Free Will
Free Will
13
1. Introduction
Incompatibilism about determinism and free will is typically formulated as
the claim that free will is impossible in a deterministic world:every possible
world where determinism obtains is a world without free will. Van Inwagens
consequence argument (van Inwagen, 1983, ch. 3)is often taken to support
that claim. Recently, however, it has been suggested that the consequence
argument does not support incompatibilism, since it rests on at least one
contingent assumptionsuch as the assumption that there is a past, or the
assumption that there is a past without any human beings (see Warfield,
2000, and Campbell, 2007).1 If the argument indeed relies on an assumption
that is at best contingently true, then it cannot be used to show that free will is
absent in every deterministic world. Furthermore, the objection appears to be
more far-reaching than previously noticed:a similar objection can be raised
against other key arguments for incompatibilism (see Bailey, 2012). In what
follows Irefer to objections of this kind as contingency objections.
Proponents of contingency objections believe that the fact that the rel-
evant incompatibilist arguments are subject to those objections shows that
the arguments are flawed in some important way.2 Iargue that the lesson we
should draw from that fact is significantly different. Contingency objections
dont show that there is any deep flaw in those arguments. What they show
is that we have been confused about the nature of the problem that motivates
incompatibilism. The so-called problem of determinism and free will is not
the problem of determinism and free will. The first thing to do, then, is to
clarify the nature of the problem.
The plan for the chapter is this. In section 2, I briefly review the conse-
quence argument for incompatibilism, I explain the contingency objection
against it, and Iargue that what the objection shows is that the problem of
determinism and free will has been misdiagnosed. In section 3, Iexplain what
Itake the real nature of the problem to be. In section 4, Idraw some impli-
cations for different aspects of the free will debate:the role of causation in
the formulation of the problem of determinism and free will, the distinction
between leeway and source incompatibilism, and the fate of incompatibilist
arguments that appeal to an ultimacy condition on freedom.
versa, the state of the world at T-last and the laws fix the state of the world
at T-first. This is so even if there is no time prior to T-first that determines
the state of the world at T-first. Of course, this is only the case because
there is no time prior to T-first. But it is still true of W that the state of the
world at T-first is determined by the state of the world at any other exist-
ing time; hence W is deterministic in the traditional sense. As Campbell
points out, the consequence argument cannot be used to show that Adam
lacks free will in W for there is no past time at which Adam doesnt exist,
so the argument cannot get off the ground.
Now, should we expect an incompatibilist argument such as the conse-
quence argument to show that Adam doesnt have free will? Isubmit that we
shouldnt. If we had lived in such a world and each of us had been like Adam,
then its likely that the problem of determinism and free will wouldnt even
have come up for us. For what could have been the motivation for thinking
that determinism is a threat to our free will in that world?
Note, first of all, that if Adam can be free at T-first, then the fact that his
acts at later times are determined by the events at T-first is not a threat to his
free will at those later times. For he could have a choice about those later acts
by virtue of having had a choice about the relevant events at T-first.3 But why
would determinism be a reason for thinking that he isnt free at T-first? Given
that Adams world is deterministic, what Adam does at T-first is determined
by the state of the world at, for example, T-last. But again, if he can affect what
happens at T-last by acting at T-first, then this cannot be a reason for thinking
that hes not free at T-first.
Compare:we dont think that the fact that the future determines what we
do in the present is a reason for thinking that we dont have a choice about our
present acts. For we think that we can causally influence the future by acting
in the present. So, even if the future determines the present, this by itself isnt
a reason to think that we dont have a choice about our present acts. This is
why, of course, van Inwagen chose the past instead of the future to build the
consequence argument for incompatibilism.4
Here, then, is an argument that determinism isnt a threat to Adams free
will at T-first (and, thus, at any othertime):
(1) Determinism could only be a reason to think that Adam is not free at
T-first if there were some time T such that the fact that the laws and the
state of the world at T determine what Adam does at T-first is a reason to
think that Adam is not free at T-first.
(2) There isnt any such time. (All other times are after T-first, and Adam can
affect what happens at those times by acting at T-first.)
(3) Therefore, determinism isnt a reason to think that Adam lacks free will
at T-first.
I believe that this argument is sound. It follows that we shouldnt expect the
consequence argument to prove that Adam lacks free will. Of course, there
could be other reasons (apart from determinism) for thinking that Adam
lacks free will at T-first. For example, one could argue that he is not free be-
cause what he does at T-first is not the result of an ordinary process of deliber-
ation (see Brueckner, 2008, and also Speak, 2011). But determinism itself isnt
a problem for Adams free will.5
Campbell would disagree with my assessment. He urges us to consider a
slightly different world that contains, in addition to Adam, an only woman,
Eve, who comes into existence some time after T-first (Campbell, 2007). Eves
situation is different from Adams situation in that there is a time before Eve
(a time when Eve doesnt exist), and thus the consequence argument could be
used to show that none of Eves actions are free. But, Campbell argues, given
that Adam and Eve live in the same world, their acts are equally determined.
Hence, if Eve lacks free will, Adam does too, and the consequence argument
should reflect that. In other words, the fact that the argument cannot be used
to show that Adam lacks free will is a weakness of the argument:it means that
it isnt as general as it should be.
However, Idont think that its a weakness at all. Perhaps both Adam and
Eve lack free will in W. But if so, its for different reasons in each case. In
particular, the argument that Adam is not free at the first moment of his exis-
tence would have to be a very different kind of argument from the argument
that Eve is not free at the first moment of her existence. One could argue that
Eve is not free at the first moment of her existence by pointing out that what
she does then is determined by the laws and the state of the world at T-first,
when she didnt yet exist. But again, if one were to argue that Adam at T-first
is not free because what he does at T-first is determined by the laws and the
state of the world at, for example, T-last, this would be like arguing that we
are not free because the laws and some future state of the world determine our
present acts. This is a bad argument.
Our discussion of the contingency objection to the consequence argument
suggests that the so-called problem of determinism and free will is not the
problem of determinism and free will. The threat to our free will is not just
determinism; it is determinism plus some additional fact: a contingent fact
about us.6 Campbell (2007) complains that this form of weak incompati-
bilism, as he calls this kind of view, is not really incompatibilism at all. But
what the objection helps to bring out, as we have seen, is that it is the only
incompatibilist thesis that is, on reflection, even remotely plausible. We just
hadnt realized this before because we had been confused about the nature of
the problem.7
Now, if the problem of determinism and free will isnt the problem of
determinism and free will, then what is the problem? What is that additional,
contingent fact that, coupled with determinism, creates a threat to our free
will? Iturn to this in the next section.
That is to say, according to (a), the threat to our free will is not just deter-
minism but the fact that, for every time at which we exist, there is a past
time such that the state of the world at that time determines what we do.
This condition isnt met for Adam in W, since Adam doesnt have a past at
T-first. Hence (a) could be used to explain why, while determinism may
contribute to undermine our free will, it doesnt contribute to undermine
Adams free will.
However, (a)wont do. Consider another possible scenario:Eternal Adam
(discussed, e.g., in Finch, 2013). Eternal Adam lives in an eternal deterministic
world and he himself is eternal:he has existed forever, and will exist forever, as
an adult rational being with all the powers and abilities necessary for respon-
sibility. Eternal Adam meets condition (a):for every time at which he exists,
there is a past time. But if determinism is not a threat to Adams free will in W,
it is also not a threat to Eternal Adams free will. For take any act by Eternal
Adam at a time T.Could the fact that his act is determined by later events and
the laws be a reason to think that he is not free when he acts at T? No (for
the same reason as before:he can affect the later events by acting at T). Could
the fact that his act is determined by earlier events and the laws be a reason to
think that he is not free when he acts at T? No, for given that he exists before
T as a fully rational being, he could have a choice about what he does at T by
virtue of having had a choice about the earlier events.
Note that (b)is stronger than (a). Eternal Adam doesnt meet this condition
(he himself exists at all past times), and neither does Adam. Hence, (b)could
be used to explain why, whereas determinism may contribute to undermine
our free will, it doesnt contribute to undermine Adams free will or Eternal
Adams free will.
However, (b) still wont do. Consider a third scenario: Time-Traveler
Adam. Time-Traveler Adam lives in a world where time-travel is physically
possible and he is in fact in possession of a working time machine that could
take him to any time, including any past time. Time-Traveler Adam has a past
when no rational beings exist. Although he could have traveled to those past
times, he didnt in fact travel to those times. (It is important that he didnt, in
fact, travel to any of those times. Otherwise they wouldnt be times when no
rational beings existed!) If determinism isnt a threat to Adams free will or
Eternal Adams free will, it also isnt a threat to Time-Traveler Adams free will.
For even if there are many times at which he doesnt exist and that, together
with the laws, determine what he does, he can still have a choice about what
happens at those times thanks to his time-traveling abilities, which allow him
to travel back to those times and affect the events that take place then.
This suggests that what contributes to the problem of determinism and
free will is not the existence of a time without any rational beings but, instead
That is to say, for each one of us, there is a time that is outside our control, and
this is why determinism is a problem for our free will. Claim (c)draws atten-
tion to an important difference between Time-Traveler Adam (and Eternal
Adam, and Adam) and us:the remote past is clearly outside our control, but
no time is clearly outside Time-Traveler Adams control (and the same goes
for Eternal Adam and Adam).
In what sense does Time-Traveler Adam have control over the past, if he
didnt travel to the past? The answer is:in at least the sense that the past is
within his causal reach. The most natural way to support this claim is to say
that, although he didnt causally influence the past (since he didnt travel to
the past), he could have causally influenced it (because he could have traveled
to the past) and this is enough for the past to be within his causal reach. But
This more complex picture of the world is a certain kind of naturalistic picture
that conceives our acts as natural processes and thus, like any natural process
or mechanism, as having an explanation in terms of other natural processes at
other times. Even if the relations between those processes are not always fully
deterministic, in the case of macroscopic events such as the acts of human
agents, they are for all practical purposes deterministic (the microscopic
indeterminacies, most seem to believe, dont result in significant macroscopic
indeterminacies).
So an important part of the picture of the world that threatens our free-
dom is the quasi-determinism that, according to this view, applies to all
human acts. But arguably, the naturalistic conception of human agency
goes beyond this. Human acts dont just (quasi-deterministically) result
from other natural processes or mechanisms; importantly, those other
natural processes or mechanisms are for the most part (and barring the
possibility of human-operated time-machines!) beyond our control. They
are so extended in space-time that we have no access to them, given the
limited scope of our causal powers. Our finite existence and powers are,
as they say, just a tiny speck in a vast universe that is mostly beyond our
reach, in both the spatial and temporal dimensions. This is as much part
of a humble, unpretentious conception of human agency as the claim that
our acts can be quasi-deterministically explained in terms of other natural
events.
So my tentative suggestion is this. The reason we tended to overlook the
fact that the problem of determinism and free will is not the problem of
determinism and free will is that determinism wasnt our real focus to start
with. Our real focus was a more complex naturalistic picture of the world
that includes, in addition to a quasi-deterministic element (one according
to which all acts have a quasi-deterministic explanation in terms of other
events at other times), an equally important element of lack of control that
characterizes human beings as in some important way disconnected from
the causal mechanisms that result in their acts. Our acts are the (for all prac-
tical purposes, inevitable) result of mechanisms or processes that are com-
pletely beyond our reach. That is how the problem for our freedom arises.
This passage suggests that, according to van Inwagen, the real source of the
problem is not causal determination but determination tout court. The fact
that the problem can be formulated without making reference to causation,
but simply in terms of certain propositions entailing other propositions,
strongly suggests this.
Now, Ihave argued that determination itself is not enough to give rise to
the problem, and that, in particular, an assumption of causal impotence (to-
ward the past events that determine our acts) is also needed. This means that,
even if van Inwagen were right that pure determination, and not causal deter-
mination, is what contributes to the problem, the concept of causation would
still play a central role in its formulation. What plays the relevant role, in this
case, is not the existence of a certain type of cause, but the absence of a cer-
tain type of cause, or of a certain type of causal access. Its a negative role
instead of a positive role (in that sense), but its an important role neverthe-
less. This means that one probably shouldnt try to avoid all questions about
causation when thinking about the problem of determinism and free will, as
van Inwagen recommends we should do.
are beyond our control (events that occurred before we were born) but that
determine how we act (Pereboom, 2001, 3).
Once one notes that the source intuition concerns actual causal histories
it becomes quite evident that any source incompatibilist argument will have
to rely on what appears to be a contingent assumption:the assumption that
our acts in fact have such a history (in particular, one that can be traced back
to remote times when no human beings existed and that are outside of our
control).10 It is perhaps less obvious in the case of the consequence argument
(or any argument that draws on the leeway intuition instead of the source
intuition) that the argument relies on a contingent assumption about actual
histories. But, as we have seen, ultimately both kinds of arguments do. So the
belief that the difference between the two incompatibilist ideas is simply the
focus on counterfactual versus actual facts turns out to be misguided. Both
incompatibilist ideas rely on a contingent fact about actual histories. And
in both cases, the fact on which each relies is a very similar fact, if not the
same one:its the fact that the causal history is one that traces back to factors
beyond our control, or some closely related fact.
To be sure, there is still a difference between the two incompatibilist routes.
According to the leeway-incompatibilist intuition, the reason the actual causal
history of our acts is a problem for our free will is that given that it obtains,
we couldnt have acted otherwise (since we dont have a choice about that his-
tory). On the other hand, according to the source-incompatibilist intuition,
the reason the actual causal history of our acts is a problem for our free will is
not that it rules out alternative possibilities; its, rather, some other thought
about sourcehood, ultimacy, or origination. But still, in both cases part of the
source of the problem is a certain kind of actual history:a causal history that
includes factors that are beyond our control. As a result, leeway and source
incompatibilism end up being much closer than they are usually taken to be.
All roads to incompatibilism share a common starting point.
We have seen that rethinking the nature of the problem of determinism
and free will has important implications for the distinction between leeway
and source incompatibilism. A related implication concerns the nature of
leeway and source incompatibilism themselves. Leeway incompatibilism is
typically regarded as the claim that determinism rules out freedom and re-
sponsibility because it rules out the ability to do otherwise. But we can see
now that this isnt what we should take leeway incompatibilism to be. Rather,
we should take it to be the claim that determination by factors beyond our con-
trol rules out freedom and responsibility, because it rules out the ability to
do otherwise. Similarly for source incompatibilism:we shouldnt take source
incompatibilism to be the thesis that determinism rules out freedom and re-
sponsibility given the type of causes our acts would have if determinism were
true. Rather, we should take it to be the thesis that determination by factors be-
yond our control rules out freedom and responsibility, given the types of causes
our acts would have if they were determined by factors beyond our control.
Pereboom, one of the main proponents of source incompatibilism, charac-
terizes what he calls the core incompatibilist claim (which he takes to be the
source-incompatibilist idea) as follows:
Here Pereboom is implying that being the ultimate sources of our actions
requires a certain kind of indeterministic causal history. He then goes on
to examine the kind of indeterministic actual causal history that would be
required for our acts to be free and argues that an agent-causal version of lib-
ertarianism would do the trick (unfortunately, he thinks, this is a version of
libertarianism that we have good reason to believe is false). But, as we have
seen, being agent-causes of our actions is not the only way in which we could
have control of the sources of our actions. Agent-causation is not in fact
needed, and neither is any form of indeterminism; being able to travel back in
time (or some other way of causally influencing the past) would be enough.11
It follows that its misguided to characterize the main source-incompatibilist
condition as one that requires indeterminism. Again, in this respect leeway
and source incompatibilism seem to be in the same boat. For, at least in princi-
ple, it seems that time-travelers would be able to meetall the main leeway and
source incompatibilist conditions, even if they lived in a deterministic world.
Clearly, the fate of this type of argument will depend on what is meant by
ultimacy. In what follows Iargue that given the way that source incompati-
bilists tend to understand this concept, the earlier results of this chapter show
that ultimacy arguments are simply question-begging.
Incompatibilists have offered different renderings of the ultimacy require-
ment. Smilansky formulates it in the followingway:
McKenna helpfully offers the following interpretation (or perhaps its more
like a friendly variation) of Smilanskys principle:
Note that all these different accounts of the notion of ultimacy seem to
entail the following minimal condition:
To see that this argument blatantly begs the question against the compatibil-
ist, insert what Ihave argued is a minimal interpretation of the ultimacy re-
quirement, Ultimacy (Causal Access), in the two premises. This results in the
following argument:
(U2**) We cannot be free unless we have causal access to all of the actual
sufficient sources of our choices.
(U3**) Therefore, if our choices are determined by causes outside of our
causal reach, we cannot be free.
Notice that the first premise has now turned trivial (claiming that the causes
that determine our choices are outside our causal reach is just claiming that we
dont have causal access to those actual sufficient sources of our choices). All
the weight rests on the second premise, which is basically a mere restatement of
incompatibilism, or, more precisely, of the thesis that Ihave argued we should
take incompatibilism to be:the claim that the lack of causal access to (some of )
the deterministic causes of our choices is sufficient to undermine our freedom.
Ultimacy arguments, then, are blatantly question-begging. If we couldnt
see this clearly, it is because we were confused about the nature of the problem
of determinism and free will.14 In conclusion, getting clear on the nature of
the problem has mixed results for incompatibilism. On the one hand, as
explained, it helps incompatibilist arguments avoid the contingency objec-
tion. But on the other hand, as we have just seen, it turns some incompatibilist
arguments into obvious non-starters.
Acknowledgments
I presented earlier versions of sections 2 and 3 of this chapter at a workshop
on Joe Campbells work on free will at the University of Arizona in March
of 2012. Iam grateful to the participants at the workshop, and especially Joe
Campbell, Michael McKenna, and Brandon Warmke, for their helpful com-
ments. Iam also grateful to Juan Comesaa, Carl Hoefer, Michael McKenna,
and Derk Pereboom for comments on early drafts. Part of the research result-
ing in this paper was made possible through the support of a grant from the
John Templeton Foundation. The opinions expressed in this publication
are those of the author and do not necessarily reflect the views of the John
Templeton Foundation. Iam particularly grateful to Al Mele, director of the
Big Questions in Free Will project, for the support.
N OT E S
2. Even Bailey, who thinks that the cost involved in retreating to a weaker form of
incompatibilism (of the kind discussed later in the text) might not be too large, is
committed to the idea that there is a cost involved. As a result, he proposes a new
incompatibilist argument that attempts to escape the objection (and so do other
philosophers like Warfield, 2000, and Finch, 2013).
3. Bailey (2012) makes a similar point.
4. Note that van Inwagen himself employs a definition of determinism according to
which the future determines the present and the past just as much as the present
and the past determine the future (see van Inwagen, 1983, ch. 7). Hoefer (2002)
has pointed out that if determinism is a purely logical thesis in this sense, then
our actions place some constraints on the past. Whereas Hoefer saw this as the
basis for a potential solution to the problem of determinism and free will, Isee it as
a reason to think that the problem itself has not been adequately formulated.
5. Finch (2013) has recently claimed that determinism alone is a threat to Adams free
will in W.She argues for a view of free will according to which doing Afreely at T
requires the existence of an earlier time T' at which it was possible for the agent to
do Aat T and also to refrain from A-ing at T.If this view of freedom were correct,
it would follow that Adam doesnt act freely at T-first. But again, the reason he
wouldnt be free is not that determinism obtains in W, but that in W there is no
time prior to T-first. So determinism alone isnt a threat to Adams free will, des-
pite what Finch suggests.
6. Note that a contingent fact about us could be ambiguous between two readings.
On the first reading, it implies that we (humans) could have been like Adam and
could have lived in his world, W.But there is a second, weaker reading, which
only implies that some rational being with similar capacities for free will (although
not necessarily human) could have lived in W.According to this weaker reading,
whats contingent is only that the type of rational being that we are dealing with
when we formulate the consequence argument for human beings is unlike Adams
type. Importantly, the weaker reading is all thats required to make the case that
determinism per se does not threaten free will.
7. Shabo (2011) argues for a similar claim. In response to a critic, van Inwagen
explains that he never intended the premises of the consequence argument to be
necessary truths (see van Inwagen, 1983, 7071). But this is only because he for-
mulated the argument in terms of a specific human being, J, who (he was assum-
ing) in fact performed a certain act at T, and then argued that, given the actual
past, J couldnt have failed to perform that act at T.Van Inwagen notes that, of
course, there are possible worlds where the past is different (there are even pos-
sible worlds where J doesnt exist at all). So, he argues, some of the premises of
the argument are only contingently true. Now, our discussion of the contingency
objection shows that some premises in the argument are only contingently true in
a stronger (and more interesting) sense. If, as we have seen, there are some possible
rational beings whose free will is not at all threatened by determinism, then the
consequence argument can only work for us to the extent that we are not beings of
that kind.
8. This view is suggested by Bailey (2012). It is also suggested by van Inwagens formu-
lation of the consequence argument (he formulates it in terms of T0, a time before
the birth of the relevant human beingthe arbitrary one the argument is about;
see van Inwagen, 1983, 6970).
9. Pereboom uses the label causal-history incompatibilism instead of source
incompatibilism. Similar sourcehood conditions are formulated in Kane (1996,
section I.5), Mele (1995, 211), McKenna (2010, 437), and Shabo (2010, 369).
10. Bailey (2012) draws attention to the fact that a main kind of source-incompatibilist
argument (an argument that appeals to the notion of ultimacy, discussed later in
the text) relies on such a contingent assumption. Interestingly, whereas he sees this
as a reason to think that this argument, too, falls prey to the contingency objec-
tion, Iargue that it is, instead, a reason to think that ultimacy arguments simply
beg the question against the compatibilist.
11. As explained earlier, an eternal life as fully rational beings would do the trick too.
As for Campbells Adam, the inhabitant of world W, Perebooms view seems to
entail that since there is no time before T-first, this means that Adam is not free
at T-first. For when he acts at that time, he is not in control of the sources of his
act, since his act doesnt have any sources (it is not caused by anything). (On this
point, see Pereboom, 2001, 5.) Again, however, note that the reason Adam is not
free has nothing to do with determinism.
12. Kane (1996), Stump (1996), Smilansky (2000), Zagzebski (2000), and Shabo
(2010) seem to endorse arguments of this kind. The first explicit formulation of the
ultimacy argument in a way that resembles my formulation appears in McKenna
(2001). See also Strawson (1994) for a stronger version of an ultimacy argument
that attempts to show that freedom and responsibility are impossible, regardless of
whether determinism is true. Pereboom (2001) endorses the ultimacy condition
on freedom, but makes a case for it on the basis of a different argument, an argu-
ment from manipulation.
13. Shabo seems to understand ultimacy in a similar way (see Shabo, 2010, 369).
14. In Sartorio (manuscript), Iargue that similar considerations threaten to under-
mine yet another type of source-incompatibilist argument, direct arguments
(or, as they are sometimes called, transfer of non-responsibility arguments), at
least under the assumption that Frankfurt cases show that responsibility is not
grounded in alternative possibilities (an assumption that source incompatibilists
typically grant).
REFERENCES