Carroll J.W., Markosian N. An Introduction to Metaphysics

Подождите немного. Документ загружается.

Freedom and determinism 49

Meanwhile, in more recent times, the determinist idea has been

understood in terms of laws of nature and physically possible futures. The

concept of a law of nature is discussed at length in Chapter 4. But it is

worth saying a bit more here about physical possibility and physical neces-

sity. The alethic modalities are the notions of possibility and necessity that

are connected with the laws of logic. (The word ‘alethic’ means having to

do with truth.) When we talk of something being logically possible or logic-

ally necessary, we are talking about alethic modalities. The nomic modalities,

on the other hand, are the notions of possibility and necessity that are con-

nected with the laws of nature. (‘Nomic’ means having to do with laws.)

To say that something is physically possible is to say that it is allowed by the

laws of nature. So a physically possible future, relative to a certain time, is

a way that things could continue from that time that is permitted by the

laws of nature. (And a physically impossible future is one that is not per-

mitted by the laws.) To say that something is physically necessary, however, is

to say, roughly, that it is required by the laws of nature.

In addition to the notion of physical necessity, our characterization of

determinism in terms of laws of nature will depend on several key pre-

suppositions. First, we will assume that it’s possible for a proposition to

“express” the state of the world at a particular time, and, in fact, that for

every time, there is a proposition that expresses the state of the world at

that time. In addition, our discussion will be much simpler if we agree to

think of propositions as “tenseless” items that have ﬁxed truth values (like

the proposition that Juan is standing at noon on June 1, 2525), rather than

thinking of propositions as “tensed” items that can change their truth

values over time (like the proposition that it will be the case in 515 years

that Juan is standing).

(It is also possible to formulate determinism in a

More precisely, to say that some event, e, is physically necessary as of time t is to say

that conditions at t are such that e occurs in every physically possible future rela-

tive to t. In the text we sometimes speak of an event (or set of conditions) “making”

another event physically necessary, which we take to be a shorter and simpler way

of saying that, given that the ﬁrst event occurs (or that the relevant conditions

obtain), the second event is physically necessary at the time of that ﬁrst event (or

those conditions).

Why “physical” possibility and necessity? Because the presumption is that all the laws

of nature are reducible to the laws of physics. But “nomic” (or “lawful”) possibility and

necessity would work just as well, if we wanted to drop that presumption.

For more on the issue of tensed vs. tenseless propositions, see Chapter 7.

An Introduction to Metaphysics50

way that is more friendly to the tensed way of thinking about proposi-

tions, but doing so would be a bit more complicated.)

The intuitive idea, then, is that at each instant, there is exactly one phys-

ically possible future. In other words, for any time, t, given the way the

world is at t, and given the laws of nature, there is only one way that things

could continue. Here is a way to formulate this thesis that is a variation on

van Inwagen’s formulation from page 65 of his An Essay on Free Will:

Determinism

For any propositions, p

and p

, and times, t

and t

, such that p

expresses

the state of the world at t

and p

expresses the state of the world at t

, the

conjunction of p

with the laws of nature entails p

This formulation has the consequence that the present always determines

a unique future. But it also has the consequence that the present always

determines a unique past. For there is nothing in the formulation requir-

ing that time t

be later than time t

. It could be earlier. This is a feature of

van Inwagen’s formulation as well, and it is a feature that he welcomes.

Nevertheless, van Inwagen points out that it would be easy to reformulate

determinism so as to avoid the relevant commitment to “backwards deter-

minism.” We would just have to restrict the thesis in a way that requires

that t

be earlier than t

, as follows:

Future-Oriented Determinism

For any propositions, p

and p

, and times, t

and t

, such that t

is earlier

than t

, p

expresses the state of the world at t

, and p

expresses the

state of the world at t

, the conjunction of p

with the laws of nature

entails p

It is important to appreciate how determinism is different from

Universal Causation. determinism is formulated in terms of laws of

nature (and hence in terms of physical necessity) rather than in terms of

caus ation. This is important because it is at least arguably possible to have

causation without having physical necessity. That is, a particular event can

be caused without being made physically necessary. (For example, suppose

there is a bomb with an indeterministic detonation device. It is a genu-

inely indeterministic matter, as far as the laws of nature are concerned,

whether Tim pushing the button will result in the bomb being detonated;

but his pushing the button ensures that there is, let’s say, a “nomic” prob-

ability of 0.99 [on a scale from 0 to 1] that the bomb will detonate. Now

Freedom and determinism 51

suppose Tim pushes the button and the bomb ends up detonating. Then

we will want to say that Tim pushing the button caused the explosion,

but that the explosion was not made physically necessary by Tim pushing

the button.) Meanwhile, you presumably cannot have physical necessity

without having causation. In other words, if it is physically necessary that

a particular event occurs, then that event must be caused. The upshot

is that determinism (as formulated here) captures a stronger version of

the general idea that the past determines the future than is captured by

Universal Causation.

Here’s an important question for students of The Problem of Freedom and

determinism: what status should we give to determinism? Traditionally,

many philosophers (including leibniz and kant) have thought that some-

thing like determinism is an a priori, necessary truth. But nowadays most

philosophers tend to think that if anything like determinism is true, then

it is contingently true, and also (relatedly) that whether determinism is

true is an empirical (rather than an a priori) matter. We tend to agree

with most contemporary philosophers on these matters (that if determin-

ism is true then it is contingently true, and that whether determinism is

true is an empirical matter), but we will not argue for these claims here.

Instead, we now turn to a consideration of the popular view known as

Hard Determinism.

3.4 Hard Determinism

Suppose that for whatever reason you think determinism is true. Maybe

you think (like leibniz and kant) that it’s a conceptual truth. or maybe

you think it’s just an empirical truth that is supported by our best evi-

dence. Now consider some action of yours, and focus on the fact that

(if determinism is true) it was physically necessary a million years ago

that you would perform that action.

It begins to look like you couldn’t

be responsible for that action, since, on our supposition, it was “pre-

determined” by past conditions and the laws of nature.

For more on these issues, see 2.6 above.

recall that this is a somewhat loose way of saying that a million years ago, condi-

tions were such that your eventual performance of the action in question was already

required by the laws of nature.

An Introduction to Metaphysics52

Here is a traditional argument against moral responsibility that is

meant to capture this intuitive idea.

A Traditional Argument against Moral Responsibility

(1) determinism is true.

(2) If determinism is true, then every human action is “pre-determined” by

conditions that obtained before the birth of its agent.

(3) If every human action is “pre-determined” by conditions that obtained

before the birth of its agent, then no human ever acts freely.

(4) If no human ever acts freely, then humans are never morally respon-

sible for their actions.

________________________

(5) Humans are never morally responsible for their actions.

We are currently supposing that determinism is true. The rationale for

premise (2) is that humans have been around for only about 100,000 years,

but the universe is billions of years old, so that (if determinism is true)

every human action was made physically necessary by conditions that

obtained before the birth of its agent. (The agent of an action is simply

the one who performs that action.) The rationale for premise (3) is that

an action that is made physically necessary by conditions that obtained

before the birth of its agent is not under the control of its agent;

and if

an action is not under the control of its agent, then it is not a free action.

And the rationale for premise (4) is that, as we said above, the kind of

freedom we are talking about here is a necessary condition for moral

responsibility.

It might be thought that we can escape the conclusion of this argument

by claiming that each person has an immaterial soul. But even if we do

have immaterial souls, they must either be subject to deterministic laws

of nature or else not. If they are, then the souls give us no way out of the

above argument. Whereas if we have immaterial souls that are not subject

to deterministic laws, then all the considerations raised below concerning

Indeterminism will be relevant. So it appears that either way, the positing

of immaterial souls will be immaterial to our discussion.

See, for example, d’Holbach, “The Illusion of Free Will;” and ree, “determinism and

the Illusion of Moral responsibility.” Not that anyone has ever presented the argu-

ment in exactly this form.

let’s set aside the question of whether it’s possible for a human to travel back in time

to before he or she was born.

Freedom and determinism 53

returning to the above argument against moral responsibility: Hard

determinism is essentially the combination of determinism with some-

thing like the reasoning contained in this argument. Here is an ofﬁcial

statement of the view:

Hard Determinism

(i) determinism is true. (ii) No human ever acts freely. (iii) Humans are

never morally responsible for their actions.

Many ﬁnd this position to be as inescapable as it is unpalatable. But as we

will see, there are ways of at least trying to escape it.

3.5 Compatibilism

one way to avoid the conclusion of the above argument is to deny

determinism. let’s set that move aside for the moment. (We’ll come back

to it shortly.) What about premise (2)? Can it be rejected? Yes, but only if

we insist either (i) that there are humans who have been around forever,

or (ii) that there was a beginning of time and also that there are humans

who have been around since that beginning. each of these is a substantive

and highly controversial empirical claim, however, and neither is the sort

of claim on which one could comfortably pin one’s hopes for freedom and

responsibility.

What about rejecting premise (4)? recall that the rationale for this

premise was our stipulation that we are talking about the kind of freedom

that is a necessary condition for moral responsibility. And this stipulation

was based on the idea that we all share an intuition according to which

there is a kind of freedom that is required for moral responsibility. Perhaps

in a more in-depth discussion of these issues it would be worth question-

ing whether we all really do have such an intuition; but we will not do so

here.

let us then consider another way to avoid the conclusion: rejecting

premise (3). Note that this premise is a conditional. To deny a conditional

is to say that its consequent does not follow from its antecedent. So reject-

ing premise (3) involves saying that even if it’s true that every human

action ever performed is made physically necessary by conditions that

obtained before the birth of its agent, it doesn’t follow that we humans

never act freely. For our purposes, this amounts to granting that every

An Introduction to Metaphysics54

human action ever performed is made physically necessary by conditions

that obtained before the birth of its agent, but nevertheless insisting that

humans sometimes act freely. That is, rejecting premise (3) amounts to

saying that someone can act freely even if their action was made physi-

cally necessary by conditions that obtained before they were born. Applied

to the above example, in which you take the money from a wallet that

you ﬁnd on your way home, this strategy involves claiming both that your

action was made physically necessary by conditions before you were born

and also that you nevertheless acted freely.

The general idea that you can be free even if your actions are determined

by conditions going on before you were born is known as Compatibilism.

Here is our ofﬁcial formulation of the view:

Compatibilism

Free Will is compatible with determinism.

Typically, the Compatibilist will say that even if all our actions are deter-

mined from long ago, there is still an important distinction to be made

between long-determined actions that are free and long-determined

actions that are not free. And a favorite Compatibilist technique for motiv-

ating this distinction is to point out that, in the ordinary business of life

(outside of “the philosophy room”, that is), the factors that we attend to in

judging whether an action is done freely have nothing to do with whether

determinism is true. Instead, the Compatibilist points out, we pay atten-

tion to whether the agent in question is unconstrained, and is able to move

his or her body in normal ways, without being subject to strange powers

like hypnosis or mind control.

Some examples that a Compatibilist might appeal to in motivating the

claim that we typically draw a distinction between free actions and unfree

actions without reference to determinism are shown in Table 3.1.

With respect to the ﬁrst pair of examples, it’s natural to say that the

agent described in the left-hand column is acting freely while the one in

the right-hand column is most deﬁnitely not acting freely. But notice that

examples of Compatibilists include Aristotle (see his Physics, Book II, chapter 4);

Hume (Inquiry Concerning Human Understanding, Section VIII); Hobart (“Free Will as

Involving determinism and Inconceivable without It”); Stace (“The Problem of Free

Will”); and lewis (“Are We Free to Break the laws?”).

Stace uses a similar table of examples in “The Problem of Free Will.”

Freedom and determinism 55

Table 3.1 A Compatibilist table of free and unfree actions

Actions that would ordinarily be

considered free

Actions that would ordinarily be

considered unfree

A man walks down the street,

stops at an intersection where

he encounters a “don’t Walk”

sign, notices a button for getting

the “Walk” sign to light up,

deliberates about what to do,

consciously decides to push the

button, and then voluntarily

pushes the button.

A man’s body is tightly bound in

such a way that his captors can

control his bodily movements.

His captors force him, against his

will, to move toward the but-

ton for a “Walk” sign. Then they

force him, still against his will, to

move in such a way that his ﬁnger

pushes the button.

A woman in a restaurant peruses

the menu, deliberates in the

normal way, and then orders

Chicken Parmesan, because it’s

her favorite dish.

A woman in a restaurant has been

hypnotized so that if she hears

the words “Afghanistan banana

stand,” she will automatically say

“Chicken Parmesan.” When it’s

her turn to order, her companion

says “Afghanistan banana stand,”

and so the woman utters the

words “Chicken Parmesan.”

A man walks, of his own volition,

into the back room of a bar. He

thoughtfully picks up a gun he

ﬁnds there. When he notices his

sworn enemy bound helplessly

to a chair, he says with a sneer,

“At long last, here is my chance

to kill you.” Then he shoots his

enemy.

A man has had a brain-control device

implanted in his head by Martians,

who are able to force him to do

and say whatever they want. The

Martians force the man to walk

into the back room of a bar, pick

up a gun, and shoot an innocent

victim, despite the fact that shoot-

ing anyone goes completely against

this man’s character.

A woman is talking to her best

friend. She wants to show her

friend the latest family snap-

shots, which are in her wallet. So

she hands over the wallet with a

big smile.

A woman is held up at gunpoint. She

hates to do so, because it contains

not just cash and credit cards but

also her cherished family photos,

but reluctantly she hands over her

wallet.

An Introduction to Metaphysics56

there is nothing in our description of these two examples about whether

determinism is true. Nor is there anything in the descriptions about

whether either man’s action is determined by conditions that obtained

before he was born. And similar things are true with respect to our other

pairs of examples.

The Compatibilist will typically take this to be strong

evidence showing that whether we consider some action to be free has

everything to do with the satisfying of various conditions – including,

perhaps, the absence of external constraints, mind-control devices, and

coercive threats, together with a proper ﬁt between what the agent wants

to do and what he or she actually does – none of which has anything to

do with the truth or falsity of determinism. And the Compatibilist will

further take this as evidence that freedom is compatible both with an

action being pre-determined in the relevant way, and also with the truth

of determinism.

In what follows we will consider three representative versions of

Compatibilism, beginning with one that we will call Hobartian Compatib-

ilism.

To appreciate the basic idea behind Hobartian Compatibilism, let’s

suppose that when you took that money from that wallet, these two things

were true of you: (i) if you had wanted to return the wallet instead of tak-

ing the money, then you would have decided to return the wallet (in short,

your decision was based on your desires); and (ii) if you had decided to

return the wallet, then you would have returned the wallet (in short, your

action was based on your decision). on these suppositions, according to

this Compatibilist way of thinking, you were acting freely when you took

the money.

More generally, the Compatibilist who takes this line will focus on

whether an action satisﬁes the following two conditions:

The Desire Condition

If the agent had desired to act differently, then he or she would have

decided to act differently.

With the possible exception of the last pair. For some Compatibilists have argued

that coerced actions like the woman handing over her wallet in the last example

are in fact free actions. See, for example, Markosian, “A Compatibilist Version of the

Theory of Agent Causation.”

This view is defended in Hobart, “Free Will as Involving determinism and

Inconceivable without It.”

Freedom and determinism 57

The Decision Condition

If the agent had decided to act differently, then he or she would have

acted differently.

And such a Compatibilist will say that free actions are those actions

that satisfy these two conditions. Here is our ofﬁcial formulation of this

view.

Hobartian Compatibilism

An action is free if and only if it satisﬁes both The desire Condition and

The decision Condition.

In addition to endorsing the above claim, Hobartian Compatibilists will

typically also hold both (i) that we humans often perform free actions,

and (ii) that we humans are in general morally responsible for our free

actions. For there would be little satisfaction in saying that the conditions

for an action being free are consistent with determinism, but that in fact

our actions never satisfy those conditions. Nor would it be very satisfying

to say that we act freely but are for some other reason not responsible for

our actions.

other Compatibilists prefer to focus on the question of whether the

agent in question is, at the time of acting, under any unusual constraints,

such as those present in the examples on the right-hand side of our table

above.

Here is a deﬁnition that will help us to formulate this version of

Compatibilism:

Action A is unconstrained =df at the time of A, the behavior of A’s agent

is not inhibited in any unusual way (by, for example, metal bars, or a

straitjacket, or strings, or wires, or hypnosis).

And here is our ofﬁcial formulation of the view:

Stacean Compatibilism

An action is free if and only if it is unconstrained.

Similar remarks apply to other forms of Compatibilism, including Stacean

Compatibilism and Combo Compatibilism (see below).

Stace, in “The Problem of Free Will,” is an example of such a Compatibilist.

Throughout this text we use the expression ‘= df’ as an abbreviation for ‘means, by

deﬁnition’.

An Introduction to Metaphysics58

of course, it is possible to combine the ideas behind both Hobartian

and Stacean Compatibilism into a single view. doing so will give us the

following:

Combo Compatibilism

An action is free if and only if (i) it satisﬁes both The desire Condition and

The decision Condition, and (ii) it is unconstrained.

So much for our representative versions of Compatibilism. We now turn to

its rival, Incompatibilism.

3.6 Incompatibilism

Most people will ﬁnd it easy enough to appreciate the intuition that our

actions cannot be free if they are made physically necessary by events

going on before our births. Here is a way to articulate this intuition.

Incompatibilism

The Free Will Principle is incompatible with determinism.

Incompatibilism is closely related to Hard determinism, but the two views

are nevertheless distinct. For one thing, Hard determinism has (what is argu-

ably) an empirical component – the claim that determinism is true – that

Incompatibilism lacks. And for another thing, Incompatibilism, in virtue

of claiming that freedom is incompatible with determinism, makes a modal

claim that is not, technically, a part of Hard determinism. Still, the t wo v iews

are closely related, in the sense that most Hard determinists endorse tenets

(ii) and (iii) of their view (the claims that humans are neither free nor morally

responsible for their actions) precisely because they think that determinism

is true and also believe Incompatibilism. (For this reason, Incompatibilism can be

seen as the philosophical engine that drives Hard determinism.) In any case,

it is worth emphasizing that Incompatibilism, unlike Hard determinism, is

not (even arguably) susceptible to empirical refutation.

There have been a number of important arguments for Incompatibilism

over the years, but the most important one in recent philosophical litera-

ture is one that has been developed and defended by Peter van Inwagen,

and that he calls “The Consequence Argument.” Here is an informal

characterization of the argument that appears in several places in van

Inwagen’s An Essay on Free Will.

See, for example, van Inwagen, An Essay on Free Will, pp. v, 16, 56, and 222.