How Grecian can Grice get?

by JLS
for the GC

In "Determinism and indeterminism", R. J. Hankinson traces the Grecian (not Gricean) story of determinism. Good to connect with some Gricean (now) points.

"The notion of universal causation," writes Hankinson, "was ubiquitous in later antiquity; to loosen those ties threatened the irruption of chaos."

---

How could the evident continuity and regularity of the world survive the intervention ofcasual elements into its structure?

Still, there is a clear distinction, one
exploited by the Epicureans, between the assertion of a universal principle
of causation and any determinism.

It is one thing to accept that every
event is caused, quite another to believe that the nature and sequence of
all events is rigidly fixed for all eternity.

The latter belief forms the core of
any determinism – and it is its prima-facie implausibility, along with what
are taken to be its unacceptable consequences (for human freedom, for the
concept of responsibility), that lays it open to attack.

The origins of the problem in the Greek world were not, however,
metaphysical.

The Sophistic movement of the late fifth century bc was
particularly interested in new forms of forensic argument, especially
defence argument. Gorgias’ Helen is a case in point: Helen of Troy is innocent
of adultery, he argues, because she did what she did either under
physical compulsion, or under the influence of love, or at the whim of
some god, or persuaded by arguments. In none of these eventualities can
she be held responsible for her actions, since in all of them she is compelled
by some external force; the list is exhaustive; hence she is not
responsible for what she did. Gorgias’ rhetorical exercise is not serious
philosophy, but it raises serious philosophical points. If our actions are
indeed conditioned by factors that lie outside our control, how can we
reasonably be held responsible for what we do? Society is, indeed, to
blame: Gorgias is the Ur-progenitor of hard determinism.

Aristotle saw the problem with characteristic clarity. At the beginning
of Book iii of his Nicomachean Ethics, he wrote:

Since ethical virtue is concerned with emotions and actions, and those
which are voluntary are praised or blamed, while those which are involuntary receive pardon and sometimes pity as well, students of ethical virtue
must presumably determine the limits of the voluntary and the
involuntary . . . Actions are regarded as involuntary when performed
under compulsion or through ignorance. (EN iii.1.1109b30–1110a1)
And he was well aware of the dangers that attended any attempt to mitigate
responsibility by blaming factors outside the agent’s control:
If it were argued that pleasurable and admirable things have a compulsive
e◊ect (because they bring external pressure to bear on us), it would
make all acts compulsory, since every act of every agent is done for the
sake of such objects . . . It is absurd for the agent to lay blame on the
external factors and not upon himself for falling easy prey to them, and
to attribute his fine acts to himself, but his disgraceful ones to the attractions
of pleasure. It seems reasonable, then, that an act is compulsory
only when its originating cause is external, and receives no contribution
from the person under compulsion. (EN iii.1.1110b9–17)
Aristotle holds that we are responsible for what we do just in case the origin
of the action is within us (he has a rather literal notion of coercion); if
we contribute at all to it, then we are responsible for it. It is, fundamentally,
in virtue of the state of our characters that we are praised or blamed;
and it is insofar as we act on the basis of choice, or after deliberation, that
we are morally responsible for what we do. When we make a choice, we
choose to perform or refrain from performing some action – that choice is
up to us, the outcome of our desires, beliefs, and deliberations. We desire
a particular end, and deliberate about and select means to that end;
actions of this sort are performed in accordance with choice and are thus
voluntary: ‘therefore virtue is ‘up to us’ (eph’ he–min)’ (EN iii.5.1113b3–6).
Helen may have been persuaded by arguments – but it was she who was so
persuaded: Penelope did not succumb. Or she may have been swept away
by powerful emotions and desires – but she shouldn’t have been: so much
the worse for her morals. Aristotle may accept Gorgias’ disjunction of
alternatives as exhaustive – he can still resist the conclusion that Helen’s
action was not her fault.
Aristotle starts from the obvious fact of moral life that we do hold people,
including ourselves, responsible for what they do; otherwise we could
not legitimately praise or blame them. But if praise and blame are legitimate,
some things must be up to us (eph’ he–min). He is not blind to the
di√culties of the picture he tries to develop. We are responsible for our
actions insofar as they derive from our choices, made in line with our preferences
– and hence derivatively because they are the products of our characters. It is because I am weak-willed that I eat another danish pastry
when you, exercising self-control, refrain from so doing. But as Aristotle
clearly saw, it is only if we are further responsible for our states of character
themselves that we can be held to account in any strong sense for
actions performed as a result of them. He scouts a possible objection:
Everyone aims at what seems to him to be good, but over this appearance
he has no control. How the end appears to each individual depends on
the nature of his character, whatever this may be. So if the individual is in
a way responsible for his state of character he will also be in a way
responsible for his view of what is good; but if he is not responsible for
the former, then no wrongdoer is responsible for doing wrong. (EN
iii.5.1114a31–b4)

Perhaps virtuousness is simply a matter of natural endowment, or proper
upbringing – but neither of these seem to be in any obvious sense up to
the individual in question. The problem is occasioned by the transitivity
of responsibility. Suppose some individual I’s action A is reasonably
pinned on his character C – nevertheless, if we can trace responsibility for
C to some set of factors F, where F are outside I’s control, then F, and not
I, are truly responsible for A. That argument has a certain plausibility to
it, and it re-opens Helen’s defence.

Central to Aristotle’s account (and many others) of human responsibility
is the notion of choice. We are free, and hence responsible, just insofar
as we can choose to do what we do. If it can be shown that this choice is a
chimaera, then human freedom may turn out to be equally illusory. And
one way of going about doing that is by arguing for determinism. If the
entire course of the universe is ineluctably mapped out in advance, its
unfolding being merely the working out of an inevitable fate, there seems
little room for genuine human agency at all. Rather we are all puppets of
the ultimate causal forces of the universe, and our autonomy is mere illusion.
That inference, from the ineluctability of fate to the impossibility of
human freedom, has been deployed by determinist and indeterminist
alike, although from di◊erent directions. The unthinkability of
human bondage has frequently formed the basis of an argument to indeterminism
via a modus tollens of that implication – that strategy was pursued
by the Epicureans. Conversely, the rationally compelling nature of
the argument for universal causation has sometimes pushed people in the
other direction: Gorgias is only the first of many. Others, most famously
Hume, have tried to reconcile some concept of human freedom with
the origins of the question 515

determinism. That was Chrysippus’ line. But first of all another source for
the debate merits brief consideration.

ii Logic and contingency

In another famous passage (Int. 9), Aristotle discusses the question of
whether singular propositions about the future have determinate truth
values.1 Suppose it is fine today; surely then it was true yesterday to say
that it would be fine today – and the proposition ‘it will be fine tomorrow’
was true yesterday. More than that: it was true 10,000 years ago, indeed it
has been true since the dawn of time. But if it is now true (let us say) that
there will (to use Aristotle’s famous example) be a sea-battle tomorrow,
surely it is now unavoidable that there will be such a battle? And similarly
with the most contingent-seeming propositions:
Consequently it is necessary that either the a√rmation or the denial [of a
certain proposition] be true. Nothing then will either be or come to be
by chance or contingently . . . everything will be of necessity and not
contingent (Int. 9.18a6–8)
and
these and other bizarre consequences follow, if at least we assume that of
every a√rmation and denial . . . of contradictory opposites one must be
true and the other false: there can be no contingency in things that come
to be, and everything that is and comes to be does so of necessity. (Int.
9.18b26–31)
However, Aristotle goes on to reject this as being plainly inconsistent
with our experience that human deliberation is a source of events (Int.
9.19a7).

Aristotle’s argument rests in part on the thesis that

(T1) the past is necessary.

What’s done cannot be undone: and if we gloss ‘necessary’ as ‘fixed, unalterable’,
then (T1) is attractive. It is surely one of the fundamental temporal
distinctions that future events, unlike past ones, can be a◊ected and
altered by what is done now. But if that is right, it seems reasonable to
think, as Aristotle apparently did, that the status of propositions about
the future should be di◊erent from that of propositions about the past,

1 At least, this is the most usual interpretation of what he is doing; a rival account was urged by G.
E. M. Anscombe (Anscombe 1956), to the e◊ect that what was at issue was not future truth but
future necessity.
Cambridge Histories Online © Cambridge University Press, 2008
and not merely for epistemic reasons (not merely, that is, because in the
nature of things we can know less about the future than the past). The
past is determinate, hence propositions about it are necessarily true. The
future is not, and so utterances referring to it must at best be contingently
true if they are true at all.2
That argument is seductive – it seduced Epicurus. But it is won at the
price of abandoning the semantic principle of bivalence, at least in its
most general form. And there is something to be said for bivalence, most
obviously that it seems guaranteed by the logic of negation. Surely, if a
proposition makes sense it must be either true or false? And in any case,
are we not merely dealing with di◊erently-tensed and indexed versions of
the same proposition in the case of today’s fine weather? It seems at best
arbitrary to say that, even if it is fine today, yesterday’s accurate weatherforecaster
wasn’t telling the truth. But if he was, then it was already going
to be fine today yesterday – and that, at least for Aristotle, entails multiple
absurdities. That sequence of argument is perhaps confused – but its confusions
are not trivial (it took a Carneades to expose them), and they still
have the power to perplex.
On the basis of (T1), and the further (also Aristotelian) thesis that
(T2) an impossibility cannot follow from a possibility,
Diodorus Cronus attempted to do away with contingency altogether by
way of his celebrated Master argument (ii.1).3 Whatever its precise form
there is no doubt that it was of enormous influence on succeeding generations
of philosophers, and its move from logic to metaphysics appears to
have been widely accepted as valid.
iii The Hellenistic response
The Epicureans countered by holding that the principle of bivalence
failed for future contingents: they were neither true nor false prior to
their being actualized, and hence a fortiori were not either necessarily true
or impossible (Cic. Fat. 21). Furthermore, the apparatus of the swerve is
introduced explicitly, Cicero says, to provide a physical, indeterminist
the hellenistic response 517
2 This is inadequately crude as it stands – no indeterminist need think that all propositions about
the future are undetermined: ‘in 100 years I shall be dead’ seems about as necessary (causally
speaking) as anything else. For Aristotle, there are certain truths about natural kinds which are
dictated by the structure of the kind itself – these can be given a future tense, but they are no less
settled for all that. Yet there are indefinitely many things about my death which are, on this
view, undetermined at the present time. Its date, its place, its mode, and so on.
3 See above, pp. 88–92.
Cambridge Histories Online © Cambridge University Press, 2008
basis for the rejection of bivalence which is taken to be a necessary concomitant
of the maintenance of human freedom:
Epicurus introduced this theory [i.e. of the swerve] because of his fear
lest, if the atom was always carried along by its natural and necessary
gravity, no freedom would be left for us, since the mind will move under
compulsion from the atoms. (Cic. Fat. 23)4
But the Epicureans’ attempt to rescue freedom along with contingency
has its price. They must reject bivalence; and they are forced to compromise
the clean lines of their physical theory by introducing the minimal
atomic swerve. Chrysippus and the Stoics, on the other hand, accepted
bivalence, since they could not see how a proposition could fail to be
either true or false – and as a result they felt themselves forced to accept a
version of determinism. Both Epicureans and Stoics, then, accept the following
theses:
(T3) if truths are timeless (or eternal), then if a proposition p is true at any
time, p is true at all times;
and
(T4) if p is true at all times, then p is necessary (in the sense of its being
determined).
The Stoics accepted the timelessness of truth,5 and hence the determination
of all truths; the Epicureans discountenanced the determination of
all truths, and hence rejected their timelessness, as well as bivalence. But
rejecting bivalence is a tricky business. After all, surely even if Caesar
might have escaped death at the hands of the conspirators, the claim that
he will not be murdered is actually false whenever uttered prior to the
Ides of March. If Teiresias or Elijah had said, centuries earlier, ‘Caesar will
be murdered on the Ides of March’, they would have been telling the
truth. One does not even need anything as strong as (T1) to do the trick
here: past truth alone will do. On the other hand, the Stoics have a di√cult
518 determinism and indeterminism
4 This translation (and subsequent ones) follows those of Long and Sedley 1987, vol. i, 102–7. For
the swerve also see above, pp. 501–2.
5 I should note here that I do not mean to imply that for the Stoics the proper unit of fundamental
semantic appraisal was a timeless proposition construed in the manner of contemporary classical
logic. See above, pp. 95–6. On the contrary, the Stoics, in common with other ancient
semanticists, held that non-indexed sentences were the basic truth-bearers, and hence that
‘propositions’ such as ‘it is fine’ may change their truth-value. But that is irrelevant to the point
at issue, which is whether or not an indexed ‘proposition’ (‘it is [tenseless] fine at spatio-temporal
region S’) is such as to be always true if true at all – the Stoics hold that it is: and in that
sense their propositions are timeless.
Cambridge Histories Online © Cambridge University Press, 2008
time of it explaining how truths can be determined without being (in
some sense) necessary.
It is one of the great achievements of the Academic sceptic Carneades to
see that this dilemma is not forced upon us. We may distinguish6 between
Causal Determinism (CD: the thesis that each event is the ineluctable
product of antecedent causes); Logical Determinism (LD: if an event is
going to happen it is already true that it will happen); and Epistemic
Determinism (ED: if it is known that an event will happen, then that
event cannot fail to occur).7 The Stoics rely on the view that if a future
proposition is now true there must be some truth-maker for it in the
world now: it must already be true that it is unavoidable. But as Carneades
showed, that is a mistake. LD does not entail CD. For any future contingent
proposition, either it or its contradictory is true. But it is not true
necessarily in virtue of anything in the world now; rather it will be made
true by the event as it turns out:
The truth of propositions like ‘Cato will come into the Senate’ is
brought about by contingent causes, not by causes bound up in nature
and the world. And yet that something will come about, when true, is as
immutable as the truth that something has come about. (Cic. Fat. 28)
If it is now true that I shall die at sea, then nothing I can now do will alter
that truth. But does that mean I am fated to die at sea? No: because it is
only true that I shall die at sea if as a matter of fact I now do nothing to
prevent it, and as a matter of fact act in such a way that eventually leads to
my maritime death. If I prevent my own watery demise, then it will never
have been true that I was going to die at sea. The assimilation of LD to CD
rests on a straightforward mistake; CD entails LD – but not vice versa, as
both Stoics and Epicureans believed. As regards ED, Carneades holds that
it entails CD (and hence LD); and so, if CD is false, ED must be too:
That is why Carneades used to say that not even Apollo could tell the
future apart from things whose causes were embodied in nature in such a
way as to render their coming about necessary. For by inspecting what
could even the god himself tell that Marcellus . . . would die at sea? This
was something that was true from eternity but did not have causes working
to bring it about. (Cic. Fat. 32–3)
That is, I think, a mistake: it is logically possible that one might know that
some future event will occur, and yet it need not be the case that the event
the hellenistic response 519
6 With Long and Sedley 1987, vol. i, 466.
7 Sorabji (1980a) discusses these categories (although not under these names) and the relations
that hold between them, see especially chs. 1–3; also see Sorabji 1980b.
Cambridge Histories Online © Cambridge University Press, 2008
is causally determined at the time of knowledge. Perhaps one can simply
see the map of time laid out before one, as it were; perhaps the fact of the
later event actually causes (or is part of the cause of ) my current knowledge
of it, even though it has not yet occurred and is not yet causally
determined. And theologians have argued that God may know how we are
going to choose, and hence whether we are doomed to damnation or are
saved, without that entailing that our choices are unfree, even for a noncompatibilist
sense of freedom.
But even if such possibilities are logically open, they are exotic suppositions,
and nothing in our experience gives us any reason to suppose that
they are true. Furthermore, it seems clearly to be the case that if ED holds
then the future events which are within its grip are in some sense already
fixed, and if they have already been fixed it surely does follow that nothing
can now be done to prevent them. This does not, perhaps, entail that they
are causally determined; but it does entail that they are determined, and
that fact alone may be enough to undermine any robust notion of freedom.
Here is Carneades’ argument, as reported by Cicero:
(1) If all things come about through antecedent causes, all things come
about through the interconnection in a natural chain. (2) If that is so, all
things are the product of necessity. (3) If that is true, nothing is in our
power. (4) But there is something in our power. (5) But if all things come
about through fate, all things come about through antecedent causes. (6)
Therefore it is not the case that whatever happens happens through fate.
(Cic. Fat. 31)
The Stoics accept (5) and (1); they reject (2), but Carneades is surely right, as
we have seen, to see that they are committed to it at least in a sense strong
enough to cast doubt on the view that human agents are anything more than
instrumental causes of their actions, and hence (3) will follow, at least for a
strongsense of ‘in our power’. (4) is simply a bald assertion: no argument is
o◊ered for it. Perhaps, like Dr Johnson, Carneades relies on the view that
‘all experience is for it’; perhaps he simply takes the Stoics to be committed
to it. At all events, (4), alongwith the other premisses, clearly entails (6).
Here we should turn to another Carneadean argument from Cicero’s
De fato. In order to counter arguments like that of (1)–(6) above, the
Epicureans felt it necessary to introduce the atomic swerve, a sudden,
unpredictable, uncaused quantum movement in atomic motion, in order
to account, among other things, for human freedom (iv.3).8 Carneades
520 determinism and indeterminism
8 ‘Quantum’ is more than mere metaphor here; the evidence suggests that the swerve involves the
minimal possible divergence from the atom’s previous trajectory.
Cambridge Histories Online © Cambridge University Press, 2008
pointed out that the swerve was superfluous to the Epicureans’ requirements:
A more penetrating line was taken by Carneades, who showed that the
Epicureans could defend their case without this fictitious swerve. For
since they taught that a certain voluntary motion of the mind was possible,
a defence of that doctrine was preferable to introducing the
swerve, especially as they could not discover its cause. And by defending
it they could easily stand up to Chrysippus. For by conceding that there
is no motion without a cause, they would not be conceding that all
events were the results of antecedent causes. For our volition has no
external antecedent causes. (Cic. Fat. 23)
If we say someone acts without a cause, Carneades continues, we mean
only without external causes: their volitions still cause their actions. But
their volitions are not themselves caused. Hence the Epicureans can avoid
uncaused events (at least if volitions are not events), with all the problems
that they entail, and still reject universal determinism, and hence support
freedom. Carneades holds that the Epicureans can admit the truth of
(7) no event occurs causelessly
and so avoid ‘incurring the scorn of the natural philosophers’ (Cic. Fat.
25),9 and yet still hold that actions are events and are caused, since
(8) actions are caused by the will;
but the volitions themselves are not caused, at least not by anything external
to us. This line of argument requires that either the volitions themselves,
or what causes them, are not themselves events (if events must have
antecedent, and hence independent, causes); ‘pure acts of the will’, or
something of the sort, are supposed in some sense to be self-caused,
brought about by their own internal nature, in just the same way as atoms
fall (on the Epicurean account) because of their intrinsic weight.
There are obscurities in this doctrine, and it will convince no Humean,
wedded to the necessary distinctness of cause and e◊ect. And there are
di√culties with reconciling the notion of the human will as a sort of selfstarting
mechanism with the evident fact, noted by Chrysippus, that
external influences are at the very least necessary conditions of our having
acts of the will at all.10 The following argument suggests itself: if autonomy
involves having control over one’s actions, then the will of the autonomous
agent must be su√cient for those actions. But that conflicts with
the hellenistic response 521
9 See above, p. 502. 10 See below, pp. 531–4 and 577–80.
Cambridge Histories Online © Cambridge University Press, 2008
the admission that externals are necessary conditions for the action
(except on the absurd supposition that the act of will is itself a su√cient
condition for the presence of external conditions). However, given the
notion of causation canvassed in the previous chapter, this need not be
the problem it appears to be; when the external object impinges, it sets
the process in motion – but it is the volition itself that keeps it going. This
will be developed later on.
iv The Epicurean position
So much for the logical argument designed to show that the principle of
bivalence entails determinism. Even if that argument fails, there is still
work to be done for the anti-determinist. The grip of the principle that
every event must have a cause was a strong one – and provided that the
principle is interpreted strongly such that every event has a specific
cause,11 determinism, with all that it entails, seems unavoidable.
Causation will become assimilated to necessitation.12 Certainly there is
reason to believe that Aristotle did not think of his causes as invariably
necessitating: if A is only for the most part causally correlated with B,
none the less there is nothing amiss in saying that A causes B, although it
does not necessitate it: for to say that A necessitated B, it would have to be
the case that A-type events were invariably followed by B-type events. But
even if that is the case, it still leaves wide open the question of whether
events (considered as individual tokens, and not under some particular
generic and allegedly explanatory description) are uniquely determined
by antecedent circumstances – and here the evidence is far more equivocal.
So if you think that universal causation entails determinism, while
determinism is incompatible with human freedom, yet you also think that
human freedom is an incontrovertible datum of ordinary experience, then
you will be bound to deny universal causation. And that is precisely the
path Epicurus and his followers took. Carneades, as Cic. Fat. 23 above
demonstrates, clearly thought that the Epicureans solved nothing simply
by introducing unpredictable swerves – randomness is not the form of
human freedom; and such views have frequently surfaced in the succeeding
debate. If my actions are detached from the fetters of necessity at the
expense of making them simply random, it is hard to see what concept of
522 determinism and indeterminism
11 See above, p. 481 (T1*).
12 There have been modern attempts to deny the link between the two, and equally some modern
commentators, notably Sorabji 1980a and 1980b, wish to ascribe such a denial to the ancients.
Cambridge Histories Online © Cambridge University Press, 2008
freedom can be embodied thereby. We surely do not think of ourselves
being free just insofar as we are loose cannons, firing uncontrollably.
Indeed, an influential concept of human freedom, one endorsed in
di◊erent ways by the Stoics, the early Christians, and Galen, has it precisely
that human freedom involves self-control, and hence regularity in
behaviour. Moreover, even if volitions may be thought of as possessing
causal e√cacy which is in some way independent of, or at least not reducible
to, the sum of the atomic motions, it is not clear why the atoms
should need to be able to swerve: all that is necessary is that acts of the will
can deflect them from their normal trajectories, as Carneades saw.
However, Epicurus may not have thought the swerve to be constitutive
of human freedom. Recently, David Sedley has argued13 that the function
of the swerve is simply to allow for acts of volition: that every swerve is of
its nature uncaused, is an unjustified inference of Epicureanism’s opponents.
Rather, the atoms’ ability to be deflected allows volition a toehold
into causal e√cacy, since the will can now actually a◊ect the mechanical
course of atomic events.14 To say (in the case of volitions at least) that the
atoms swerve is simply to say that their motions are not exhaustively
determined by the force, momentum, shape and velocity of the atoms
themselves, that is by their intrinsic, essential properties. This is not of
course to say that all the swerves are caused by volitions – just that there is
available, as it were, a set of alternative trajectories into which the atoms
can be forced by acts of the will; the will can then take advantage of the
causal elasticity so permitted, and, within limits at least, mould events in
its image.
This interpretation at least avoids the reduction of freedom to randomness.
But it is highly controversial;15 and it is clear neither whether
Epicurus actually held it, nor if he did how it was to be developed. Sedley
bases his case upon a papyrus fragment of the On Nature, in which
Epicurus discusses the proper attitude to take towards those who squander
their natural gifts:
The nature of their atoms has contributed nothing to some of their behaviour,
and degrees of behaviour, and character, but it is their developments
which themselves possess all or most of the responsibility for
certain things. It is a result of that nature that some of their atoms move
with disordered motions, but it is not on the atoms should be placed>. Thus when a development occurs which takes on
the epicurean position 523
13 Sedley 1983b and 1988; see also the brief account in Long and Sedley 1987, vol. i, 107–12; it is
assessed and criticized in Laursen 1988.
14 See Long and Sedley 1987, vol.i, 110–12. 15 See the criticisms of Everson, pp. 553–7.
Cambridge Histories Online © Cambridge University Press, 2008
some distinctness from the atoms in a di◊erential way – not in the way
which is like viewing from a di◊erent distance – he acquires responsibility
which proceeds from himself. Then he immediately transmits this to
the primary substances. (Ep. Nat. xxv, [34] 21–2 Arr.2)
That passage is obscure, and its readings are disputed.16 But Epicurus
does appear to make a distinction between the ordinary, bottom-up causality
of atomic motion and the way in which the individual’s settled disposition
must be conceived of as operating: it is not, so he suggests,
merely one of scale or perspective (this is presumably the point of the
aside about viewing from a distance). Rather (if this account is correct), he
appears to envisage a new level of causal power emerging above the microscopic
level, whose activity is in some way autonomous, and can even feed
back into the workings of that micro-level.
Yet, as far as we know, if Epicurus did adopt such a two-level causal
position, he made no attempt to elaborate upon it; furthermore, the passage
may be read in such a way as not to involve the emergence at the
macroscopic level of new causal forces. Epicurus may simply be denying
the inference from the fact that the individual is composed of atoms with
their own laws of working to the claim that it is the atoms themselves, and
not the individual, who is responsible for that individual’s actions. The
target of the passage would then be not physicalism (or reductionism) as
such, but rather physicalist determinism.17
Even so, it is quite unclear how, even on that view, Epicurus’ thought is
supposed to be developed here, and there is no real indication (although
given the fragmentary nature of the sources, such remarks should be
handled with care) that Epicurus did develop it. Perhaps he felt that a
simple outline was enough to establish its possibility, and hence preferability
to an utterly deterministic view of mind and action of the type he
ascribes to Democritus and his followers, and which he takes to be absolutely
rationally unacceptable. In a relatively well-preserved fragment
(Nat. xxv, [34] 26–30 Arr.2), Epicurus takes his Democritean opponents
to task for not seeing the self-refuting nature of the thesis of determinism:
they ‘debate this very question on the assumption that their opponent is
himself responsible for talking nonsense’. And if the opponent alleges that
this behaviour is itself necessitated, then he will be forced into a regress.
Once again, the exact nature of Epicurus’ complaint is di√cult to establish
– he does not appear to think that regress is itself vicious, merely that
in some sense the fact of it renders the determinist’s stance empty. The
524 determinism and indeterminism
16 As is its location in On Nature, see below, p. 532, n. 30. 17 See below, pp. 529–31.
Cambridge Histories Online © Cambridge University Press, 2008
point seems to be that even the determinist relies on the persuasiveness of
his arguments – that is, they are supposed to supply reasons why his opponent
should modify his views. But one can only act upon reasons if one
can act, that is, if one is free (if there is, in Epicurus’ language, some ‘auxiliary
element or impulse within us’: Nat. xxv, [34] 29 Arr.2) – hence the
determinist’s position is pragmatically at least self-stultifying. That argument
is not convincing (no committed determinist will find it too di√cult
to evade) – but it is subtle.18
There is, however, a less subtle but still important general line of criticism
to be examined. It is put most succinctly in one of the collections of
Epicurean sayings:
The man who says that all events are necessitated has no ground for criticizing
the man who says that not all events are necessitated, since
according to him it too is necessitated. (Ep. Sent.Vat. 40)
Universal determinism, so it is said (and has been repeated innumerable
times since) makes the practices of praise and blame, reward and punishment,
ethically null. We shall return to the Stoic response to this later.
Lucretius leaves us with no doubt that it is the apparent undeniability
of the real existence of independent volitions which pushes the
Epicureans to take the line they do;
Furthermore, if all motion is always linked, and new motion arises out of
old in a fixed order, and atoms do not by their swerve make some beginning
of motion to break the decrees of fate, so that cause should not follow
cause from infinity, from where does free volition exist for animals
throughout the world? From where, I ask, comes this volition wrested
away from the fates, through which we proceed wherever each of us is
led by his pleasure and likewise swerve o◊ our motions at no fixed time
or fixed region of space, but wherever the mind carries us? For without a
doubt it is volition that gives these things beginnings for each of us, and
it is from volition that motions are spread through the limbs . . . Nor is it
the same when we move forward impelled by a blow through another
person’s great strength . . . For then it is plain that all the matter of the
whole body moves and is driven against our wish, while volition reined it
back through the limbs . . . So in the seeds too you must admit . . . that
there is another cause of motion besides impact and weight, from which
this power [i.e. volition] is born in us, since we see that nothing can come
to be out of nothing . . . That the mind should not itself possess an internal
necessity in all its behaviour, and be overcome, and as it were forced
the epicurean position 525
18 See once again below, pp. 529–31.
Cambridge Histories Online © Cambridge University Press, 2008
to su◊er and be acted upon – that is brought about by a tiny swerve of the
atoms at no fixed place or time. (Lucr. ii.251–93)
Lucretius apparently invokes volition to avoid the complete linearity of
causal behaviour, to explain interruptions in motion, irregularities of
behaviour, and so on. It is indeed an evident fact of animal behaviour that
animals do not react in the same way to stimuli as inanimate objects. Kick
a stone in Johnsonian fashion and, even if you fail to refute Berkeley thus,
at the expense of a certain foreseeable discomfort you may none the less
propel it a certain distance – and the trajectory it follows will be determinate,
and in principle predictable on the basis of familiar physical laws.
Kick a puma, on the other hand, and the outcome is a good deal less certain,
both in terms of puma-trajectory and of your own subsequent discomfort.
Animals quite clearly do, as Aristotle put it, have an internal
principle of motion that stones do not.
But it is one thing to point out that fact – quite another to refute determinism
on its basis, and it is far from clear whether it calls for anything
resembling an atomic swerve. Lucretius seems to think that because animals
swerve ‘at no fixed place or time’, i.e. not in response to crude
mechanical laws, so too must the atoms. But that involves a gross fallacy,
one moreover that the atomists’ own insistence on the possibility of emergence
should have warned them against. In short, nothing Lucretius says
in this passage seems to require the swerve, or indeed to militate against
any but the crassest determinisms.
v The Stoic response to the Master argument:
fate and necessity
The Stoics too were perplexed by Diodorus’ argument – and while they
wished to reject the conclusion (that all truths are eternal and necessary),
they were apparently unsure as to how to do so. Cleanthes, Zeno’s immediate
successor, seems to have rejected (T1);19 Chrysippus his successor,
on the other hand, denied (T2) (Cic. Div. i.14).20 At all events they found
it deeply troubling – and something had to be wrong with it.
This may appear puzzling. The Stoics, after all, are partisans of a universal
causality. Numerous texts attest to their belief in an all-embracing
ineluctable fate, which they identify variously with the will of Zeus, Zeus
526 determinism and indeterminism
19 See above, p. 516.
20 Chrysippus’ adoption of this fairly desperate manoeuvre need not detain us: his defence of it
turns on a peculiarity of Stoic semantics; see Sorabji 1980b, 263.
Cambridge Histories Online © Cambridge University Press, 2008
himself, the Logos of the world, and so on. Furthermore, several texts
ascribe to Chrysippus belief in the necessity of fate:
Chrysippus . . . said that there was no di◊erence between what was
necessitated and fate, saying that fate was an eternal, continuous, and
ordered movement. (SVF ii 916; cf. 926)
Fate is described as a chain (or rope) of causes (SVF ii 915, 917, 920, etc.),
unravelling in an ineluctably determined manner; and as Sorabji writes ‘the
Stoics had a battery of other words for the inevitability which they applied
to this all-embracing fate’.21 Finally, Chrysippus is expressly said to have
composed the first book of his On Fate in order to show ‘that everything is
encompassed by necessity and fate’ (Diogenianus ap. Eus. PE vi.8.1). How
do they manage what Sorabji calls their ‘retreat from necessity’?
First of all, note that Chrysippus says that things are necessitated by
fate, not that they are themselves necessary. This is not as trivial as it
seems. We need here briefly to examine the Stoic treatment of the modal
concepts of necessity and possibility, whose proper interpretation was
already a matter of philosophical dispute among Diodorus’ circle.
Diodorus held that necessity was a matter of eternal truth;22 the Stoics’
view was apparently more generous:
A proposition is possible which admits of being true, there being no
external factor to prevent its being true . . . Necessary is that which,
being true, does not admit of being false, or if it does so admit is prevented
from being false by external factors. (D.L. vii.75)
The interpretation of the definitions given here is a matter of scholarly
controversy (and the examples o◊ered by Diogenes Laertius are not very
illuminating).23 The notion of ‘admitting of ’ is murky; and the precise
role of the second clause in each definition, referring to ‘external factors’
is a matter of dispute.
None the less, on the most reasonable interpretation it appears that the
Stoics are prepared to treat as necessary those things which simply as a
matter of fact have turned out to be true, and whose truth is now unassailable
(that fits in well with Chrysippus’ acceptance of (T1), the first
premiss of the Master argument). That last sentence may be misleading –
the stoic response to the master argument 527
21 Sorabji 1980b, 261–2; cf. SVF ii 202, 528, 913–14, 917–18, 923–4, etc.
22 This is a little loose, but it will do: the actual definition he o◊ers is ‘that which, being true, will
not be false’ (Boethius Int. 2, p. 234 Meiser). But he is clearly committed to eternal truth by the
Master argument.
23 See most clearly Frede 1974a, 107–17; also Bobzien 1986, Mignucci 1978, and see above,
pp. 118–20.
Cambridge Histories Online © Cambridge University Press, 2008
for the Stoics, there will always be a causal explanation as to why things
have turned out thus and so – there is no such thing as simply turning out
true. It is tempting to try to interpret this on the basis of the time of the
preventing. A statement that we would normally consider contingent will
only be necessary for the Stoics if there is as a matter of actual fact some
causal factor operative at the time to prevent its failing to be true.
If this is right, the Stoics can evade an obvious objection: if your definition
of necessity holds, then anything that will as a matter of fact turn out
true must be necessary, not for logical reasons, but because, given the
iron-clad necessity of the unfolding of fate, there are reasons in the world
now (in the form of the total nexus of its causal processes) why things will
turn out thus. Determinism should, after all, be temporally indi◊erent.
But, the Stoics will reply, consider what it is to be a cause, or at least a perfect
cause.24 If A is a perfect cause of B, A is actually acting to bring B
about. In this sense, there are no perfect causes of future events (crucially
causal perfection is not simply a matter of causal necessity and su√ciency
in a Davidsonian fashion).
This yields two distinct types of modality. The first one might label
‘species possibility’. In this case some predicate P is possibly applicable to
an individual of natural kind K just in case K’s can, other things being
equal, be P’s. Thus Philo of Megara apparently held that a piece of wood
at the bottom of the ocean could be burnt, just because wood is naturally
flammable. But secondly there is what might be called actual possibility,
according to which the submerged wood is not now flammable because of
actually obtaining circumstances. The Stoics, on this view, restrict nonactual
species-possibilities to future cases; but they do none the less admit
some of them. The Stoics buy Philo’s account in forward-looking cases
only; otherwise the actual prevention condition in their modal definitions
kicks in.
If this is right, it is false to say that the only type of possibility available
to the Stoics is epistemic. Consider an example of Aristotle’s: a new cloak
might perish as a result of ordinary wear, or it might be cut. For the Stoics,
sub specie aeternitatis there is only one thing that can happen to it – the
unravelling of fate will see to that. However, there is nothing now in the
world that prevents either outcome, for no causally e√cient state of
a◊airs is now making it the case that it will (or will not) be cut. There is
thus a point to Chrysippus’ insistence that fate is an ineluctable chain of
antecedent causes.
528 determinism and indeterminism
24 On the notion of a ‘perfect’ cause, see above, pp. 488–90.
Cambridge Histories Online © Cambridge University Press, 2008
vi The Chrysippean notion of fate: soft
determinism
Cicero reports that
Between the two views held by the old philosophers, one being that of
those who held that everything takes place by fate in the sense that fate
exercises the force of necessity – the view of Democritus, Heraclitus,
Empedocles, and Aristotle, the other that of those who said that the
movements of the mind are voluntary and not at all controlled by fate,
Chrysippus stands as an honorary arbiter and wished to strike a mean
between the two; though he leans rather towards those who hold that
mind is free from all necessity of motion . . . none the less he slips into
such di√culties that against his will he lends support to the necessity of
fate. (Cic. Fat. 39)
Cicero has Chrysippus impaled on the horns of a dilemma. On the one
hand reluctant to abandon proposition (4) above,25 he nevertheless
wishes to a√rm universal causal determinism. To drive a wedge between
the two positions, he attempts to disengage assent, sunkatathesis, from the
grip of fate. Since
those old philosophers who used to say that everything takes place by
fate held that assent is given by force and necessity. But those who disagreed
with them released assent from fate and denied that if assent were
tied to fate it would be possible to disentangle it from necessity. They
argued as follows: if (9) all things occur by fate, all things occur by an
antecedent cause; (10) and if desire is caused, those things which follow
desire are also caused; therefore (11) assent is also caused. But (12) if the
cause of the desire is not situated within us, even desire itself is not in our
power; (13) and if this is so, those things which are caused by desire do
not rest with us. Thus it follows (14) that neither assent nor action are in
our power. Hence (15) there is no justice in either praise or blame, honours
or punishments. (Cic. Fat. 40)
We have seen the bulk of this argument elsewhere. It has (for the antideterminist)
the form of a multiple modus tollens: deny (15), and you are
committed to denying the antecedent of (9). By contrast, a hard determinist
will take the truth of (9) to entail (15), and the emptiness of conventional
morality. Chrysippus tries to avoid either conclusion, impugning
the argument’s validity by distinguishing antecedent causes from internal
causes, as we saw in the last chapter.26
the chrysippean notion of fate 529
25 See p. 520. 26 See above, pp. 490–4.
Cambridge Histories Online © Cambridge University Press, 2008
Chrysippus does not deny that there are antecedent causes of our
desires, and hence (since such causes are transitive) of our assents to them:
‘assent cannot take place unless prompted by a sense-impression’ (Cic.
Fat. 42). But antecedent causes are not of themselves determining. They
determine only in conjunction with other factors, the internal states and
conditions they operate upon: and these causes are not transitive in form.
Thus our assents are caused, but not determined. Chrysippus accepts (11)
in a sense – and equally in a sense he accepts (12): but he holds that they
equivocate on the notion of cause at play in each of them. In no sense of
‘cause’ for which (11) is true will the antecedent of (12) follow; hence the
argument fails.
But this is immediately puzzling: while elsewhere it is indeed confirmed
that Chrysippus identifies fate with antecedent causes (Cic. Top.
59; cf. [Plu.] Fat. 574d). However, Plutarch (Stoic. Rep. 1056b–c) complains
that if Chrysippus does make fate merely the antecedent, and not
the perfect cause of right action and thought, he will contradict himself,
since the antecedent cause is supposedly weaker than the perfect, yet
nothing is more powerful than the will of Zeus (with which Chrysippus
wants to identify fate). Furthermore, fate is supposed to be unconquerable,
ineluctable, and unavoidable.
Sorabji distinguishes three ways in which Chrysippus’ argument may
be taken.27 According to the first (to which he inclines), Chrysippus is
trying to avoid the necessitation of assent – but it is, as Sorabji admits,
hard to see how the argument could begin to show this with any generality.
The second view has Chrysippus making a point about moral responsibility
rather than necessity – if our actions derive partially from
something internal to us, then we are responsible for them.28 But then it
is hard to see how the argument goes; for, as we have noted, Aristotle was
already well aware that to show that our dispositions and so on are
responsible for our actions in this sense will still not establish any genuine
responsibility. Finally, it might be that Chrysippus wants to deny not
necessity as such, but the necessity of fate.29 This accords well with
Cicero’s language, and with the view that fate is a sequence of external
causes.
If this is right, then there will be no relaxation of the fetters of necessity
– but fate in and of itself will not be the sole producer and determiner of
that necessity. Even so, it is hard to see how the Stoics could have thought
530 determinism and indeterminism
27 Sorabji 1980b. 28 Donini 1974–5.
29 Sorabji 1980b, 274, attributes this interpretation to Frede.
Cambridge Histories Online © Cambridge University Press, 2008
of fate on its own, if it is simply the sequence of antecedent causes, as
being determining – for antecedent causes do not themselves determine.
It is only when they are combined with the potentials contained within
things in temporal and causal sequences that the determination of events
arises at all. However, there will be another sense in which fate is determining:
given the structure of natural potentialities which obtains in the
world, plus various contingent (contingent in a relaxed sense: nothing
turns on the notion here) facts about their physical and spatial relationships,
then their entire history of interaction will be plottable by a
Laplacean super-scientist – their impingement upon one another being
precisely the play of antecedent causes. Thus, looking backward, we can
say that fate necessitated the outcome of every event given that spatiotemporal
structure.
We can still give a perfectly clear and coherent sense to the claim that
things might have turned out otherwise: they might, indeed would, have
done so had that structure, per (causal) impossibile, been di◊erent. If it had
been you rather than I who was tempted by that extra helping of
zabaione, the pudding would have remained virtuously on the plate. And
the truth-makers for these counterfactuals of possibility will be realworld
situations that are relevantly similar in all important aspects (last
week you were o◊ered a second helping of zabaione, but virtuously
declined). Hence possibility does not collapse into necessity; and nor does
it become merely epistemic in form.
Thus Plutarch’s charge of self-contradiction also fails. Considered
simply qua collection of antecedent causes, fate is not all-powerful; but
there is a perfectly clear real-world sense in which, given the way things
are, things could not turn out di◊erently given that set of causes. The will
of Zeus really is ineluctable.
vii Fate and responsibility: confatalia and the
eph’ he–min
One may still ask where this leaves the concept of responsibility. We may
allow Chrysippus the distinction between internal and external causes –
we may even admit that, in some metaphysical sense, the inner causes
really are the causes that keep the system going. But even if that shows
that in some sense the things we do are eph’ he–min, ‘up to us’, in that it is
the structure of our individual desires and beliefs and so on which causally
determines how we react individually to stimuli (as in the zabaione case),
is that sense strong enough to justify the ascription of praise and blame?
fate and responsibility 531
Cambridge Histories Online © Cambridge University Press, 2008
While my gourmandise is evidently responsible for my custardly akrasia
(and equally your self-control conditions your restraint under the same
circumstances), am I responsible for the state of my dispositions?
This is of course Aristotle’s problem: only if I am responsible for the
development of my states of character can I be held responsible for the
excesses of that character. Aristotle thought, and Epicurus appears to
have echoed him here,30 that there was a stage in the development of our
characters, before the dispositions hardened into hexeis, where they were
su√ciently elastic for us to be able to influence which way they went. We
become good, on the Aristotelian model, by performing good actions, by
making them a part of our make-up (EN ii.2.1104a11–4.1105b18). But
while that process is under way we are genuinely free to choose either the
straight-and-narrow or the primrose path. So, while it may be true of the
old reprobate that he genuinely cannot now refrain from vice, the young
blade he was thirty years ago could have done so. There is a window of
opportunity, as it were, in the development of our characters when that
development is up to us.
This is not easy to make sense of; when I ‘freely’ choose vice over virtue,
why assume that my choice is somehow free of the constraints which later
condition my vicious life-style? The fashionable distinction between firstand
second-order desires does no work here (was not my second-order
choice to develop a vicious disposition itself the result of existing traits of
character, or external pressures?). Moreover, Aristotle stresses the importance
of a proper upbringing in developing a good character – but that is
surely definitionally outside the control of the individual properly
brought up.
Su√ce it to say here that the mere fact that the agent’s desires, beliefs,
etc. play an instrumental role in causing his actions does nothing to support
the view that what he does is up to him in any strong sense. It does
show that such actions are attributable to him – it is he who does them, in
a sense in which those ‘actions’ of someone physically compelled by main
force are not. That Humean response will deflect the Epicurean accusation
that anyone who espouses this kind of determinism can make no
sense of the notion of coercion (although the subtler point that there is no
532 determinism and indeterminism
30 Nat. xxv, [34] 26 Arr.2: ‘from the very outset we always have seeds directing us some towards
these, some towards those, some towards these and those, actions and thoughts and characters,
in greater and smaller numbers. Consequently that which we wish to develop – characteristics
of this or that kind – is at first absolutely up to us; and the things which of necessity flow in
through our passages from that which surrounds us are at one stage up to us and dependent
upon beliefs of our own making.’ Quite what Epicurus’ account is here is di√cult to determine
– but the overall tenor of his developmental view is unmistakeable.
Cambridge Histories Online © Cambridge University Press, 2008
morally relevant di◊erence between coercion and voluntary action may of
course still be defensible).31 But the thornier problem remains of explaining
how ascription of responsibility is to be justified if everything that
happens is indeed necessitated by fate, even if an individual’s actual dispositions
are part of the causal nexus that determines those outcomes. The
Stoics held that we were responsible for what we did because our natures
were such as to determine how we chose – our natures, that is, are partially
(and for evaluative purposes relevantly) causally responsible for our
making the decisions we make, and hence our decisions are ‘up to us’.32
But the rejoinder is obvious:
If they assign impulse as being up to us on the grounds that we have it by
nature, what is to stop us from saying that burning is up to fire since fire
burns by nature? (Nemes. 106.7–9)
The Stoics seem to have thought that, the more convoluted the internal
causal process leading from impression to action in an individual, the
more attributable to that individual becomes the action. But it is hard to
see how a mere increase in complexity can deliver the required result, at
least if responsibility, like causation, is taken to be transitive.33
Hard determinism would be an option – but as Sorabji stresses, it was
not one much canvassed in antiquity.34 There is a famous story about Zeno:
He was once beating a slave for stealing, and when the latter said ‘I was
fated to steal’, he replied ‘and to be flogged’. (D.L. vii.23)
But, as Sorabji says, while compatible with a hard determinist stance this
by no means entails it: and it is clearly intended rather to illustrate the
Stoic doctrine of ‘co-fated events’ (confatalia).
A standard objection to Stoic determinism was that known as the ‘lazy
argument’, the argos logos; if everything is preordained since time began,
what is the point of my making any interventions in the world at all?
Cicero again:
fate and responsibility 533
31 It is worth pointing out that such a response can also make room for a notion of coercion richer
and more interesting than simply that of main force – an action (handing over money to an
armed robber, for instance) is coerced in this sense just in case the agents’ desires are not
directly involved in their coming to be in a situation in which their freedom of action (i.e. the
set of options over which their desires can range) is curtailed by the actions of another (in this
case the robber o◊ering the alternatives of ‘your money or your life’). I owe these points to
numerous discussions with my wife, Jennifer Greene.
32 See e.g. Alex. Fat. 181.3–182.20; 205.24–206.2.
33 Of course, one may deny that responsibility is, in the appropriate sense, thoroughly transitive –
complex systems will, on this view, take on ‘a life of their own’, and serve as an appropriate locus
for evaluation – but it is one thing to state this position – quite another to argue for it.
34 Sorabji 1980b, 280–2. An exception is Galen: see Hankinson 1992, which deals in greater detail
with the issues of causation, determination, and responsibility.
Cambridge Histories Online © Cambridge University Press, 2008
They pose it [i.e. the lazy argument] thus: ‘if it is your fate to recover
from illness, you will recover, regardless of whether or not you call the
doctor. Similarly if it is your fate not to recover . . . you will not recover
whether or not you call the doctor. And one or the other is your fate.
Therefore it is pointless to call the doctor.’ (Cic. Fat. 28–9)
The lazy argument confounds determinism with what one might call
Islamic fatalism, the notion that no matter what you do a particular fate is
in store for you. By contrast genuine determinists must reject this – they
should (and the Stoics did) say that whatever we do is predetermined,
including our reaction to the lazy argument. Here is Chrysippus’ response:
Some events in the world are simple, some complex. ‘Socrates will die on
such and such a day’ is simple: his day of dying is fixed, regardless of what
he may or may not do. But if a fate is of the form ‘Oedipus will be born to
Laius’, it will not be possible to add ‘regardless of whether or not Laius
has intercourse with a woman’. For the event is complex and co-fated
(confatale). He [i.e. Chrysippus] uses this term because what is fated is
both that Laius will have intercourse with his wife and that by her he will
beget an Oedipus . . . All fallacies of this sort are refuted in the same way.
‘You will recover regardless of whether or not you call the doctor’ is fallacious.
For it is just as much fated for you to call the doctor as for you to
recover. (Cic. Fat. 30)
Elsewhere, Chrysippus employed a di◊erent, pointed example – even if it
is fated for Hegesarchus to win the bout without taking a punch, it would
still certainly be absurd to expect him to fight with a dropped guard on
the grounds that he was so fated (SVF ii 998). In general, the doctrine of
confatalia has it that if some event E is fated, then so are all the necessary
conditions of E. Thus I eat the zabaione because I am greedy – I am fated
not only to eat it, but to be greedy too. And my being greedy explains why
I eat it. My actions are not robbed of point simply because in one sense at
least I could not do otherwise. Persistent acratics can still try to mend
their ways in a thoroughly determined universe; and they may still succeed
through their own e◊orts (although of course the fact that they make
the e◊ort is determined too). But whether or not they do so will be determined
by causes remote from their own control.
viii Divination and fate
Whether or not the universe is determinist in character is, as we saw earlier,
35 logically distinct from the question of whether that future can be
534 determinism and indeterminism
35 See p. 519 above.
Cambridge Histories Online © Cambridge University Press, 2008
known – but if one assumes, as the ancients in general did, that the future
can be known only if there are determinate causes of it knowable in the present,
then the possibility of large scale accurate forecasting of the future will
seem to depend on that future’s beingdetermined. The Stoics, by and large
(Panaetius was an exception) believed in divination; and they found that
belief comfortingly compatible with their determinism. Indeed their views
of the relations between the two were sometimes charged with circularity:
Chrysippus gives this demonstration to us, proving each one by way of
the other. For he wants to establish that everything comes to be according
to fate on the grounds of divination, while that divination exists he is
able to show by no other means than by assuming that everything comes
about according to fate. (Diogenianus ap. Eus. PE iv.3.2)
The circularity is not, however, vicious; and our text confuses explanation
with support. The (supposedly) empirical fact of divinatory success supports
the hypothesis of determinism (indeed perhaps on their, mistaken,
view it entails it); conversely, the deterministic hypothesis explains divination,
or at any rate is part of its explanation.
Of course, the mere fact that the universe is deterministic would be no
guarantee that its future course would be patent to the miserable human
intelligence. Perhaps it is just vastly too complex for that. In that case, it
will help to have a benevolent deity who is au fait with those complexities
– but that there is such a deity is a feature of Stoic theology. Thus it would
be good for us if we could know the future; the gods (or God) can tell us
what it will be like; they have concern for us; hence they will tell us (cf.
Cic. Div. i.101–2; ND ii.161–8). Of course, that argument is vulnerable at
every turn – and the Stoics’ ancient critics exploited that vulnerability (cf.
Carneades, in Cic. ND i.4; Favorinus ap. Gell. xiv.1–36).36 But for all that,
the Stoics have a consistent set of at least mutually supportive doctrines in
theology and metaphysics.
The ins and outs of the debate on divination are beyond the scope of
this chapter.37 But one feature of the dispute is of importance. The Stoics
defined divination as ‘the foretelling of events that come about by chance’
(Cic. Div. ii.13–15, 26): but in the Stoic universe there is no room for
chance. Indeed, the Stoics’ stubborn refusal to admit that there is such a
thing as genuine chance forms the core of the Peripatetic attack on Stoic
determinism.38 Cicero indeed attempts to convict the Stoics of a formal
divination and fate 535
36 On the Stoics and divination in general see Hankinson 1988b, and Long 1982a.
37 Arguments both for and against may be found in Cic. Div.; and S.E. M v is a compendium of
arguments, many probably Carneadean in origin, against divination.
38 To be found principally in Alex. Fat.; I do not deal with the Peripatetic views in detail in their
own right, since it is not certain that they were developed in the period covered by this volume.
None the less, they are of much intrinsic interest, see Sharples 1983.
Cambridge Histories Online © Cambridge University Press, 2008
self-contradiction on this score. The future cannot be predicted if it is the
result of chance, since if we can know an event is going to occur, it is not
now possible that it will not occur; hence it cannot be a chance event.
Furthermore, Cicero argues, the only basis for any such knowledge would
have to be causal – the event in question follows in a lawlike manner from
known initial conditions. But in that case too, it cannot be chance. But the
Stoics have no di√culty with this, for they specifically define chance epistemically
as ‘a cause obscure to human understanding’.39 To describe an
event as a matter of chance is to say that it was not predictable by us on the
basis of known causal laws and initial conditions. Of course, that is quite
compatible with the idea of its being causally predictable by some superintelligence;
and hence of such events being foreshadowed for us in some
other manner (by the cleft in an ox-liver, for instance).
What is at issue here is the nature of the divinatory sign. Sign-theory and
its ramifications are fully treated elsewhere in this volume.40 But crudely
the issue is this. In the case of a certain type of sign, the so-called indicative
sign, the sign-event is such that it is more than materially tied to that which
it signifies. The sign is a sign in virtue of a determinate causal relationship
which holds between it and what it signifies (further conditions also need
to be fulfilled – but they need not concern us). By contrast, a commemorative
sign merely serves to call to mind that of which it is a signifier: there
may be, but crucially need not be, any causal connection between the two.
This distinction lies behind Chrysippus’ famous attempt to evade the
unpalatable implications of treating divinatory ‘theorems’ as conditionals
in the Stoic sense. If
(T5) whoever is born at the rising of the Dog-star will not die at sea
is such a theorem, and Fabius was born at that time, then
(16) if Fabius was born at the rising of the Dog-star, Fabius will not die at
sea
looks as though it is a sound conditional, a simple substitution instance of
(T5). But, given the Stoics’ own account of its truth-conditions,41 for (16)
to be sound there must be a connection of relevance (it is tempting to say
causal relevance) between antecedent and consequent – the consequent is
true in virtue of the antecedent’s truth. But (16) does not appear to satisfy
that condition, since Fabius’ being born at that time does not seem to be a
536 determinism and indeterminism
39 Alex. Fat. 172.12, 173.13, 174.1; SVF ii 965–71.
40 See above, pp. 286–94 and below, pp. 611–3. 41 See above, pp. 106–8.
Cambridge Histories Online © Cambridge University Press, 2008
cause of his avoiding a watery grave. Hence Chrysippus reformulated
such propositions as negated conjunctions:
(16*) it is not the case both that Fabius is born at the rising of the Dog-star
and that he dies at sea.
Cicero, who reports this (Fat. 15), also ridicules it – why cannot doctors
and geometricians simply do the same with their own theorems:
What is there that cannot be carried over in that sort of way from the
form of a necessary consequence to that of a negated conjunction? (Cic.
Fat. 16)
But he misses the point, which is that such reformulations carry no commitment
to there being any necessary relationship between the two component
propositions (or perhaps more accurately if there is a necessary
relation between them it is derivative, not direct). This does not imply
that there can be no causal relation between them (they may, for example,
be collateral e◊ects of some more remote cause), although for reasons
briefly canvassed earlier it does not seem that there need be. Thus,
Cicero’s jocularity notwithstanding, Chrysippus is making a serious
point with his plea for reformulation.
ix Soft determinism
None of the Stoics wished to invoke their determinism to exonerate the
wrong-doer; indeed, their stern morality lays great store by the individual’s
own e◊orts at self-improvement and moral progress. But the suspicion
still remains that, for all Chrysippus’ attempts to find a middle way
between hard determinism and the causal chaos of the Epicurean swerve,
there is something missing in the determinist’s universe, namely the justification
for punishment and reward – the sense in which things are, for
them, eph’ he–min is not strong enough to bear the weight of moral evaluation.
To be sure, the determinist can explain why they occur; perhaps he
can even do so in terms of their beneficial e◊ects (by developing an evolutionary
theory of them, for instance). Crimes may still be punished pour
encourager les autres; deterrence works, indeed perhaps works best, in a
determinist cosmos. Furthermore, a determinist is not debarred by his
views from attempting to reform a criminal, or from protecting society
from him. But the notion of desert seems to be in bad order – and the
ancients were not prepared readily to abandon it. Let us conclude by
examining some texts which bear upon the issue.
soft determinism 537
Cambridge Histories Online © Cambridge University Press, 2008
The Stoics adduce a number of arguments in order to defend their
thesis of the compatibility of moral judgement with their form of determinism.
First of all they argue that, since such judgements arise of necessity
out of the natural order, they must themselves be natural – it is part of
our natures to praise and blame. Man is rational, mortal – and censorious:
They [i.e. the Stoics] suppose that everything naturally constituted is
such as it is in accordance with fate, ‘natural’ being the same thing as ‘in
accordance with fate’, and they add ‘consequently it will be in accordance
with fate that animals have perceptions and impulses. And some
animals will be merely active, while others will perform rational actions.
And some will do wrong, while others will do right actions. For these are
natural to them. But so long as wrong and right actions remain, and their
natures and qualities are not removed, there also remain commendations
and censures, punishments and honours. For such are the sequence and
order to which they are subject.’ (Alex. Fat. 205.24–206.2; cf. 207.5–21)
Right and wrong actions are just as much written into the causal sequence
of things as anything else; and so, consequently, are their consequences.
Zeno’s slave has no right to complain. This raises some immediate questions.
Alexander himself asks pertinently how the concepts of right and
wrong can retain any content in the Stoic universe. It is precisely because,
he avers, the actions which invite moral appraisal are up to us in the sense
of not being externally compelled that we can meaningfully label them
right or wrong. But such a position is not open to the Stoics, since whatever
they say, all actions are compelled (Alex. Fat. 206.2–207.3).
Alexander e◊ectively challenges the Stoics to show not merely how
such institutions might be explained, but how they can retain their moral
value. The Stoics can, he allows, give a natural history of their genesis –
but to give a natural historical explanation is precisely to explain them
away, to show how they can have the appearance of content without actually
possessing any: but that is hard, and not soft, determinism.
The Stoics cannot, I think, avoid accepting some shift in the content of
these moral notions. It is a further question how damaging that shift need
be, and whether it need empty them of all recognizable evaluative content.
The following is a sketch only of what the Stoics need to say on the
issue (whether they actually did or not is a further question). They borrow
from Aristotle the notion of natural hierarchy of functions; and like him
they interpret this in a teleological fashion (the di◊erences between
Aristotelian and Stoic teleology need not detain us). Di◊erent animals do
di◊erent things: and what they do uniquely or best is their proper or
definitive function. It is their nature for them to act thus; and hence it is
538 determinism and indeterminism
Cambridge Histories Online © Cambridge University Press, 2008
right for them to do so. Man censures; hence it is right for man to be censorious.
It is not di√cult, however, to see where that argument is vulnerable.
First of all, the Stoics can be accused of an equivocation on the notion of
rightness here. Perhaps ‘right’ just means ‘fitting’ or ‘appropriate’; in that
sense it is clearly right for a carving-knife to slice meat, and wrong to use
it to slice flesh. And, at a stretch, you might censure people who do so for
misusing their tools. But it would be absurd to censure the knife itself,
since its capacity is merely instrumental; and yet it is di√cult to see how,
in the Stoic cosmos, human beings have any morally richer role simply in
virtue of their structural complexity.
But the Stoics have a further line of defence. Divine providence sees
that the world is a hierarchy not merely of functions but of good ones. But
if that is the case, surely we can view the gradual approximation of the
world to a perfect condition as being itself a good thing. Thus there is a
further sense over and above their mere functionality why it is good that
knives cut, or men blame: it is part of the benevolent, providential ordering
of things; God has made things as good (compatible with the material
restraints imposed upon him: Epict. Diss. i.1.7–12, following Plato Tim.
29e–30b; 75a–c) as they could possibly be.42 But the Stoics notoriously
held that everything that occurred was good (see e.g. Marc. Aurel. iv.10,
23, 26; cf. ii.3; and cf. Chrysippus’ attitude to bed-bugs);43 consequently,
it is not a mere fact that we praise and blame – it is good that we do so;
praise and blame are part of the causal working of the best of all possible
worlds. But if that is right, how can the Stoics di◊erentiate between good
and bad actions? Since the world is as good as it could be (sc. at this particular
time), and since its evolution is towards a state of perfection, and
since every event that occurs within it is part of and contributes towards
that evolution, every event must be good.
The Stoics are in some sense committed, I believe, to that. But they are
not debarred thereby from developing a non-trivial conception of the relative
rightness and wrongness of individual actions and events which
relies on the distinction between temporally-indexed and timeless judgements
of worth. Let us consider some action that would be considered
wrong, both conventionally and by the Stoics (for the Stoics notoriously
do not subscribe to a conventional morality: S.E. PH iii.200–1, 205–6,
soft determinism 539
42 The limitations on creative possibility supplied by material recalcitrance common to the
Platonic, Stoic, and Galenic view of providential creation allows them to sidestep some of the
notorious di√culties associated with maintaining that ours is the best of all possible worlds: see
Hankinson 1989. 43 See above, pp. 467, 504.
Cambridge Histories Online © Cambridge University Press, 2008
etc.): premeditated murder, for example. A murder is an event, and hence
datable. Suppose x kills y at spatio-temporal region s. We may consider
the action either as (a) a murder tout court, or as (b) a murder-at-s. Other
things being equal, murder is wrong; hence sub specie (a) we condemn it;
however viewed sub specie (b), it is an event conducive towards the perfection
of the world, and hence we can welcome it, not because it is an action
of the type that it is (that is still reprehensible), but simply because it is so
conducive. We can even add a little further flesh to that: the truth-maker
for the claim that the action is wrong is the fact that when the world
attains its most developed condition there will no (a)-type acts, which
allows us to evaluate (b) the way we do.
Thus the Stoics can rescue a non-trivial sense in which actions can be
genuinely evaluated. But of course it is one thing to evaluate the act –
quite another to evaluate the agent as being responsible. Here, I think, the
opponents of the Stoics can make out their case. As a Stoic, I can think
that you are a frightful bore; I can wish it were otherwise; and I can look
forward to the blessed day when the world will contain only interesting
people. But it seems unfair to hold you any more responsible for that than
you are for your basic physical nature.
x Fate and moral progress
Finally I turn to the issue of how the Stoic view of moral progress relates
to their determinism. Earlier I distinguished determinism of the Stoic
type from fatalism, the view that no matter what you do, your fate is
bound to be thus and so. The Stoics do, however, appear at times to stray
close to a position that resembles at least a limited fatalism. This is most
apparent in their famous comparison of human fate to a dog tied behind a
cart – it can choose to go willingly along, or it can choose to resist: but
either way the end result is the same (Hipp. Ref. i.21). That appears to suggest
that, contrary to the burden of the argument so far, the Stoics conceive
of human beings as having a sort of Humean liberty of spontaneity.
They cannot choose how things will be; but they can choose, and apparently
in some strong sense, whether or not to like it. Thus human choices
are in a sense genuine – but they are causally insulated from the working
of the world.
But if what has gone before is even remotely correct, that cannot be
right. The dog and cart image is not a particularly happy one, precisely
because it does have these fatalistic overtones: none the less it can be interpreted
consistently with a non-fatalist view of the structure of causation.
540 determinism and indeterminism
Cambridge Histories Online © Cambridge University Press, 2008
The point is primarily, I think, about our attitude to our projects. The
Stoics held that, until we actually arrive at the condition of the sage, in
which we will simply never make mistakes, every expression of projected
action should be hedged with a mental rider, a ‘reservation (huphexairesis)’
to the e◊ect ‘God willing’. I should want to do things only on condition
that they will as a matter of fact take place (and hence by definition are part
of the unravelling of fate). What the Stoic aims to do is to bring his own
impulses and desires as closely as possible into harmony with the way
things are actually going to go; he will not strive for the impossible. Now
that can be interpreted fatalistically, if one assumes that the way things are
going to go is fixed independently of human decisions and desires. But it
does not require such an interpretation; and all that we have seen of the
Stoic position so far tells against it. As I approach moral perfection, the
extent to which my desires will be frustrated will diminish. In the perfect
world there will be no such frustrations at all. And the Stoics hold that the
world is evolving towards such perfection. Of course its evolution is determined
– and part of that evolution is driven by human desires and their
frustrations. But, as we have seen, we can make sense of the claim that it is
good that there are such frustrations even though frustrations are not
good. That position is coherent, and non-fatalistic (it does not assert that
things would have been the same whatever decisions were made). Whether
there is anything else to be said for it is, however, another matter