Free Will

1
Free Will

• PHIL30088
• Markus Schlosser
• University of Bristol
• markus.schlosser_at_bristol.ac.uk

2
Lecture 7

• Libertarianism Overview and a Modest Version
• Compatibilist Options
• New Compatibilism
• Version 1 Hierarchical Theory of Free Will

3
Two Kinds of Libertarianism

• Event-causal libertarianism
• Strengths
• Based on widely accepted view of causation
• Compatible with a broadly naturalistic view of
  human nature
• Rich account of the self
• Weaknesses
• Cannot account for control over open alternatives
  due to residual degrees of chance (indeterminism)
• Does not show that agents have more control than
  agents in a deterministic world (mere openness of
  future)

4
Two Kinds of Libertarianism

• Agent-causal libertarianism
• Strengths
• Makes possible what appeared to be impossible
  choices are undetermined (by events) and
  determined (by agents)
• Appears to do justice to the phenomenology of
  choice it is up to agents to choose between open
  alternatives
• Weaknesses
• The notion of causation by a substance is very
  controversial
• It does not explain how free will works
• It cannot easily be reconciled with a
  naturalistic view of human nature

5
A Third Possibility

• Modest libertarianism
• Even if agents do not have more control than
  agents in a deterministic world, there is still
  something to be said against compatibilism and in
  favour of libertarianism
• If determinism is true, then alternatives are not
  genuinely open to us. We would do otherwise, only
  if the past is different
• If determinism is false, then the future can be
  genuinely open to us, given the same past
• This gives us independence from the past,
  something that cannot be had by compatibilism

6
Components of Free Will

• (a) Absence from interference(b) Ability to do
  otherwise(c) Being an ultimate source or origin
• Condition (c) is controversial even among
  libertarians. There is disagreement about how it
  is to be understood. Some (Van Inwagen, for
  instance) do not explicitly require (c)
• (c) is very controversial in the overall debate.
  Most compatibilists deny that being an ultimate
  source is necessary for free will and moral
  responsibility. Their views are more firmly
  grounded in naturalistic views of human nature
• Nothing in this world is an origin or uncaused
  cause. The idea that we are ultimate source is
  nothing more than metaphysical megalomania
  (John M. Fischer)

7
Components of Free Will

• (b) Ability to do otherwise
• The consequence argument says that (b) is
  incompatible with determinism
• Compatibilists can use the conditional analysis
  of (b) to show that the consequence argument is
  invalid
• This does not trouble incompatibilists because,
  firstly, this move appears to beg the question
• Secondly, it has been argued that the conditional
  analysis is subject to counterexamples

8
Components of Free Will

• S is able to do otherwise if and only ifS would
  have done otherwise, had S wanted to
• Incompatibilist have put forward counterexamples
• Assume that Peter is incapable (psychologically)
  to want to pick up a spider. The sight (or mere
  image) of a spider gives him the shivers he
  cannot even bring himself to want to pick it up
• Consider a case where Peter encountered a spider.
  Peter did not go near it, and it seems clear that
  he was unable to pick it up
• But it may be true that he would have picked it
  up, if he had wanted to
• Because, had he wanted to, he would not have had
  the mentioned psychological condition
• Hence, the conditional analysis gets it wrong

9
New Compatibilism

• Some compatibilists tried to mend the conditional
  analysis
• Most have given up on this, pursing alternative
  compatibilist options
• Traditional compatibilism is a conjunction of
• (a) Absence from interference, and(b) Ability to
  do otherwise
• All versions of new compatibilism give up on
  (b)
• The first version we will consider can be
  understood as a refinement of condition (a)

10
New Compatibilism

• Condition (a) has itself two components
• (1) The ability to do what one wants to do
• (2) Absence from various forms of interference
  (coercion, force, manipulation etc.)
• This goes back to Hobbes. One main idea is that
  what we call the will is just the sum of what
  we want (the sum of our desires)
• There is one obvious problem with this cases
  where one cannot help doing what they desire to.
  That is, compulsive behaviour that stems from
  irresistible desires or urges
• Conditions (1) and (2) can be satisfied, but the
  agent does not seem to be free, nor responsible

11
New Compatibilism

• The will is not simply the sum of our desires,
  because we can be driven to do something by our
  desires
• But we do not have to conclude that the will must
  be something distinct from our desires. We can
  think of the will, more narrowly, as a collection
  of some of our desires
• Which desires form (are part of) the will?
• The characterisation implicit in the problem is
  negative irresistible, compulsive desires and
  urges are not part of the will
• Harry Frankfurt has suggested a positive
  characterisation

12
Hierarchical Models of Free Will

• The basic idea is this a desire is part of your
  will just in case you can identify with the
  desire, if you endorse the desire as yours
• Identification (endorsement) is construed in
  terms of higher-order desires
• To identify with a desire D is to have a
  higher-order desire D to have D and to be moved
  to action by D
• Frankfurt calls second-order desires to be
  motivated by a certain first-order desire
  second-order volitions
• If you identify with D, then D is part of your
  will and the resulting actions are free (given
  that there are no external interference)
• Frankfurt uses various metaphors to describe
  cases where the agent is motivated by a desire
  that he or she cannot identify with. The desires
  is alien, and the agent is a bystander who is
  not really active but passively driven by a
  desire

13
Hierarchical Models of Free Will

• It is an hierarchical theory there is a
  hierarchy of desires. Frankfurt thought that in
  all ordinary cases there would be no more than
  two levels first-order desires, some of which
  form the will, and second-order desire to be
  moved by certain first-order desires
• Higher-order desires are not a special kind of
  desire they are just more desires. What
  separates them from first-order desires is their
  content, rather than their nature as desires
• A first-order desire is directed towards action
  a desire to do something
• A second-order desire is directed towards a
  first-order desire a desire to have certain
  first-order desire or a desire to be moved to
  action by a certain first-order desire

14
Hierarchical Models of Free Will

• The theory is sometimes also called a mesh
  account of free will. What is crucial to having
  free will is an inner harmony, accordance or mesh
  between desires. Frankfurt was also talking of
  wholehearted action
• Note the contrast to Kanes view, where free will
  springs from cases of inner conflict
• Also, sometimes it is called a time-slice
  theory Free will consists in features of the
  agent at the time of acting
• It neglects historical aspects What matters is
  the mesh between first-order and second-order
  desires, not how desires have been acquired

15
Some Problems for the Hierarchical Theory

• 1. The regress problem
• Second-order desires are just desires
• Frankfurt says that if an agent does not want to
  be motivated by a first-order desire, the agent
  is alienated from that desire
• What about second-order desires?
• Given that they are only desires, the same
  problem may arise for them the agent may be
  alienated from a second-order desire
• So, the fact that one has a second-order desire
  towards a first-order desire does not guarantee
  that one can identify with that desire
• One might refer to third-order desires
• But it seems clear that this leads to a regress
  of ever higher-order desires

16
Some Problems for the Hierarchical Theory

• 2. Manipulation worries
• This problem stems from the mesh and
  time-slice nature of the view free will
  consists in an inner harmony at the time of
  action
• But just as first-order desires may be the result
  of manipulation, brainwashing, or the like,
  second-order desires might be the result of
  manipulation as well
• So, the inner harmony between desires might be
  the outcome of manipulation, brainwashing, or the
  like. In that case, the agent is neither free nor
  responsible
• Hence, inner harmony alone does not guarantee
  freedom and responsibility

17
Some Problems for the Hierarchical Theory

• 3. Moral responsibility
• It is commonly thought that there is a close
  connection between free will and moral
  responsibility
• The problem is that identification with desires
  and wholehearted action does not seem to be
  necessary for moral responsibility, nor
  sufficient
• Consider a case of weak-willed action Suppose
  Susan owes Anne 50. Knowing Anne, Susan suspects
  that Anne has forgotten about it. Susan wants to
  pay her back, but she is also tempted to take
  advantage. Giving into this temptation, Susan
  acts on a desire that she does not identify with
  (she has a second-order desire not to give into
  temptation)
• Is Susan morally responsible? It seems so. The
  fact that the action was not wholehearted does
  not absolve from responsibility identification
  and wholeheartedness are not necessary for moral
  responsibility

18
Some Problems for the Hierarchical Theory

• 3. Moral responsibility Cont.
• Are identification and wholeheartedness
  sufficient for moral responsibility?
• Manipulation worries suggest that they are not
• Suppose Robert has been tricked into a serious
  drug addiction. The condition affects his
  first-order and his second-order desires. He
  develops a strong first-order desires for the
  drug, and his fully identifies with those
  first-order desires
• Roberts actions satisfy the internal mesh
  condition of the theory, but it seems clear that
  he is not responsible for the actions that result
  from his desire for the drug

19
Two Solutions to the Regress Problem

• Frankfurt tried to solve the regress problem. But
  it is generally thought that his own attempts are
  unsuccessful
• Nevertheless, the theory was very influential,
  and many attempts have been made to solve the
  internal problems with the view
• The regress problem seems to stem from the fact
  that Frankfurt refers only to desires. Many have
  suggested to incorporate other higher-order
  attitudes into the model
• Michael Bratman, for instance, suggests to
  replace second-order desires with self-governing
  policies higher-order attitudes that specify
  which desires to treat as a justifying reason for
  action

20
Two Solutions to the Regress Problem

• You might, for instance, have the policy not to
  act out of revenge this policy says, basically,
  that a desire for revenge does not give you a
  reason to seek revenge
• (One might compare this to Kants maxims of the
  will they are subjective principles of choice
  and action)
• These policies are better suited for the job than
  second-order volitions, because they are stable
  and considered attitudes
• They speak for the agent with authority, they
  organize the agents actions across time
  (long-term principles of choice)
• Human action is not a time-slice phenomenon. It
  is organized across time. Self-governing policies
  speak for the agent, because they organize
  actions across time

21
Two Solutions to the Regress Problem

• David Velleman has proposed a different kind of
  solution
• We need to find mental states that speak for
  the agent, and he also emphasizes that role of
  reasons (as opposed to the role of desires)
• But Velleman thinks that referring to mental
  states within the hierarchy cannot solve the
  problem
• Identification with desires is driven by
  reflection you identify with a desire as a
  result of reflecting on it
• Potentially, reflection may question any kind of
  mental attitude. No mental state is immune to
  being rejected by reflection
• In order to find a mental state that truly speaks
  for the agent, we need to look elsewhere

22
Two Solutions to the Regress Problem

• Velleman thinks that there is a desire that
  drives the very process of reflection the desire
  to act in accordance with reasons
• He thinks that this desire truly speaks for the
  agent, because it is the very motive that drives
  reflection and deliberation
• Velleman also thinks that this desire joins
  forces with conflicting first-order desires
• Say, an agent S has a desire D to do A, and S
  thinks that she has reason to do A. But S is also
  tempted to do something else. Then Ss desire D
  to act in accordance with reasons may tip the
  balance

23
What about Free Will and Moral Responsibility?

• Surprisingly, Bratman and Velleman do not claim
  to have provided an account of free will
• Velleman says that he is interested in human
  action par excellence, Bratman seeks to account
  for full-blooded human action
• And they make no explicit claims about moral
  responsibility
• Plausibly, action par excellence is sufficient
  for being responsible
• But it seems clear that it is not necessary for
  moral responsibility
• We can be responsible for weak-willed actions
• And sometimes we are responsible precisely
  because we did not do what we had reason to do

24
What about Free Will and Moral Responsibility?

• This is not a problem for Bratmans and
  Vellemans views, as they make no claims
  concerning responsibility
• But what about the hierarchical approach to free
  will?
• Free will is thought to be closely connected to
  moral responsibility
• In particular, free will is thought to be
  necessary for moral responsibility
• If we make this a condition on free will, then
  the hierarchical theory fails Identification and
  wholeheartedness are not necessary (nor are
  actions par excellence and full-blooded actions)

25
What about Free Will and Moral Responsibility?

• Why think that free will is necessary for moral
  responsibility?
• Perhaps free will is a special kind of freedom or
  control that is not necessarily required for
  moral agency, but that is valuable in itself?
• This question brings us back to the very starting
  point what is free will? Whats in the concept
  of free will?
• (a) Absence from interference(b) Ability to do
  otherwise(c) Being an ultimate source or
  origin((d) Requirement for moral responsibility)
• The hierarchical theory is a refined version of
  (a) absence from external and internal
  disturbance (conflict within the self)
• It does not satisfy (b), (c) and (d)