will imminently be done to someone else, and this is the exhaustive list of reasons
as to why it would be impossible for one to be liable to self-defensive harm. This does not simply
preclude killing a bystander as a means to save oneself, but also killing one in a way that is merely
coincidental and not as a means at all. An example of this latter type of killing is as follows: a spear
is thrown at a victim with a shield and that person’s only hope of not being killed by the spear is
to move their shield in its path. The person victim only has time to move the shield in such a way
that, if the spear hit it, it would bounce off and impale an innocent bystander. In this case, I do not
think that it would be permissible to move the shield, as the victim still has a duty not to play a
causal role in killing the bystander, and the bystander has not violated any duty of their own not to
harm the victim.
10

9
As mentioned, these remarks are made in McMahan, Jeff (1994). Self-defense and the problem of the innocent
attacker. Ethics 104 (2):252-290., and Otsuka, Michael (1994). Killing the Innocent in Self-Defense. Philosophy and
Public Affairs 23 (1):74-94.
10
It may be asked here whether we can say that the bystander has in fact caused the victim’s death by omission by
virtue of not saving the victim, if we hold that that would be possible in this case. I think that whether or not that

23
I want to make clear, however, that this position does not make impermissible every type
of self-defense that may result in the non-instrumental death of a bystander. Take for example a
similar situation in which a spear is thrown at a victim, but this time they may dodge out of the
way to avoid the spear, and an innocent bystander is standing directly behind them in the trajectory
of the spear. Here, I think that it would be perfectly permissible for the victim to move out of the
way, even if doing so will result in the bystander’s death. I say “result” here, rather than cause,
because I do not think that it would follow from the conception of causation that I have laid out
that the victim would in fact be causing the death of the bystander – and this is why the self-
defensive action in this case would be permissible. Simply moving out of the way of a projectile
does not constitute any causal contribution to the death of another who may be in its path.
11

Another duty that is at play in the two situation I have just laid out, besides the duty not to
harm, is the duty to save someone from being harmed. And when it comes to this duty, which the
bystander in the first case and the victim in the second case certainly violate, I do not think that it
has the moral force necessary to make the actions of those two moral agents impermissible. I think
it is a view felt strongly enough in common sense that the duty to save another’s life, while perhaps
inviolable in cases where it would only cost one, say, a hand, is not strong enough to warrant
someone losing their own life in order to abide by it. There is also the question of enforceability,
and whether or not the victim in the first case, or the bystander in the second case, would be acting
permissibly if seeking to enforce the duty that the bystander and victim respectively have to save
them. I do not think that this would be the case – as I have said previously an easy rule when
looking at the enforceability of those duties which are at times enforceable is that one cannot
enforce another’s duty towards them if doing so would violate their duties of greater moral strength
towards that person. I take it to be the case that the duty to save someone from death has a weaker

determination can be made relies on a notion of normality – such that the bystander’s omission can merely be
regarded as a background condition. If it were normal for a bystander not to save a victim (as I believe it would be in
a majority of cases considering feasibility restraints, then we can say that this is so. If, on the other hand, the victim
may be easily saved and is crying for help, it would be abnormal for the bystander not to and their omission may in
fact be a considered a causal contribution to the victim’s death. For a similar discussion about normality and
omission, see the entry “The Metaphysics of Causation” in the Stanford Encyclopedia of Philosophy, section 1.2.3
11
In strictly counterfactual terms such jumping out of the way would in fact count as causing the death of the victim.
However, I think that we can appeal to normality concerns here as well. Someone jumping out of the way, if they
were able to, is normal human behavior that can be seen as in the background of any threat, in much the same way
that bystanders’ acts of omission in not always diving in to save the day is. And, I think that we could think of many
cases where the avoidance of a threat by some individuals is merely a background condition of that threat’s eventual
killing of another victim. This is obviously a tricky problem, and the question what exactly can be regarded as a
background condition in regards to human nature is something which deserves more space than I have here.

24
moral force upon us than the duty not to kill, so I do not think that enforcing a duty to save by
requiring someone else to die would be morally permissible for the victim in the first case or the
bystander in the second.

IV.b: Innocent Threats and Aggressors
Looking towards the other end of the spectrum from bystanders, it is clear that the causal
account makes permissible the self-defensive killing of villainous aggressors (those who
intentionally seek to a kill a victim) and I do not think that I need to spend the time to elaborate
further on that topic here. Where theories of liability to self-defensive harm in general find trouble,
and where the causal account has been dismissed in the past, is in the grey area between those
obviously immune from self-defensive harm and those obviously liable. The two categories of
individuals within this area that most discussion has centered around are innocent threats and
innocent aggressors – both being non-responsible people whose bodies stand to harm a victim in
the case of the former and whose actions stand to harm a victim in the case of the former. Classic
examples of these two types of people would be a person who has been pushed and is going to
land on and kill a victim and a person who has been mind-controlled into attacking a victim
respectively. My causal account ultimately aligns with our intuitions and holds that it is permissible
to kill both of these types of people in self-defense.
To first talk about innocent threats, they are, as I said, those whose body poses a threat to
your life insofar as its impact with you will kill you and who are non-responsible for this
movement. It is this lack of responsibility, and a similar lack of responsibility on the part of
bystanders, which lead those like Otsuka to claim that we must have the same stance towards the
permissibility of killing innocent threats as we do regarding bystanders. Unlike Otsuka, however,
I do not think that responsibility, or a lack of it, is the only morally relevant difference that could
possibly be pointed to between innocent threats and bystanders, and believe that the causal role
played by an innocent threat’s body in a situation where their movements stand to kill someone
makes all the difference.
One very popular version of a case in which we see an innocent aggressor is one that will
call Falling Man.
12
In this case, a villain shoves a man (our innocent threat) off of a balcony and

12
This case originally comes from Thomson, Judith Jarvis (1991). Self-defense. Philosophy and Public Affairs 20
(4):283-310.

25
down towards a café below. At this café sits another man, who the falling man will land on and
kill (but himself survive) if he does nothing, and who has time to move the awning above him such
that the falling man will bounce off of it and land in the street, dying there upon impact. Intuitively,
it seems as though we should be able to, in the position of the man in the café, move the awning
and save ourselves from dying. Certainly, our duty to save the falling man does not extend so far
as to make self-sacrifice morally required, and so we may have an excuse for moving that awning
and killing the falling man with our intercession on events. However, I think that we can go further,
and say that such self-defensive killing would be permissible. That falling man, and any other
possible innocent threats, have a moral ownership of their body’s presence, movement, and actions
such that the causal role played by their body in any lethal sequence of events does have a real
moral force and can constitute the violation of someone else’s claim-right not to be killed. Otsuka
claims that maintaining such a position would require someone to also hold that a something like
a falling rock could violate rights. And, specifically responding to an excerpt from Frances
Kamm’s book Creation and Abortion in which she says that “The position of [a] threat is different
from that of a natural object . . . because she is not a stone but, rather, a person who should not be
in an inappropriate position relative to others,” Otsuka says that following this rationale, even
infants as innocent threats could be permissibly killed and that it does not make sense that an infant
could violate a right by landing on you but a stone not (Otsuka, 1994, p. 81; Kamm, 1992, p. 47).
I think that under my causal account of liability, however, it is not the case that an infant
could violate a right by landing on you, and it would not be permissible to kill one in self-defense.
And, I think that the true rationale behind our intuition that an adult innocent threat could be
permissibly killed in self-defense is not, as Otsuka posits, that they are “capable of taking
precautions” against being in an inappropriate location (Otsuka, 1994, p. 81). This, while
explaining why we would not find it permissible to kill an infant threat, would obviously not cover
all of the instances in which we may permissibly kill an innocent threat in self-defense. Rather, I
think that our intuitions about adult innocent threats comes from the fact that they are moral agents
and as such must take some moral ownership over their body – and this explains why we would
not say that things (like stones) and mere moral patients (like infants) cannot violate a right of
yours even if their causal role in your death would be sufficient to if they were in fact moral agents.
As another example of a situation in which an innocent threat stands to a causal role in
someone’s death can be found in a case that I will call Human Bludgeon. In this case, a villain

26
kidnaps the future innocent threat and uses some poison to paralyze them. Then, they spring upon
a victim while wielding the paralyzed innocent threat like some sort of polearm, and use the
human-turned-weapon to attempt to beat the victim to death. The victim in this case has a sword
which they may use to slice up the innocent threat and render the villain weapon-less, but this
would obviously kill the innocent threat. In this case as well, I believe that the victim’s killing the
innocent threat in self-defense would be permissible. Even though the innocent threat is paralyzed,
and is not willingly serving as a weapon to be used to bludgeon the victim, their body stands to
play an insensitive causal role in the death of the victim insofar as it is the direct means with which
the victim will be killed, and any change in circumstance would have to be massive to make this
not be the case. This is a case in which I believe the killing of the innocent threat will be harder to
swallow than Falling Man for many, but I think that when it comes down to it our intuitions would
tell us that it is permissible for us to take such an action to save ourselves. I think that it is right
that we have such difficulty in reaching this conclusion, though, and that it is fantasy to think that
every decision our intuitions guide us to make will be one that we have no difficulty reaching.
Moving on to innocent aggressors, Otsuka and others defending the moral responsibility
account of liability to self-defensive harm would say that they have the same morally relevant
properties as innocent threats (being a lack of responsibility for their movements or actions), and
as morally innocent threats have the same morally relevant properties as bystanders we may not
kill innocent aggressors either.
13
I think that he and others are correct in their assessment that the
two have the same morally relevant properties, but that due to the causal role that the actions or
movement of innocent threats and aggressors could play in the death of another, these morally
relevant properties are not the same as those of a bystander. Taking causation to be morally relevant
as I do, the only difference between an innocent threat and aggressor is the fact that it is the actions
of an aggressor but the movements of a threat that stand to do lethal harm to a victim. But, given
that both the movements of the threat and the actions of the aggressor are insensitive contributions
in the causal chain leading towards death, this really makes little difference as that causal
contribution renders both of them liable to self-defensive harm on its own without any need for
further specification between actions and movements. An example of such an innocent aggressor

13
See Otsuka, Michael (2016). The moral responsibility account of liability to defensive killing. In Christian Coons
& Michael Weber (eds.), The Ethics of Self-Defense. Oxford, U.K.: Oxford University Press. and Otsuka, Michael
(1994). Killing the Innocent in Self-Defense. Philosophy and Public Affairs 23 (1):74-94. and McMahan, Jeff
(1994). Self-defense and the problem of the innocent attacker. Ethics 104 (2):252-290.

27
can be found in a case that I will call Drug Zombie and is as follows: villain has kidnapped a future
innocent aggressor and injected them with a drug that takes over their nervous system and causes
them to aggressively attack anybody in their vicinity. Upon releasing this drug zombie onto the
streets, they begin to aggressively attack a victim, although doing so not out of their own will
whatsoever. The victim can shoot the drug zombie innocent aggressor, killing them and saving
themselves. I think that here, it is even easier to reach the intuitive judgement that it would be
permissible for the victim to take self-defensive action and kill the innocent aggressor than it is in
cases with innocent threats.

28
Section V: Remaining Questions
I have at this point established both the theoretical grounding for my causal account of
liability to self-defensive harm, and why it meets the challenges that has made similar theories lose
their intuitive appeal in the past, while still maintaining the parts of the theory that allows it to
align with our intuitions in other areas. There are, however still some areas of confusion for my
account which I do feel I need to address – specifically being the issues revolving around
preemption and similar cases, the decisions that must be made in cases raised by Helen Frowe
regarding her novel category of indirect threats, and problems arising from situations of
questionable moral agency and from the vagueness of a line between agent and patient in general.

V.a: Preemption
The first of these is the issue of preemption which I mentioned earlier on in this thesis – or
casual overdetermination in general – which poses a particular problem for accounts of liability to
self-defense which rely on a theory of causation that is counterfactual like my own. The example
for preemption that I gave previously was Two Hitmen, and another simple case of causal
overdetermination is one that I will call Two Murderers, and is as follows: a victim is walking
through a dark intersection at night when he is confronted by two serial murderers. The murderers
are positioned such that they are hidden from each other’s sight by a building, as they are on
perpendicular streets, and each is intent on killing the victim such that if one murderer misses their
shot the other will succeed. The victim just so happens to have two handguns, and could therefore
kill both murderers before they kill them. The problem here is that my causal account would, as it
stands, find such action by the victim impermissible. This is because the victim’s death in this case
is not counterfactually dependent on either of the murderers’ shooting them, and therefore cannot
be said under the counterfactual theory of causation that I have adopted to have actually caused
that death. If one murderer didn’t kill the victim, the other would, so the victim in killing both of
the murderers would be killing two people each of whose not shooting the victim would still see
the victim dying by another bullet. There is a way to try to eliminate this problem by fine-graining
events down to minutiae (with the resulting event not simply being the victim’s death, but the
victim’s death by a specific bullet, going a specific speed, from a specific angle), and I am not sure
of any other adequate solutions to this problem. As far as this one goes, there is certainly a tradeoff
between having to have a more convoluted conception of events and being able to capture our

29
intuition that even a causal role in cases where preemption occurs does count as causation, but I
think that such a tradeoff seems worth making.

V.b: Indirect Threats
There is also the problem posed by Helen Frowe in chapter 1 of her book Defensive Killing,
in which she posits two additional types of persons beyond those which have largely been
discussed in situations of self-defense.
14
These are innocent and malicious obstructors, each of
whom would block a person’s most viable path of escape from a lethal threat. Frowe puts both of
these kinds of people in the new category of “indirect threats” who, as she says, “is a person who
endangers Victim, but is not going to kill Victim”, by their actions, movement, or presence (Frowe
2014, p. 22). While innocent threats and innocent aggressors cause harm to a victim by their
movement and actions respectively, in the case of innocent and malicious obstructors it is their
mere presence that stands to harm a victim by virtue of that presence’s “reduce[s] the valuable
courses of defensive action morally available to Victim” (Frowe 2014, p.33). But, I use the phrase
cause harm here cautiously, because even though I think that Frowe makes a valid move by indirect
threats like obstructors from bystanders (claiming that bystanders increase a victim’s valuable
courses of defensive action while indirect threats decrease them), the question of whether or not it
is permissible to kill obstructors and other indirect threats does not immediately have a clear
answer in principle or intuition.
Take, for example, the following two cases, Fire and Malicious Fire. In Fire, a victim is in
a building that has been set ablaze by a villain seeking to kill the, and their only path of escape is
through a doorway leading to the street. However, a large man has unknowingly blocked this route
of escape by nudging himself in the doorway so tightly that the only way the victim could get past
him, and escape would be to slice him in half with the sword they conveniently have. In Malicious
Fire, the circumstances are exactly the same, except for the fact that the large man in this case is
well aware of the fact that the doorway is the victim’s only path of escape and nudges himself in
the door with the intent of preventing that escape and leaving the victim to be consumed by the
fire. I think that everybody shares the intuitions that it would not be permissible for the victim to
slice open the innocent obstructor in Fire, but that it would certainly be permissible to slice open

14
For more information on this topic, and the origin of the discussion of indirect threats provided in the next few
paragraphs, see Chapter 1 of Frowe, Helen (2014). Defensive Killing. Oxford, UK: Oxford University Press.

30
the malicious obstructor in Malicious Fire. However, the only differentiating factor in these two
cases is intent rather than causation, which I have maintained is not the thing from which liability
to self-defensive harm springs. So, having to decide what to do about innocent and malicious
obstructors purely from a causal standpoint, it seems impossible to hold that the event of one’s
presence in a location (or, maybe, the event of their actions in moving to that location) cannot be
said to play a causal role in another’s death. With this being said, it seems that I must swallow the
permissibility of killing innocent obstructors in order to maintain not only the permissibility of
killing malicious obstructors but also the causal account in general. This move on my part
essentially does away with the idea of “riding roughshod” over bystanders by saying that
individuals who would obstruct movement in such a way are not truly bystanders, and although I
think that it is preferable to saying that we cannot kill malicious obstructors, it does still have some
troubling implications. For instance, what may I say could be done by a victim whose path of
escape from a villain is blocked by a large crowd of bystanders? Could he simply mow them down
with a handy machine gun? I think not, but my only possible route of justifying this response is to
once again lean on normality concerns to perhaps say that standing in a crowd is relatively normal
and the bystanders’ being there is merely a background condition, whereas doing something such
as blocking off the only path of escape from a burning building in the case of even an innocent
obstructor is not.

V.c: Problems from Agency
Yet another problem is one that was first brought to my attention in feedback from Jonathan
Quong in a separate essay of mine. And, while not the exact same as the question posed by him I
think that it still could represent a problem in my causal account of liability to self-defensive
harm.
15
Much of my account turns on the idea that we may only make moral judgments about
those who are moral agents, and this applies even to those who may not be exercising that moral
agency at a particular time. This generally means that moral judgments may be made about the
causal role that an adult’s body may play in a lethal threat to a potential victim at any time. What
if, however, there was an adult man who had been in a coma since early childhood, and therefore

15
The original question posed by Jonathan Quong here was in response to a previous idea of mine about moral
agents possessing some degree of responsibility at all times, and whether or not a man who was in comatose could
be said to have this sort of responsibility.

31
was never conscious at the time where their mental development was sufficient to make them a
true moral agent rather than simply a moral patient. If this man were dropped off of a building by
a villain towards a victim below, would it be morally permissible for the victim to vaporize that
man with a ray gun? I think that the only way my account can remain consistent is to say that it
would not be, as this person is still not a moral agent and moral judgments cannot therefore be
made about their actions, movement, or presence. The problem develops further when we consider
whether our answer to this question would change if the man briefly woke up out of their comatose
before being thrown for a second, a minute, an hour, or a day, or even if they are held in captivity
as a fully adult moral agent for a week. And, while I think that I am saved the trouble of dealing
with this problem for now given the fact that a man waking up out of a comatose that they have
been in since childhood would probably still have the moral faculties of a child, I am sure the
philosophers could arrange it, as they often do, that the man awakens with the faculties of an adult.
The question about the comatose man is clearly a problem in its own right, but it also leads
me to discuss briefly the final issue that I feel needs to be addressed in this section. This is the
obvious vagueness in when I say that a person becomes a moral agent and can be judged on the
actions, movement, and presence of their body. While it is clear that a rock, an infant, and a young
child cannot reasonably be subject to such judgments, and that an adult person can, there is a grey
area between the two where it remains uncertain whether or not someone is a moral patient or an
agent. I do not have a firm answer to this concern, and swallow the fact that it is a problem for my
theory that a firm line here cannot be established (or at least that it remains outside of my abilities
to do so at this time), and in any case where exactly that line is probably depends on mental
faculties that are unique in each person and impossible to perfectly determine. I would hope
however, that situations of self-defense involving individuals who are in this grey area between
moral agent and moral patient are few and far between, and as such do not constitute too large of
a gap in my causal account of liability to self-defensive harm.

32
Conclusion
All told, I don’t claim for this theory to be a complete and perfect dressing of the many
considerations, or even those of liability, at play in situations of self-defense. And, as I have
mentioned throughout this essay, there are many rough edges that I have not had the time to
address here. However, I do think that this thesis gives a renewed case for an account of where
liability to self-defensive harm comes from which seems to have been in disfavor for some time
and labeled as perhaps too draconian and unflinching in regard to our pre-theoretical intuitions. I
have discussed the metaphysical view of causation which underpins my account, the view of
rights-based morality that I take to be the correct one, how the two interact, what a causal
account of liability coming from the connection between the two ultimately gives us as far as
answers, and where some problems may still be present. And, I believe that I have shown how a
causal account of this highly debated issue in the philosophy of self-defense is capable of lining
up with our intuitions in ways that other accounts simply cannot (especially in regards to
innocent threats and aggressors), while being flexible enough when getting down to the
metaphysics of causation to cover some of the areas where it has been seen as weak in the past. I
truly believe that causation is at the root of all of our ascriptions of liability, and hope that others
will make better attempts to show this than mine in the future.

Many thanks to my thesis advisor, Prof. Collis Tahzib. This would not have been possible
without your generous help throughout the process, and I am extremely grateful.

33
Bibliography

Frowe, Helen (2014). Defensive Killing. Oxford, UK: Oxford University Press.

Frowe, Helen and Jonathan Parry, "Self-Defense", The Stanford Encyclopedia of Philosophy
(Summer 2022 Edition), Edward N. Zalta (ed.),
https://plato.stanford.edu/archives/sum2022/entries/self-defense/

Gallow, J. Dmitri, "The Metaphysics of Causation", The Stanford Encyclopedia of Philosophy
(Fall 2022 Edition), Edward N. Zalta & Uri Nodelman (eds.),
https://plato.stanford.edu/archives/fall2022/entries/causation-metaphysics/

Gallow, J. Dmitri, "The Metaphysics of Causation", The Stanford Encyclopedia of Philosophy
(Fall 2022 Edition), Edward N. Zalta & Uri Nodelman (eds.),
https://plato.stanford.edu/archives/fall2022/entries/causation-metaphysics/

Kamm, Frances Myrna (1992). Creation and Abortion: A Study in Moral and Legal Philosophy.
Oxford University Press.

Lewis, David (1973). Causation. Journal of Philosophy 70 (17):556-567.

Lewis, David (1986). Causation. In Tim Crane & Katalin Farkas (eds.), Philosophical Papers Ii.
Oxford University Press. pp. 159-213.

McMahan, Jeff (1994). Self-defense and the problem of the innocent attacker. Ethics 104
(2):252-290.

Menzies, Peter and Helen Beebee, "Counterfactual Theories of Causation", The Stanford
Encyclopedia of Philosophy (Winter 2020 Edition), Edward N. Zalta (ed.),
https://plato.stanford.edu/archives/win2020/entries/causation-counterfactual/

Otsuka, Michael (1994). Killing the Innocent in Self-Defense. Philosophy and Public Affairs 23
(1):74-94.

Otsuka, Michael (2016). The moral responsibility account of liability to defensive killing. In
Christian Coons & Michael Weber (eds.), The Ethics of Self-Defense. Oxford, U.K.:
Oxford University Press.

Tadros, Victor (2011). The ends of harm: the moral foundations of criminal law. New York:
Oxford University Press.

Thomson, Judith Jarvis (1990). The Realm of Rights. Harvard University Press.

Thomson, Judith Jarvis (1991). Self-defense. Philosophy and Public Affairs 20 (4):283-310.

34
Wenar, Leif, "Rights", The Stanford Encyclopedia of Philosophy (Spring 2023 Edition), Edward
N. Zalta & Uri Nodelman (eds.),
https://plato.stanford.edu/archives/spr2023/entries/rights/

Abstract (if available)

Abstract There are many hotly-debated questions within the philosophy of self-defense. These range from those about which self-defensive actions are proportional or necessary, how our theories of self-defense apply to a wartime context, and even which sorts of killing may be more or less morally wrong than others. Further, there is also the question (perhaps the most important in the discipline) of when and why a person may be liable, at bottom, to self-defensive harm. It is my answer to this question that I wish to put forth in this thesis, and it will proceed roughly as follows. After first establishing the counterfactual metaphysics of causation that will be used to determine what makes a person count has having caused something (and more specifically someone's death), I proceed give a version of Hohfeldian claim rights that allows for ascribing rights violations to individuals who stand to harm another. Next, I discuss what the interaction between a person's causal role in events and their duties to other people is that can ultimately lead us to say that they are liable to self-defensive harm, defend this conception of liability against our intuitions, and try to address some questions that may still remain. After all of this is done, I will hopefully have established an account of liability to self-defensive harm based on causation that is more fully fleshed-out and palatable than those furnished in the past.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

On law enforcement officers’ liability to defensive harm

PDF

Responding to harm

PDF

Four stages of blame

PDF

Responsibility and rights: a search for a principled distinction between criminal law and tort law

PDF

Mistaken defense and normative conventions

PDF

On intimacy: a philosophical account

PDF

Toward a more perfect liberalism: perfectionism in Kantian political philosophy

PDF

Behavioral and neural correlates of core-self and autobiographical-self processes

PDF

An account of the normative structure of human rights

PDF

Habits and Friction: Actual and Perceived Effects on Behavior

PDF

Getting our act(s) together: the normativity of diachronic choice

PDF

Links between discrimination and sleep difficulties: romantic relationships as risk and resilience factors

PDF

Characterizing and developing E. coli Type I-E CRISPR adaptation as a DNA recording and genome engineering tool

PDF

Aggregating complaints

PDF

Defining novel molecular mechanisms to restore balance of a stress response homeostat across spatial and temporal boundaries

PDF

Units of agency in ethics

PDF

Mark making: an exploration of the meaning of marking and self

PDF

What do my (online) friends think? A topic modeling approach to patterns of response to self-injurious behaviors on Reddit

PDF

The virtue of reasonableness: on the normative epistemic implications of public reason liberalism

PDF

The effects of dual coding theory on memory retention

Asset Metadata

Creator Cameron, Reeves Davis (author)

Core Title The causal account of liability to self-defensive harm reconsidered

School College of Letters, Arts and Sciences

Degree Master of Arts

Degree Program Philosophy and Law

Degree Conferral Date 2023-08

Publication Date 05/30/2023

Defense Date 05/10/2023

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag causation,ethics,legal philosophy,liability,OAI-PMH Harvest,Philosophy,rights,self-defense

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Quong, Jonathan (committee chair), Prieto, Marcela (committee member), Tahzib, Collis (committee member)

Creator Email rdcamero@usc.edu,reeves.cameron@att.net

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113138309

Unique identifier UC113138309

Identifier etd-CameronRee-11897.pdf (filename)

Legacy Identifier etd-CameronRee-11897

Document Type Thesis

Format theses (aat)

Rights Cameron, Reeves Davis

Internet Media Type application/pdf

Type texts

Source 20230530-usctheses-batch-1049 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu